Introduction To Genetic Analysis (12th Edition)
Introduction To Genetic Analysis (12th Edition)
Introduction To Genetic Analysis (12th Edition)
this cornerstone textbook to the next level. The hallmark focus on ge- Doebley
netic analysis, quantitative problem solving, and experimentation con-
tinues in this new edition. Peichel
Wassarman
The twelfth edition also introduces SaplingPlus, the best online
resource to teach students the problem-solving skills they need to
succeed in genetics. SaplingPlus combines Sapling’s acclaimed
GENETIC ANALYSIS
INTRODUCTION TO
automatically graded online homework with an extensive suite of
engaging multimedia learning resources.
ORDERING OPTIONS:
Hardcover + SaplingPlus
6-month access 978 -1-319-34103-9 INTRODUCTION TO
GENETIC
12-month access 978 -1-319-34106-0
Loose-leaf + SaplingPlus
6-month access 978 -1-319 -34109-1
12-month access 978-1-319-34114-5
ANALYSIS
TWELFTH EDITION
TWELFTH
EDITION
Cover images: Butterflies photo: Courtesy of
Anthony J. F. Griffiths
John Doebley
Mathieu Joron; DNA art: Emiko Paul
Catherine Peichel
macmillanlearning.com
David A. Wassarman
Contents in Brief
PREFACE xii
2 Single-Gene Inheritance 29
3 Independent Assortment of Genes 79
4 Mapping Eukaryote Chromosomes by Recombination 113
5 Gene Interaction 153
6 The Genetics of Bacteria and Their Viruses 193
GLOSSARY 736
INDEX 770
Genetic Analysis
Anthony J. F. Griffiths
University of British Columbia
John Doebley
University of Wisconsin–Madison
Catherine Peichel
University of Bern
David A. Wassarman
University of Wisconsin–Madison
I n 1 9 4 6 , Wi l l i a m F re e m a n
Printed in the United States of America founded W. H. Freeman and
Company and published Linus
1 2 3 4 5 6 24 23 22 21 20 19 Pauling's General Chemistry, which revolutionized the chemistry
curriculum and established the prototype for a Freeman text.
W. H. Freeman quickly became a publishing house where
leading researchers can make significant contributions to
Macmillan Learning
mathematics and science. In 1996, W. H. Freeman joined
One New York Plaza Macmillan and we have since proudly continued the legacy of
Suite 4600 providing revolutionary, quality educational tools for teaching
New York, NY 10004-1562 and learning in STEM.
www.macmillanlearning.com
PREFACE xii
2 Single-Gene Inheritance 29
3 Independent Assortment of Genes 79
4 Mapping Eukaryote Chromosomes by Recombination 113
5 Gene Interaction 153
6 The Genetics of Bacteria and Their Viruses 193
GLOSSARY 736
INDEX 770
2.4 Some Genes Discovered by Observing 4.2 Mapping by Recombinant Frequency 119
Segregation Ratios 45 Map units 119
A gene active in the development of flower color 46 Three-point testcross 122
A gene for wing development 46 Deducing gene order by inspection 123
A gene for hyphal branching 47 Interference 124
Predicting progeny proportions or parental genotypes by Using ratios as diagnostics 129
applying the principles of single-gene inheritance 47
4.3 Mapping with Molecular Markers 130
2.5 Sex-Linked Single-Gene Inheritance Patterns 48
Sex chromosomes 48
4.4 Using the Chi-Square Test to Infer Linkage 131
Sex-linked patterns of inheritance 49 4.5 The Molecular Mechanism of Crossing Over 132
X-linked inheritance 49
4.6 Using Recombination-Based Maps in
2.6 Human Pedigree Analysis 52 Conjunction with Physical Maps 134
Autosomal recessive disorders 53
Autosomal dominant disorders 55 5 GENE INTERACTION 153
Autosomal polymorphisms 56
X-linked recessive disorders 58 5.1 Interactions Between the Alleles of a Single Gene:
Variations on Dominance 154
X-linked dominant disorders 60
Complete dominance and recessiveness 154
Y-linked inheritance 60
Incomplete dominance 156
Calculating risks in pedigree analysis 60
vii
Analyzing double mutants of random mutations 166 8.1 RNA Structure 269
RNA is the information-carrying intermediate between DNA and
proteins 269
6 THE GENETICS OF BACTERIA
Consequences of the distinct chemical properties of RNA 270
AND THEIR VIRUSES 193 Classes of RNA 270
6.1 Working with Microorganisms 195 8.2 Transcription and Decay of mRNA in
6.2 Bacterial Conjugation 197 Bacteria 271
Discovery of conjugation 197 Overview: DNA as transcription template 272
Discovery of the fertility factor (F) 199 Stages of transcription 273
Hfr strains 200 mRNA decay in bacteria 276
Mapping of bacterial chromosomes 203 8.3 Transcription in Eukaryotes 277
F plasmids that carry genomic fragments 206 Transcription initiation in eukaryotes 277
R plasmids 206 RNA polymerase II transcription elongation 281
6.3 Bacterial Transformation 208 Transcription termination in eukaryotes 281
The nature of transformation 208 8.4 Processing of mRNA in Eukaryotes 283
Chromosome mapping using transformation 209 Capping 283
6.4 Bacteriophage Genetics 209 Polyadenylation 284
Infection of bacteria by phages 209 The discovery of splicing 284
Mapping phage chromosomes by using phage crosses 211 The splicing mechanism 284
snRNAs in the spliceosome may carry out the catalytic steps of
6.5 Transduction 213 splicing 287
Discovery of transduction 213 Alternative splicing can expand the proteome 288
Generalized transduction 213 RNA editing 290
Specialized transduction 215 RNA nucleotide modification 290
Mechanism of specialized transduction 216 RNA export from the nucleus 291
6.6 Physical Maps and Linkage Maps Compared 217 8.5 Decay of mRNA in Eukaryotes 291
mRNA decay mechanisms 292
The discovery of RNA interference (RNAi) 292
PART II CORE PRINCIPLES IN MOLECULAR AND siRNA-mediated RNA decay and transcriptional silencing 294
DEVELOPMENTAL GENETICS 232 RNAi protects the genome from foreign DNA 295
7.3 DNA Replication Is Semiconservative 249 9.3 tRNAs and Ribosomes 310
Evidence that DNA replication is semiconservative: tRNAs are adaptors 310
the Meselson–Stahl experiment 250 Wobble base pairing allows tRNAs to recognize more than one
Evidence for a replication fork: the Cairns experiment 250 codon 312
Ribosome structure and function 313
7.4 DNA Replication in Bacteria 251
Unwinding the DNA double helix 251 9.4 Translation 315
Assembling the replisome: replication initiation 252 Translation initiation 315
DNA polymerases catalyze DNA chain elongation 253 Translation elongation 318
DNA replication is semidiscontinuous 254 Translation termination 319
DNA replication is accurate and rapid 255 Nonsense suppressor mutations 320
11.3 Catabolite Repression of the lac Operon: 13.4 Spatial Regulation of Gene Expression in
Positive Regulation 379 Development 442
The basics of lac catabolite repression: choosing the best sugar Maternal gradients and gene activation 442
to metabolize 379 Drawing stripes: integration of gap-protein inputs 443
The structures of target DNA sites 381 Making segments different: integration of Hox inputs 444
A summary of the lac operon 382
13.5 Post-transcriptional Regulation of Gene
11.4 Dual Positive and Negative Regulation: Expression in Development 448
The Arabinose Operon 383 RNA splicing and sex determination in Drosophila 448
Regulation of mRNA translation and cell lineage in C. elegans 449
11.5 Metabolic Pathways and Additional Levels of
Regulation: Attenuation 384 Translational control in the early embryo 449
miRNA control of developmental timing in C. elegans and other
11.6 Bacteriophage Life Cycles: More Regulators, species 452
Complex Operons 386
13.6 From Flies to Fingers, Feathers, and Floor Plates:
Regulation of the bacteriophage λ life cycle 386
The Many Roles of Individual Toolkit Genes 453
Molecular anatomy of the genetic switch 389
Sequence-specific binding of regulatory proteins to 13.7 Development and Disease 455
DNA 391 Polydactyly 455
11.7 Alternative Sigma Factors Regulate Large Sets of Holoprosencephaly 455
Genes 392 Cancer as a developmental disease 456
15.1 Molecular Consequences of Point Mutations 503 Applications of inversions and translocations 584
15.3 Molecular Basis of Induced Mutations 511 Single nucleotide polymorphisms (SNPs) 605
Identifying mutagens in the environment: the Ames test 514 Haplotypes 606
Other sources and forms of variation 608
15.4 DNA Repair Mechanisms 515
Direct repair of damaged DNA 516 18.2 The Gene-Pool Concept and the Hardy–Weinberg
Base excision repair 516
Law 609
Nucleotide excision repair 517 18.3 Mating Systems 613
Mismatch repair 519 Assortative mating 613
Translesion synthesis 520 Isolation by distance 614
Repair of double-strand breaks 521 Inbreeding 614
The inbreeding coefficient 615
16 THE DYNAMIC GENOME: Population size and inbreeding 617
TRANSPOSABLE ELEMENTS 527 18.4 Genetic Variation and Its Measurement 618
16.1 Discovery of Transposable Elements in Maize 529 18.5 The Modulation of Genetic Variation 621
McClintock’s experiments: the Ds element 530 New alleles enter the population: mutation and migration 621
19.2 A Simple Genetic Model for Quantitative Traits 648 20.5 Evolution of Traits 697
Genetic and environmental deviations 649 Adaptive changes in a pigment-regulating protein 697
Genetic and environmental variances 649 Gene inactivation 698
Correlation between variables 651 Regulatory-sequence evolution 699
Loss of characters through regulatory-sequence evolution 700
19.3 Broad-Sense Heritability: Nature versus
Regulatory evolution in humans 702
Nurture 652
Measuring heritability in humans using twin studies 653 20.6 Evolution of Species 703
Species concepts 703
19.4 Narrow-Sense Heritability: Predicting
Mechanisms of reproductive isolation 703
Phenotypes 655
Genetics of reproductive isolation 704
Gene action and the transmission of genetic variation 656
The additive and dominance effects 656
A model with additivity and dominance 657
Narrow-sense heritability 660
A BRIEF GUIDE TO MODEL ORGANISMS 713
Predicting offspring phenotypes 661
Selection on complex traits 662 APPENDIX A: Genetic Nomenclature 733
19.5 Mapping QTL in Populations with Known APPENDIX B: Bioinformatic Resources for Genetics and
Pedigrees 663 Genomics 734
The basic method for QTL mapping 664
GLOSSARY 736
From QTL to gene 667
ANSWERS TO SELECTED PROBLEMS 757
19.6 Association Mapping in Random-Mating
Populations 669 INDEX 770
The basic method for GWAS 669
GWA, genes, disease, and heritability 671
Chapter Objectives —New to the twelfth edition, each chapter begins with a two-
sentence paragraph describing the goals of the chapter and placing the chapter topic into
context of the surrounding chapters. This helps students “see the forest” before stepping
into the “trees.”
xii
Problem-Solving Skills
31. Write the sequence of the telomerase RNA that
serves as a template for the telomere repeat sequence
45. Would the Meselson–Stahl experiment have worked if
diploid eukaryotic cells had been used instead?
for Success
5′-TTAGGG-3′. 46. Consider the following segment of DNA, which is part
32. Why might Werner syndrome increase the chances of of a much longer molecule constituting a chromosome:
getting cancer?
5′ . . . ATTCGTACGATCGACTGACTGACAGTC . . . 3′
33. Draw 2′,3′ dideoxyadenosine and predict what would
Introduction
happen to Genetic
if this nucleotide Analysisinto
was incorporated hasthe
always 3′
been known for its rigorous
. . . TAAGCATGCTAGCTGACTGACTGTCAG . . . 5′
and powerful
growing DNA chainproblem sets. The twelfth edition expands
during replication. on this tradition both
If the DNA polymerase starts replicating this segment
in the how
34. Explain textDNA
and online in three
fulfills the SaplingPlus.
main requirements from the right,
for a hereditary molecule: (1) the ability to store infor- a. which will be the template for the leading strand?
mation, (2) the ability to be replicated, and (3) the
ability to mutate. b. draw the molecule when the DNA polymerase is
halfway along this segment.
35. Match the protein with its Working
function. with the Figures
c. draw the two complete daughter molecules.
A. DNA polymerase creates RNA primers
47. The DNA polymerases are positioned over the fol-
B. Helicase links short DNA chains
lowing DNA segment (which is part of a much larger
C. Ligase helps hold polymerase on DNA molecule) and moving from rightWorking with
to left. If we the
assume
D. Primase separates DNA strands Figures questions
that an Okazaki fragment is made from this segment,
prompt
what will be the fragment’s sequence? Labelstudents to
its 5′ and
E. Gyrase prevents reannealing of DNA
3′ ends. examine book fig-
F. Sliding clamp extends DNA strand
ures and tease out
G. SSB removes supercoils in DNA 5′ . . . CCTTAAGACTAACTACTTACTGGGATC . . . 3′
important information,
36. Why is telomerase not required for replication of the 3′ . . . GGAATTCTGATTGATGAATGACCCTAG understand the. . . intri-
5′
bacterial genome? cacies of an experi-
37. Explain what is meant by the terms conservative and CHALLENGING PROBLEMS mental protocol, relate
semiconservative replication. concepts, or draw
48. If you extract the DNA of the coliphage φX174, you
38. Describe two pieces of evidence indicating that DNA a conclusion about
will find that its composition is 25 percent A, 33 per-
polymerase I is not the chromosomal replicase. what is shown.
cent T, 24 percent G, and 18 percent C. Does this
39. What is meant by a primer, and why are primers nec- composition make sense in regard to Chargaff’s rules?
essary for DNA replication? How would you interpret this result? How might such
a phage replicate its DNA?
40. A molecule of composition
49. Given what you know about the structure and func-
5′-AAAAAAAAAAAAA-3′
tion of telomerase, provide a plausible model to
3′-TTTTTTTTTTTTT-5′ explain how a species could exist with a combination
is replicated in a solution containing unlabeled (not of two different repeats (for example, TTAGGG and
radioactive) dGTP, dCTP, and dTTP plus dATP with all TTGTGG) on each of their telomeres.
its phosphorus atoms in the form of the
NEW Genetics
radioactive iso-
tope 32P. Will both daughter molecules be radioactive?
50.and Society
Why is Questions
it unlikely that continuous replication of both
DNA strands occurs but is yet to be discovered?
Explain. Then repeat the question for the molecule
5′-ATATATATATATAT-3′
GENETICS AND SOCIETY
3′-TATATATATATATA-5′
NEW Genetics and
In this chapter, you learned that the shortening of chro-
SocietyonQuestions
41. Why is DNA synthesis continuous one strand and mosome telomeres due to diminished telomerase activity
at strand?
discontinuous on the opposite the end of each is associated with aging. This raises the possibility that
chapter
42. Explain why cutting one strand ask stu- DNA
of supercoiled gene therapy aimed at overexpression of telomerase will
removes the supercoiling. dents to consider the increase longevity. Do you think that it is ethical to use this
societal relevance approach to increase the longevity of normal, healthy peo-
43. Describe how the enzymatic activities of DNA poly-
merases I and III are similar of
anda topic in the
different. ple? Does your answer change if you consider that there
chapter; good for are nongenetic means such as calorie restriction that may
44. If the GC content of a DNA is 48 percent, what are the increase longevity, or that gene therapy is being pursued to
initiating classroom
percentages of the four bases (A, T, G, and C) in this treat numerous diseases?
discussion.
molecule? www
www
xiv
xv
xvi
SaplingPlus also
includes powerful
multimedia assets
such as videos and
animations to help
students visualize
complex processes.
xvii
done with it, or why it protein has been inserted into the chromosomes of the glowing mice. The other
mice are normal. [Eye of Science/Science Source.]
The mouse genome can be modified by the insertion
of specific fragments of DNA into a fertilized egg or into
serves as a good model somatic cells. The mice in the photograph have received a
for the studies described T he laboratory mouse is descended from the house
mouse Mus musculus. The pure lines used today as
jellyfish gene for green fluorescent protein (GFP) that makes
them glow green under special lights. Gene knockouts and
in the main text. standards are derived from mice bred in past centuries by replacements also are possible.
mouse “fanciers.” Among model organisms, it is the one A major limitation of mouse genetics is its cost. Whereas
whose genome most closely resembles the human genome. working with a million individuals of E. coli or S. cerevisiae is
Its diploid chromosome number is 40 (compared with 46 a trivial matter, working with a million mice requires a factory-
in humans), and the genome is slightly smaller than that of size building. Furthermore, although mice do breed rapidly
humans (the human genome being 3000 Mb) and contains compared with humans, they cannot compete with microor-
approximately the same number of genes (current esti- ganisms for speedy life cycle. Hence, the large-scale selec-
mate 25,000). Furthermore, all mouse genes seem to have tions and screens necessary to detect rare genetic events
counterparts in humans. A large proportion of genes are are not possible.
xviii
rather dark overall pigmentation. A mutation called yellow The expected monohybrid ratio of 1: 2:1 would be
(a lighter coat color) shows a curious inheritance pattern. If found among the zygotes, but it is altered to a 2:1 ratio in
any yellow mouse is mated with a homozygous wild-type the progeny actually seen at birth because zygotes with a
01_GriffitITGA12e_11478_FM_i_xx.indd 18 mouse, a 1:1 ratio of yellow to wild-type mice is always lethal AY /AY genotype do not survive to be counted. This 23/10/19 9:42 AM
Acknowledgments
We extend our thanks and gratitude to our colleagues who reviewed this edition and
whose insights and advice were most helpful:
Rashid Abu-Ghazalah, McMaster Teresa Donze-Reiner, West Chester Whitney M. Jones, North Carolina State
University University University
Uduak Afangideh, Faulkner University David Durica, University of Oklahoma Katie Vermillion Kalmon, University of
Faiz Ahmad, Brandon University Deborah Eastman, Connecticut College Wisconsin, Madison
Shivanthi Anandan, Drexel University Edward Eivers, California State University, Kathleen Karrer, Marquette University
Katsura Asano, Kansas State University Los Angeles Christin Mercedes Kastl, University of
Rao Ayyagari, Lindenwood University Nancy L. Elwess, State University of New Maine at Fort Kent
Guy F. Barbato, Stockton University York, Plattsburgh Kathrin Schrick, Kansas State University
Isabelle H. Barrette-Ng, University of Bert Ely, University of South Carolina Oliver Kerscher, The College of William &
Calgary Yiwen Fang, Loyola Marymount University Mary
Bruce Bejcek, Western Michigan University Robert E. Farrell Jr., Penn State University Nobuaki Kikyo, University of Minnesota
John Belote, Syracuse University Steven D. Fenster, Fort Lewis College Miriam K. Konkel, Clemson University
Renaud Berlemont, California State Victor Fet, Marshall University Lori Koziol, New England College
University, Long Beach Christy Fillman, University of Colorado Brian Kreiser, University of Southern
Jaime E. Blair, Franklin & Marshall College Wayne Forrester, Indiana University Mississippi
Nicole Bournias-Vardiabasis, California Richard D. Gardner, Southern Virginia Jason N. Kuehner, Emmanuel College
State University, San Bernardino University Dana Robert Kurpius, Elgin Community
Mirjana Milosevic Brockett, Georgia Topher Gee, University of North Carolina, College
Institute of Technology Charlotte Howard Laten, Loyola University Chicago
Judy Brusslan, California State University, Vaughn Gehle, Southwest Minnesota State Jason P. Lee, Lander University
Long Beach University John Loike, Touro College
Patrick Calie, Eastern Kentucky University Matthew Gilg, University of North Florida Mark S. Longo, University of Connecticut
Jeffrey D. Camper, Francis Marion Michael Gleason, Georgia College & State Xu Lu, University of Findlay
University University Bethany Lucas, Regis University
John C. Carmen, Northern Kentucky Paul Goldstein, University of Texas at El Michael Martin, John Carroll University
University Paso Endre Mathe, University of Debrecen,
Steven M. Carr, Memorial University of Eli Greenbaum, University of Texas at El Hungary, Vasile Goldis University of
Newfoundland Paso Arad, Romania
J. Aaron Cassill, University of Texas at San Joanna Gress, Abraham Baldwin P. J. Maughan, Brigham Young University
Antonio Agricultural College Herman Mays, University of Cincinnati,
Maria V. Cattell, University of Colorado Chris Griffin, Ohio University Andrew McCubbin, Washington State
Sarah Certel, University of Montana Patrick J. Gulick, Concordia University University
Henry C. Chang, Purdue University Jody Hall, Brown University Virginia McDonough, Hope College
Hexin Chen, University of South Carolina Mike Harrington, University of Alberta Thomas Mennella, Bay Path University
Ian Chin-Sang, Queens University J. Scott Harrison, Georgia Southern Julie E. Minbiole, Columbia College
Youngkoo Cho, Eastern New Mexico University Chicago
University Elizabeth Hart, University of Ekaterina Mirkin, Tufts University
Sara G. Cline, Athens State University Massachusetts, Dartmouth Anni Moore, Morningside College
Craig E. Coleman, Brigham Young George Haughn, University of British Sarah Mordan-McCombs, Franklin College
University Columbia of Indiana
Diane M. Cook, Louisburg College Christopher J. Hickey, Wilkes Honors Jeanelle M. Morgan, University of North
Sarah Adelaide Crawford, Southern College at Florida Atlantic University Georgia
Connecticut State University Gregory Hocutt, Mesa Community College Gary Z. Morris, Glenville State College
Hongchang Cui, Florida State University Liza Holeski, Northern Arizona University Christopher O’Connor, Maryville
Cristina M. Cummings, Stockton Margaret Hollingsworth, University at University
University Buffalo Daniel Odom, California State University,
Ann Marie Davison, Kwantlen Polytechnic Adam W. Hrincevich, Louisiana State Northridge
University University Greg Odorizzi, University of Colorado at
Elizabeth A. De Stasio, Lawrence Jeffrey A. Hughes, Millikin University Boulder
University Diana S. Ivankovic, Anderson University Maria E. Orive, University of Kansas
Matt Dean, University of Southern Varuni Jamburuthugoda, Fordham Pamela Osenkowski, Loyola University
California University Chicago
Tracie Delgado, Northwest University Zhenyu Jia, University of California, Ana Otero, Emmanuel College
Alyce DeMarais, University of Puget Riverside Paul Overvoorde, Macalester College
Sound Lan Jiang, Oakland University Leocadia Paliulis, Bucknell University
xix
Holly Paquette, College of Western Idaho Melanie A. Sacco, California State Abe Tucker, Southern Arkansas University
Sally G. Pasion, San Francisco State University, Fullerton Jennifer C. Tudor, Saint Joseph’s University
University Jon Schnorr, Pacific University L. K. Tuominen, John Carroll University
Thomas R. Peavy, California State Aaron Schrey, Georgia Southern University Ludmila Tyler, University of Massachusetts,
University, Sacramento Dana Schroeder, University of Manitoba Amherst
Guy M. L. Perry, University of Prince Sandra Schulze, Western Washington Philip Villani, Butler University
Edward Island University Darlene Walro, Walsh University
Lynn A. Petrullo, College of New Rochelle Bin Shuai, Wichita State University Yunqiu Wang, University of Miami
Susanne Pfeifer, Arizona State University, Elaine Sia, University of Rochester Randal Westrick, Oakland University
Tempe Amanda Simons, Framingham State Matt White, Ohio University
Ruth Phillips, Syracuse University University Daniel Williams, Coastal Carolina
Helen Piontkivska, Kent State University Elspeth Smith, University of Guelph University
Andres Posso-Terranova, University of Marc Spingola, University of Missouri, Darla J. Wise, Concord University
Saskatchewan St. Louis Donald Withers, Husson University
Heather Prior, The King’s University Amy E. Sprowles, Humboldt State Glenn Yasuda, Seattle University
Jeffrey L. Reinking, State University of University Mary Alice Yund, University of California,
New York, New Paltz Emily Stowe, Bucknell University Berkeley Extension
Keefe Riedel Reuther, University of Alice Tarun, Alfred State College Xing-Hai Zhang, Florida Atlantic
California, San Diego Michael A. Thomas, Idaho State University, Boca Raton
Eugenia Ribeiro-Hurley, Fordham University Jianmin Zhong, Humboldt State
University Judith M. Thorn, Knox College University
Todd Rimkus, Marymount University Douglas Thrower, University of California, David S. Zuzga, La Salle University
Edmund Rucker, University of Kentucky Santa Barbara
Tony Griffiths would like to acknowledge the pedagogical insights of David Suzuki, who was
a co-author of the early editions of this book, and whose teaching in the media is now an
inspiration to the general public around the world. Great credit is also due to Jolie Mayer-
Smith and Barbara Moon, who introduced Tony to the power of the constructivist approach
applied to teaching genetics.
John Doebley would like to thank his University of Wisconsin colleagues Bill Engels, Carter
Denniston, and Jim Crow, who shaped his approach to teaching genetics, as well as Jim
Birchler, Allen Laughon, and Anna-Lisa Doebley for helpful comments of select chapters.
Katie Peichel would like to thank Jasper Rine for inspiring her love of genetics as an
undergraduate, Tom Vogt for teaching her how to actually be a geneticist, and David Kingsley
for mentoring throughout her career as a geneticist.
David Wassarman is particularly grateful for the teaching influences of Joe Pelliccia, Tom
Wenzel, Joan Steitz, Karen Wassarman, Doug Wassarman, and Kelly Wassarman.
The authors also thank the team at W. H. Freeman for their hard work and patience. In
particular we thank our developmental editors, Erica Champion, Erica Frost, and Michael
Zierler; program director Sandy Lindelof; senior content project manager Harold Chester; and
copy editor Matthew Van Atta. We also thank Paul Rohloff, senior workflow project manager;
Natasha Wolfe, design services manager; Matthew McAdams, art manager; Robin Fadool,
executive permissions editor; Richard Fox, permissions project manager; Cassandra Korsvik,
senior media editor; Jennifer Compton, media editor; and Casey Blanchard, editorial assistant.
Finally, we especially appreciate the marketing and sales efforts of Will Moore, executive
marketing manager, and the entire sales force.
T
he science of genetics was born about 120 years ago. Like begets like
Since that time, genetics has profoundly changed our
understanding of life, from the level of the individ-
ual cell to that of a population of organisms evolving over
millions of years. In 1900, William Bateson, a prominent
British biologist, wrote presciently that an “exact deter-
mination of the laws of heredity will probably work more
change in man’s outlook on the world, and in his power
over nature, than any other advance in natural knowledge
that can be foreseen.” Throughout this text, you will see
the realization of Bateson’s prediction. Genetics has driven
a revolution in both the biological sciences and society in
general.
In this first chapter, we will look back briefly at the his-
tory of genetics, and in doing so, we will review some of
the basic concepts of genetics that were discovered over the
last century. After that, we will look at a few examples of
how genetic analysis is being applied to critical problems in
biology, agriculture, and human health today. You will see
how contemporary research in genetics integrates concepts
discovered decades ago with recent technological advances.
You will see that genetics today is a dynamic field of inves-
tigation in which new discoveries continually advance our
understanding of the biological world.
FIGURE 1-1 Family groups in the gray wolf show familial
resemblances for coat colors and patterning. [(Top) DLILLC/Corbis/
VCG/Getty Images; (bottom) Bev McConnell/Getty Images.]
1.1 THE BIRTH OF GENETICS
LO 1.1 Know the experiments by which genetics developed
from Mendel to today. the Hopi farmers hoped to harvest. Upon receiving this
LO 1.2 Know the molecules involved in storage
message, the gods would faithfully return them a plant that
and expression of genetic information. produced kernels of the desired color.
In the 1800s in Europe, horticulturalists, animal breed-
Throughout recorded history, people around the world ers, and biologists also sought to explain the resemblance
have understood that “like begets like.” Children resemble between parents and offspring. A commonly held view at
their parents, the seed from a tree bearing flavorful fruit that time was the blending theory of inheritance, or the
will in turn grow into a tree laden with flavorful fruit, and belief that inheritance worked like the mixing of fluids
even members of wolf packs show familial resemblances such as paints. Red and white paints, when mixed, give
(Figure 1-1). Although people were confident in these obser- pink; and so a child of one tall parent and one short par-
vations, they were left to wonder as to the underlying ent could be expected to grow to a middling height. While
mechanism. The Native American Hopi tribe of the south- blending theory works at times, it is also clear that there
western United States understood that if they planted a red are exceptions, such as tall children born to parents of aver-
kernel of maize in their fields, it would grow into a plant age height. Blending theory also provides no mechanism by
that also gave red kernels. The same was true for blue, which the imagined “heredity fluids,” once mixed, could be
white, or yellow kernels. So they thought of the kernel as separated—the red and white paints cannot be reconstituted
a message to the gods in the Earth about the type of maize from the pink. Thus, the long-term expectation of blending
2
theory over many generations of intermating among indi- One of Mendel’s experiments
viduals is that all members of the population will come to
express the same average value of a trait. Clearly, this is
not how nature works. There are people with a range of
heights, from short to tall, and we have not all narrowed in
Parents ×
on a single average height despite the many generations that
humans have dwelled on Earth.
Mendel’s 1866 publication it, “There are people who William Bateson gave
seem to be born in a van- genetics its name
ishing cap. Mendel was one
of them.”
Mendel
rediscovered
As the legend goes, when
the British biologist William
Bateson (Figure 1-5) boarded
a train bound for a confer-
ence in London in 1900, he
had no idea how profoundly
his world would change
during the brief journey. FIGURE 1-5 William Bateson,
Bateson carried with him the British zoologist and
a copy of Mendel’s 1866 evolutionist who introduced
paper on the hybridization the term genetics for the study
of inheritance and promoted
of plant varieties. Bateson Mendel’s work. [SPL/Science
had recently learned that Source.]
biologists in Germany, the
Netherlands, and Austria had each independently reproduced
Mendel’s 3:1 ratio, and they each cited Mendel’s original
work. Bateson needed to read Mendel’s paper. By the time
FIGURE 1-4 Excerpts from Mendel’s 1866 publication, Versuche
über Pflanzen-Hybriden (Experiments on Plant Hybrids). [Augustinian
he stepped off the train, Bateson had a new mission in life.
Abbey in Old Brno, Courtesy of the Masaryk University, Mendel Museum.] He understood that the mystery of inheritance had been
solved. He soon became a relentless apostle of Mendel’s laws
of inheritance. A few years later in 1905, Bateson coined the
dominant to the other, enabled him to explain the lack of
term genetics—the study of inheritance. The genetics revolu-
blending in the first-generation hybrids and the re-appearance
tion had begun.
of white-flowered plants in the second-generation hybrids
When Mendel’s laws of inheritance were rediscovered
with a 3:1 ratio of purple- to white-flowered plants. This rev-
in 1900, a flood of new thinking was unleashed. Mendel-
olutionary advance in our understanding of inheritance will
ism became the organizing principle for much of biology.
be fully discussed in Chapter 2.
There were many new questions to be asked about inheri-
KEY CONCEPT Mendel concluded that (1) genes behave tance. Table 1-1 summarizes the chronology of seminal dis-
like particles and do not blend together, and (2) one allele is coveries made over the coming decades and the chapters of
dominant to the other. this text that cover each of these topics. Let’s look briefly at
a few of the questions and their answers that transformed
How did Mendel get it right when so many others the biological sciences.
before him were wrong? Mendel chose a good organism Where in the cell are Mendel’s genes? The answer came in
and good traits to study. The traits he studied were all con- 1910, when Thomas H. Morgan at Columbia University in
trolled by single
Introduction genes.
to Genetic Traits 11e
Analysis, that are controlled by several New York demonstrated that Mendel’s genes are located on
genes,
Figure as many
01.04 #104traits are, would not have allowed him to chromosomes—he proved the chromosome theory of inher-
04/15/14 the laws of inheritance so easily. Mendel was also
discover itance. The idea was not new. Walter Sutton, who was raised
05/01/14
aDragonfly
careful observer, and he kept detailed records of each of
Media Group
on a farm in Kansas and later served as a surgeon for the
his experiments. Finally, Mendel was a creative thinker U.S. army during WWI had proposed the chromosome the-
capable of reasoning well beyond the ideas of his times. ory of inheritance in 1903. Theodor Boveri, a German biol-
Mendel’s particulate theory of inheritance was pub- ogist, independently proposed it at the same time. It was a
lished in 1866 in the Proceedings of the Natural History compelling hypothesis, but there were no experimental data
Society of Brünn (see Figure 1-4). At that time, his work to support it. This changed in 1910, when Morgan proved
was read by some other biologists, but its implications the chromosome theory of inheritance using the fruit fly as
and importance went unappreciated for almost 40 years. his experimental organism. In Chapter 4, you will retrace
Unlike Charles Darwin, whose theory of evolution by natu- Morgan’s experiments that proved genes are on chromosomes.
ral selection made him world-renowned virtually overnight, Can Mendelian genes explain the inheritance of con-
when Mendel died in 1884, he was more or less unknown tinuously variable traits such as human height? While
in the world of science. As biochemist Erwin Chargaff put 3:1 segregation ratios could be directly observed for simple
1865 Gregor Mendel showed that traits are controlled by discrete factors now known as genes. 2, 3
1903 Walter Sutton and Theodor Boveri hypothesized that chromosomes are the hereditary elements. 4
1905 William Bateson introduced the term genetics for the study of inheritance. 2
1908 G. H. Hardy and Wilhelm Weinberg proposed the Hardy–Weinberg law, the foundation for population 18
genetics.
1913 Alfred Sturtevant made a genetic linkage map of the Drosophila X chromosome, the first genetic map. 4
1918 Ronald Fisher proposed that multiple Mendelian factors can explain continuous variation for traits, 19
founding the field of quantitative genetics.
1931 Harriet Creighton and Barbara McClintock showed that crossing over is the cause of recombination. 4, 15
1941 Edward Tatum and George Beadle proposed the one-gene–one-polypeptide hypothesis. 5
1944 Oswald Avery, Colin MacLeod, and Maclyn McCarty provided compelling evidence that DNA is the 7
genetic material in bacterial cells.
1948 Barbara McClintock discovered mobile elements (transposons) that move from one place to another 16
in the genome.
1950 Erwin Chargaff showed DNA composition follows some simple rules for the relative amounts of A, C, 7
G, and T.
1952 Alfred Hershey and Martha Chase proved that DNA is the molecule that encodes genetic information. 7
1953 James Watson and Francis Crick, using data produced by Rosalind Franklin and Maurice Wilkins, 7
determined that DNA forms a double helix.
1958 Matthew Meselson and Franklin Stahl demonstrated the semiconservative nature of DNA replication. 7
1958 Jérôme Lejeune discovered that Down syndrome resulted from an extra copy of the 21st 17
chromosome.
1961 François Jacob and Jacques Monod proposed that enzyme levels in cells are controlled by feedback 11
mechanisms.
1961–1967 Marshall Nirenberg, Har Gobind Khorana, Sydney Brenner, and Francis Crick “cracked” the genetic 9
code.
1968 Motoo Kimura proposed the neutral theory of molecular evolution. 18, 20
1977 Fred Sanger, Walter Gilbert, and Allan Maxam invented methods for determining the nucleotide 10
sequences of DNA molecules.
1980 Christiane Nüsslein-Volhard and Eric F. Wieschaus defined the complex of genes that regulate body 13
plan development in Drosophila.
1989 Francis Collins and Lap-Chee Tsui discovered the gene causing cystic fibrosis. 4, 10
1998 Andrew Fire and Craig Mello discover a mechanism of gene silencing by double-stranded RNA. 8, 13
2009 Elizabeth H. Blackburn, Carol W. Greider, and Jack W. Szostak win the Nobel prize for their discovery 7
of how chromosomes are protected by telomeres and the enzyme telomerase.
2012 John Gurdon and Shinya Yamanaka win the Nobel Prize for their discovery that just four regulatory 8, 12
genes can convert adult cells into stem cells.
4:10 4:11 5:0 5:1 5:2 5:3 5:4 5:5 5:6 5:7 5:8 5:9 5:10 5:11 6:0 6:1 6:2
FIGURE 1-6 Students at the Connecticut Agriculture College in 1914 show a range of heights.
Ronald Fisher proposed that continuously variable traits such as human height are controlled by
multiple Mendelian genes.
traits such as flower color, many traits show a continuous Tatum and Beadle’s breakthrough became known as the
range of values in second-generation hybrids without sim- one-gene–one-enzyme hypothesis. You will see how they
ple ratios such as 3:1. In 1918, Ronald Fisher, the British developed this hypothesis in Chapter 5.
statistician and geneticist, resolved how Mendelian genes What is the physical nature of the gene? Are genes com-
explained the inheritance of continuously variable traits posed of protein, nucleic acid, or some other substance?
such as height in people (Figure 1-6). Fisher’s core idea was In 1944, Oswald Avery, Colin MacLeod, and Maclyn Mc
that continuous traits are each controlled by multiple Men- Carty offered the first compelling experimental evidence
delian genes. Fisher’s insight is known as the multifactorial that genes are made of deoxyribonucleic acid (DNA). They
hypothesis. In Chapter 19, we will dissect the experimental showed that DNA extracted from a virulent strain of bac-
evidence for Fisher’s hypothesis. teria carried the necessary genetic information to transform
a nonvirulent strain into a virulent one. Their inference was
KEY CONCEPT The multifactorial hypothesis states that confirmed in 1952 by Alfred Hersey and Martha Chase.
continuously variable traits are each controlled by multiple You will learn exactly how they demonstrated this in
Mendelian genes. Chapter 7.
How can DNA molecules store information? In the
1950s, there was something of a race among several
How do genes function inside cells in a way that groups of scientists to answer this question. In 1953, James
enables them to control different states for a trait such as Watson and Francis Crick, working at Cambridge Univer-
flower color? In 1941, Edward Tatum and George Beadle sity in England, won that race. They determined that the
proposed that genes encode enzymes. Using bread mold molecular structure of DNA was in the form of a double
(Neurospora crassa) as their experimental organism, they helix—two strands of DNA wound side-by-side in a spi-
demonstrated that genes encode the enzymes that perform ral. Their structure of the double helix is like a twisted lad-
metabolic functions within cells (Figure 1-7). In the case der (Figure 1-8). The sides of the ladder are made of sugar
of the pea plant, there is a gene that encodes an enzyme and phosphate groups. The rungs of the ladder are made of
required to make the purple pigment in the cells of a flower. four bases: adenine (A), thymine (T), guanine (G), and
(a) (b) 5´
O O
P 3´
O O
N H O H
CH2
O A N H N T O
O
O O O
P CH2
O O N H O O
O
CH2 P
C N O O
O H NG
O H N
O
O O CH2
P
O O O O O
H N
CH2 P
T N H O O
O NA
O
O
O O
P CH2
O O
O H N O O
CH2 P
O G N H N C O O
N H O
O
O
CH2
H
3´ O O
P
O O
5´
FIGURE 1-8 (a) The double-helical structure of DNA, showing the sugar–phosphate backbone in blue
and paired bases in brown. (b) A flattened representation of DNA showing how A always pairs with T,
and G always pairs with C. Each row of dots between the bases represents a hydrogen bond.
cytosine (C). The bases face the center, and each base is demonstrated that genes have regulatory elements that
hydrogen bonded to the base facing it in the opposite strand. control gene expression—that is, whether a gene is turned
Adenine in one strand is always paired with thymine in the on or off (Figure 1-9). The regulatory elements are specific
other by a double hydrogen bond, whereas guanine is always DNA sequences to which a regulatory protein binds and
paired with cytosine by a triple hydrogen bond. The bonding acts as either an activator or repressor of the expression of
specificity is based on the complementary shapes and charges the gene. In Chapter 11, you will explore the logic behind
of the bases. The sequence of A, T, G, and C represents the the experiments of Jacob and Monod with E. coli, and in
coded information carried by the DNA molecule. You will Chapter 12, you will explore the details of gene regulation
learn in Chapter 7 how this was all worked out. in eukaryotes.
How is the information stored in DNA decoded to syn-
KEY CONCEPT DNA is a double helix in which the nucle- thesize proteins? While the discovery of the double-helical
otide bases of one strand are paired with those of the other structure of DNA was a watershed for biology, many details
strand. Adenine always pairs with thymine, and guanine always
were still unknown. Precisely how information was encoded
pairs with cytosine.
into DNA and how it was decoded to form the enzymes
that Tatum and Beadle had shown to be the workhorses of
How are genes regulated? Cells need mechanisms to gene action remained unknown. From 1961 through 1967,
turn genes on or off in specific cell and tissue types and at teams of geneticists and chemists working in several coun-
specific times during development. In 1961, François Jacob tries answered these questions when they “cracked the genetic
and Jacques Monod made a conceptual breakthrough on code.” What this means is that they deduced how a string
this question. Working on the genes necessary to metabo- of DNA nucleotides, each with one of four different bases
lize the sugar lactose in the bacterium Escherichia coli, they (A, T, C, or G), encodes the set of 20 different amino acids
Genes have regulatory and coding regions He curiously used the word dogma, “a belief that is to be
accepted without doubt,” when he intended hypothesis, “a
RNA polymerase testable explanation for an observed phenomenon.” Despite
Regulatory complex
this awkward beginning, the phrase had an undeniable
protein
Direction of power and it has survived.
transcription Figure 1-10b captures much of what was learned about
the biochemistry of inheritance from 1905 until 1967. Let’s
GGGCCC
review the wealth of knowledge that this simple figure cap-
Regulatory Site where the Protein coding
element RNA polymerase sequence tures. At the left, you see DNA and a circular arrow repre-
complex binds senting DNA replication , the process by which a copy of
the DNA is produced. This process enables each of the two
FIGURE 1-9 A protein-coding gene includes a regulatory DNA
daughter cells that result from cell division to have a complete
element (GGGCCC) to which a regulatory protein binds, the site
where a group of proteins called the RNA polymerase complex binds copy of all the DNA in the parent cell. In Chapter 7, you will
to initiate transcription, and a protein-coding sequence. explore the details of the structure of DNA and its replication.
Another arrow connects DNA to RNA, symbolizing
that are the building blocks of proteins. They also discovered how the sequence of base pairs in a gene (DNA) is copied
that there is a messenger molecule made of ribonucleic acid to an RNA molecule. The process of RNA synthesis from
(RNA) that carries information in the DNA in the nucleus to the a DNA template is called transcription. One class of RNA
cytoplasm where proteins are synthesized. By 1967, the basic molecules made by transcription is messenger RNA , or
flowchart for information transmission in cells was known. mRNA for short. mRNA is the template for protein syn-
This flowchart is called the central dogma of molecular biology. thesis. In Chapter 8, you will discover how transcription is
accomplished.
KEY CONCEPT Genes reside on chromosomes and are The final arrow in Figure 1-10b connects mRNA and
made of DNA. Genes encode proteins that conduct the basic protein. This arrow symbolizes protein synthesis, or the
enzymatic work within cells. translation of the information in the specific sequence of
bases in the mRNA into the sequence of amino acids that
compose a protein. Proteins are the workhorses of cells,
The central dogma of molecular biology comprising enzymes, structural components of the cell,
In 1958, Francis Crick introduced the phrase “central and molecules for cell signaling. The process of translation
dogma” to represent the flow of genetic information within takes place at the ribosomes in the cytoplasm of each cell.
cells from DNA to RNA to protein, and he drew a sim- In Chapter 9, you will learn how the genetic code is writ-
ple diagram to summarize these relationships (Figure 1-10a). ten in three-letter words called codons. A codon is a set of
(a)
Replication
Transcription Translation
DNA RNA Protein
(b)
Protein
FIGURE 1-10 (a) One version of Francis Crick’s sketch of the central dogma, showing informa-
tion flow between biological molecules. The circular arrow represents DNA replication, the central
straight arrow represents the transcription of DNA into RNA, and the right arrow the translation of
RNA into protein. (b) More detailed sketch showing how the two strands of the DNA double helix A N I M ATED A RT
are independently replicated, how the two strands are disassociated for transcription, and how the The central dogma
messenger RNA (mRNA) is translated into protein at the ribosome.
three consecutive nucleotides in the mRNA that specifies What features make a species suitable as a model organ-
an amino acid in a protein. For example, CGC specifies the ism? (1) Small organisms that are easy and inexpensive to
amino acid arginine, AGC specifies serine, and so forth. maintain are very convenient for research. So fruit flies
Since Crick proposed the central dogma, additional are good, blue whales not so good. (2) A short generation
pathways of genetic information flow have been discov- time is imperative because geneticists, like Mendel, need
ered. We now know that there are classes of RNA that do to cross different strains and then study their first- and
not code for proteins, instances in which mRNA is edited second-generation hybrids. The shorter the generation time,
after transcription, and cases in which the information in the sooner the experiments can be completed. (3) A small
RNA is copied back to DNA (see Chapters 8, 9, and 16). genome is useful. As you will learn in Chapter 16, some
species have large genomes and others small genomes in
KEY CONCEPT Genes are made of DNA, which is terms of the total number of DNA base pairs. Much of the
transcribed to RNA molecules that serve as the template extra size of large genome species is composed of repetitive
for protein synthesis. DNA elements between the genes. If a geneticist is looking
for genes, these can be more easily found in organisms with
smaller genomes and fewer repetitive elements. (4) Organ-
1.2 AFTER CRACKING THE CODE isms that are easy to cross or mate and that produce large
numbers of offspring are best.
LO 1.3 Know the basic tools for genetic research including As you read this textbook, you will encounter certain
model organisms. organisms over and over. Organisms such as Escherichia
coli (a bacterium), Saccharomyces cerevisiae (baker’s
With the basic laws of inheritance largely worked out, the yeast), Caenorhabditis elegans (nematode or round-
1970s and beyond witnessed an era of applying genetic worm), Drosophila melanogaster (fruit fly), and Mus
analysis to many questions in biology. Much effort has musculus (mice) have been used repeatedly in experiments
been and continues to be invested in developing tools to and revealed much of what we know about how inheri-
address these questions. Geneticists focused their research tance works. Model organisms can be found on diverse
on a small number of species known as “model organisms” branches of the tree of life (see Figure 1-11), represent-
that are well suited for genetic analysis. Then in the late ing bacteria, fungi, algae, plants, and invertebrate and
1990s, the first complete genome sequences were published, vertebrate animals. This diversity enables each geneticist
launching the genomics era and the ability to study all the to use a model best suited to a particular question. Each
genes in the genome simultaneously. model organism has a community of scientists work-
ing on it who share information and resources, thereby
Model organisms facilitating each other’s research. More information on
each of the most commonly used model organisms can be
Geneticists make special use of a small set of model organ-
found in “A Brief Guide to Model Organisms” at the end
isms for genetic analysis. A model organism is a species used
of this book.
in experimental biology with the presumption that what is
Mendel’s experiments were possible because he had
learned from the analysis of that species will hold true for
several different varieties of pea plants, each of which
other species, especially other closely related species. The phi-
carried a different genetic variant for traits such as pur-
losophy underlying the use of model organisms in biology
ple versus white flowers, or tall versus dwarf stems. For
was wryly expressed by Jacques Monod: “Anything found to
each of the model species, geneticists have assembled
be true of E. coli must also be true of elephants.”1
large numbers of varieties (also called strains or stocks)
As genetics matured and focused on model organisms,
with special genetic characters that make them useful in
Mendel’s pea plants fell to the wayside, but Morgan’s fruit
research. For example, there are strains of fruit flies that
flies rose to prominence to become one of the most impor
have trait variants such as red versus white eyes. Similarly,
tant model organisms for genetic research. New species were
there are strains of mice that are prone to develop specific
added to the list. An inconspicuous little plant that grows as
forms of cancer or other diseases such as diabetes. Genetic
a weed called Arabidopsis thaliana became the model plant
strains enable geneticists to study how genes influence
species, and a minute roundworm called Caenorhabditis ele-
physiology, development, and disease. The different
gans that lives in compost heaps became a star of genetic
strains of each model organism are available to research-
analysis in developmental biology (Figure 1-11).
ers through stock centers that maintain and distribute the
KEY CONCEPT Genetic discoveries made in a model organ-
strains.
ism are often true of related species and may even apply to all
KEY CONCEPT Model organisms have features that make
forms of life.
them well-suited for genetic studies, such as small size, small
genome, large numbers of offspring, and short generation
time. Geneticists working with the same model organism share
1
F. Jacob and J. Monod, Cold Spring Harbor Quant. Symp. Biol. 26, stocks and information with one another.
1963, 393.
Fruit fly
Drosophila Nematode
melanogaster Caenorhabditis elegans
Yeast
Saccharomyces
cerevisiae
Mouse
Mus musculus
Mouse-eared cress
Arabidopsis
thaliana
Eukaryotes
Mycoplasma Archaea
gentalium
Bacillus
subtilis
Helicobacter
pylori E. coli
Eubacteria
FIGURE 1-11 The tree shows evolutionary relationships among the major groups of organisms:
Bacteria, Archaea, and Eukaryota (plants, fungi, and animals). [(Clockwise,
Clockwise, from top, center
center) Sinclair
Stammers/Science Source; SCIMAT/Science Source; Darwin Dale/Science Source; Biophoto Associates/
Science Source; imageBROKER/Superstock; blickwinkel/Alamy.]
Tools for genetic analysis or other enzymes, DNA can also be “labeled” or “tagged”
Geneticists and biochemists have created an incredible with a fluorescent dye or radioactive element so that the DNA
array of tools for characterizing and manipulating DNA, can be detected using a fluorescence or radiation detector.
RNA, and proteins. Many of these tools are described in Second, geneticists have developed methods to clone DNA
Chapter 10 or in other chapters relevant to a specific tool. molecules. Here, cloning refers to making many copies (clones)
There are a few themes to mention here. of a DNA molecule. The common way of doing this involves
First, geneticists have harnessed the cell’s own enzymatic isolating a relatively small DNA molecule (up to a few thou-
machinery for copying, pasting, cutting, and transcribing sand base pairs in length) from an organism of interest. The
DNA, enabling researchers to perform these reactions inside DNA molecule might be an entire gene or a portion of a gene.
test tubes. The enzymes that perform each of these func- The molecule is inserted into a host organism (often E. coli)
tions in living cells have been purified and are available to where it is replicated many times by the host’s DNA poly-
researchers: DNA polymerases can make a copy of a single merase. Having many copies of a gene is important for a vast
DNA strand by synthesizing a matching strand with the com- array of experiments used to characterize and manipulate it.
plementary sequence of A’s, C’s, G’s, and T’s. Nucleases can Third, geneticists have developed methods to insert foreign
cut DNA molecules in specific locations or degrade an entire DNA molecules into the genomes of many species, including
DNA molecule into single nucleotides. Ligases can join two those of all the model organisms (Figure 1-12). This process
DNA molecules together end-to-end. Using DNA polymerase is called transformation, and it is possible, for instance, to
Genetically modified transform genes from one cloned DNA of a gene can be tagged with a fluorescent dye
tobacco species into the genome and then hybridized to chromosomes fixed on a microscope
of another. The recipi- slide, revealing the chromosome on which the gene is located
ent species then becomes (Figure 1-13b).
a g e ne t ic a l ly mo d i f ie d Fifth, geneticists and biochemists have developed mul-
organism (GMO). In the last tiple methods for determining the exact sequence of all the
few years, geneticists have A’s, C’s, G’s, and T’s in a DNA molecule. These methods
developed an exciting new are collectively called DNA sequencing, and they have
method called CRISPR/Cas9 allowed geneticists to read the language of life. Recently,
that facilitates editing the cost-effective, high-through-put methods to sequence both
genes of an organism and very short (100 bp) and very long (10,000 bp) DNA mole-
is expected to r evolutionize cules were developed, enabling sequencing of the complete
not just laboratory genetics, genomes of thousands of individuals of a single species
but also medicine and agri- such as humans (see Chapter 14).
culture (see Chapter 10). Finally, over the last 20 years, researchers have created
Fourth, geneticists have molecular and computational tools for analyzing the entire
FIGURE 1-12 This genetically
modified tobacco plant has a
developed a large set of genome of an organism. These efforts gave birth to the field
gene from the firefly inserted methods based on hybridiz- of genomics—the study of the structure and function of entire
into its genome, giving it the ing DNA molecules to one genomes (see Chapter 14). Geneticists and genomicists have
capability to emit light. [Republished another (or to RNA mol- assembled mind-boggling amounts of information on model
with permission of the American ecules). The two comple- organisms and their genomes, including the complete DNA
Association for the Advancement
of Science, from D.W. Ow et al.,
mentary strands of DNA in sequence of their genomes, lists of all their genes, catalogs
“Transient and Stable Expression the double helix are bound of variants in these genes, data on the cell and tissue types in
of the Firefly Luciferase Gene in together by hydrogen bonds, which each gene is expressed, and much more. To get an idea
Plant Cells and Transgenic Plants” either G ≡ C or A = T. These of what is available, try browsing Fly Base (http://flybase.org/),
Science 234, 4778: (1986) bonds can be broken by heat the genomics Web site for the fruit fly (see also Appendix B).
pp. 856–859, Figure 5. Permission
conveyed through Copyright
(denatured) in an aqueous
solution to give two sin- KEY CONCEPT Geneticists developed tools to replicate,
Clearance Center, Inc.]
gle-stranded DNA molecules cut, label, and degrade DNA as well as use it as a template to
be transcribed into RNA. These tools allow the assembly of the
(Figure 1-13a). When the solution is cooled under controlled
DNA sequence of whole genomes. Computational tools allow
conditions, DNA molecules with complementary strands will
biological questions to be answered by the analysis of genome
preferentially hybridize with one another. DNA hybridization sequences and associated information.
methods have enabled many discoveries. For example, the
eementary
to complementary
sequencessequences Strands
Strands ofStrands
of nucleic
nucleic ofStrands
nucleic
acids
Strands
acids hybridize
of
acids
of
nucleic
hybridize nucleic
hybridize
to toacids
complementary
acids
to
hybridize
complementary
hybridize
complementary tosequences
to
complementary
complementary
sequences sequencessequences
sequences
Heat
Heat Heat Heat
Heat CoolCool Cool Cool
Cool
l DenatureDenature Denature
Denature Denature Anneal Anneal
Anneal Anneal
Anneal
3′ 5′ 3′ 5′ 5′3′ 3′ 5′ 3′ 5′ 5′
3′ 3′ 5′ 3′ 5′ 3′ 5′ 5′
3′ 3′
3′ 3′ 5′ 3′
5′ 5′ 3′ 3′ 5′5′5′3′
FIGURE 1-13 (a) The two strands of the DNA double helix can be dissociated by heat in aqueous
solutions. Upon cooling under controlled conditions, strands reassociate, or hybridize, with their
complement. (b) A cloned copy of the human gene for muscle glycogen phosphorylase was tagged
with a yellow fluorescent dye. The fluorescent-tagged DNA was then denatured and allowed to
hybridize to the chromosomes in a single cell. The fluorescent-tagged clone hybridized to the loca-
tion on chromosome 11 (yellow fluorescent regions) where the gene is located. [(b) Republished with
permission of the American Association for the Advancement of Science, from P Lichter, CJ Tang, K Call,
G Hermanson, GA Evans, D Housman, DC Ward, “High-resolution mapping of human chromosome 11 by
in situ hybridization with cosmid clones” Science 05 Jan 1990: Vol. 247, Issue 4938, pp. 64–69, Figure 1B.
Permission conveyed through Copyright Clearance Center, Inc.]
Au-ED, Au-ED,
Au-ED, Au-ED,
Au-ED,
Resolution
Resolution Resolution
is over
is over 300dpi
Resolution
is over
300dpi Resolution
forfor
300dpi
photo.
is over
photo. isfor
over
photo.
300dpi
300dpi
for for
photo.
photo.
Thanks, Thanks, Thanks,
Thanks, Thanks,
Craig Craig
Craig Craig
Craig
02_GriffitITGA12e_11478_Ch01_001_024.indd 11 03/10/19 9:48 AM
12 C H A P T E R 1 The Genetics Revolution
1.3 GENETICS TODAY should convey a dual message—the science of genetics has
profoundly changed our understanding of life, but it is
LO 1.4 Give examples of how genetics has influenced our also a youthful field in the midst of a dynamic phase of its
society. development.
FIGURE 1-14 (a) Louise Benge developed an undiagnosed disease as a young woman. (b) An
X ray revealed that Louise Benge’s disease condition caused calcification of the arteries in her legs.
[(a) Jeannine Mjoseth, NHGRI/www.genome.gov; (b) National Human Genome Research Institute (NHGRI).]
What intrigued the UDP team about Benge’s case was copies would be defective. Each of Benge’s siblings would
that she was not alone in having this disease. Benge had two also need to have inherited two mutant copies from their
brothers and two sisters, and all of them had arterial calci- parents to explain the fact that they have ACDC. In Chap-
fication. Remarkably, however, Benge’s parents were unaf- ter 2, you will learn how to calculate the probability of this
fected. Moreover, Benge and her siblings all had children, and actually happening.
none of these children had arterial calcification. This pattern With this hint from the family history, the UDP team
of inheritance suggested that the underlying cause might be now knew where to look in the genome for the mutant
genetic. Specifically, it suggested that Benge and all of her sib- gene. They needed to look for a segment on one of the
lings inherited two defective copies of either CD73 or a gene chromosomes for which the copy that Benge inherited from
that influences CD73 expression—one from their mother her mother is identical to the copy she inherited from her
and one from their father. A person with one good copy and father. Moreover, each of Benge’s siblings must also have
one defective copy can be normal, but if both of a person’s two copies of this segment identical to Benge’s. Such regions
copies are defective, then they lack the function that the gene are very rare in people unless their parents are related, as in
provides. The situation is just like Mendel’s white-flowered the case of Benge since her parents are third cousins. Gener-
pea plants. Since the functional allele is dominant to the dys- ally, a segment of a chromosome that is just a few hundred
functional allele, ACDC, like white flowers, appears only if base pairs long will have several differences in the sequence
an individual carries two defective alleles. of A’s, C’s, G’s, and T’s between the copy we inherited from
The UDP team delved further into Benge’s family his- our mother and the one we inherited from our father. These
tory and learned that Benge’s parents were third cousins differences are known as single nucleotide polymorphisms,
(Figure 1-15). This revelation fit well with the idea that the or SNPs for short (see Box 1-1).
cause was a defective gene. When a husband and wife are The UDP team used a genomic technology, called a DNA
close relatives such as third cousins, there is an increased microarray (see Chapter 18), that allowed them to study one
chance that they will both have inherited the same version million base-pair positions across the genome. At each of
of a defective gene from their common ancestor and that these base-pair positions along the chromosomes, the team
they will both pass on this defective gene to their children. could see where Benge’s two chromosomal segments were
Children with one copy of a defective gene are often nor- identical, and whether all of Benge’s siblings also carried
mal, but a child who inherits a defective copy from both two identical copies in this segment. The UDP team found
parents is likely to have a genetic disorder. exactly the type of chromosome segment for which they were
In Figure 1-15, we can see how this works. Benge’s looking, and furthermore, they discovered that the gene that
mother and father (individuals V-1 and V-2 in the figure) encodes the CD73 enzyme is located in this segment. This
have the same great-great-grandparents (I-1 and I-2). If result suggested that Benge and her siblings all had two iden-
one of these great-great-grandparents had a mutant gene tical copies of the same defective CD73-encoding gene. The
for CD73, then it could have been passed down over the team seemed to have found the needle in a haystack; how-
generations to both Benge’s mother and father (follow the ever, there was one last experiment to perform.
red arrows). After that, if Benge received the mutant copy The team needed to identify the specific defect in the
from both her mother and her father, then both of her defective CD73 gene that Benge and her siblings had
I ? ?
1 2
II
inherited. After determining the DNA sequence for the of inheritance. The genealogies of some British families sug-
CD73 gene from Benge and her siblings, the team found gested that new mutations for the blood-clotting disorder
the defect in the gene—“the smoking gun.” The defective hemophilia tended to arise in men more frequently than in
gene encoded only a short, or truncated, protein—it did women. Taken together, these two observations suggested
not encode the complete sequence of amino acids. One of that the risk of an inherited disorder for a child is greater
the DNA codons with letters TCG that encodes the amino as the parents age and also that fathers are more likely than
acid serine was mutated to TAG, which signals the end of mothers to contribute new mutations to their children.
the protein. The protein made from Benge’s version of the Advances in genomics and DNA sequencing technology
CD73 gene was truncated, so it could not signal cells in the (see Chapter 14) allowed new analyses proving that Wein-
arteries to keep the calcification pathway turned off. berg’s and Haldane’s suspicions were correct and provid-
Louise Benge’s journey from first experiencing pain in ing a very detailed picture of the origin of new mutations
her legs to learning that she had a new disease called ACDC within families. Here is how it was accomplished. A team of
was a long one. The diagnosis of her disease was a triumph geneticists in Iceland studied 78 “trios”—a family group of
made possible by the integration of classic transmission a mother, a father, and their child.
genetics and genomics. Knowing the defect underlying the
disease ACDC allowed the doctors to try a medication that
Simple trio
they would never have considered before they knew that
the cause was a defective CD73 enzyme. The medication in
question is called etidronate, and it can substitute for CD73 The researchers determined the complete genome sequence
in signaling cells to keep the calcification pathway turned of each individual with DNA isolated from their blood cells,
off. Clinical trials with etidronate began in 2012 and are compiling genome sequences from a total of 219 individuals.
scheduled for completion in 2020. With these genome sequences in hand, the research-
ers could comb through the data for new or de novo
KEY CONCEPT The integration of classical genetics and mutations—unique DNA variants that exist in a child but in
genomic technologies allows the causes of inherited diseases neither of its parents. Their focus was on point mutations,
to be readily identified and appropriate therapies applied.
changes of one letter in the DNA code to another—for
example, a change of an adenosine (A) to a guanine (G)
(Figure 1-16). A point mutation creates a SNP.
The logic of the discovery process used by the Icelandic
Investigating mutation and disease risk geneticists is outlined in Figure 1-16, which shows a seg-
Shortly after the rediscovery of Mendel’s work, the German ment of DNA for each member of a trio. Each individual
physician Wilhelm Weinberg reported that there appeared has two copies of the segment. Notice that copy M1 in the
to be a higher incidence of short-limbed dwarfism (achon- mother has a SNP (green letter) that distinguishes it from
droplasia) among children born last in German families copy M2. Similarly, there are two SNPs (purple letters) that
than among those born first. A few decades later, British distinguish the father’s two copies of this segment. Compar-
geneticist J. B. S. Haldane observed another unusual pattern ing the child to the parents, we see that the child inherited
Mother Father
Copy M1 • • C AGCAGA T TGCTGC T T TGT A TGAG • • Copy F1 • • C AGC TGA T TGCTGC T T TGT AGGAG • •
Copy M2 • • CAGC TGA T TGCTGC T T TGT A TGAG • • Copy F2 • • CAA C TGA T TGCTGC T T TGT A TGAG • •
Child
FIGURE 1-16 A short segment of DNA from a specific location in the genome is depicted using
the nucleotide base letters of just one strand of the DNA duplex. Each individual has two copies of
the DNA segment. In the mother, these are labeled M1 and M2; in the father, F1 and F2. The child
inherited copy M1 from its mother and F2 from its father. The version of F2 in the child carries a new
point mutation (red, arrow). Single nucleotide polymorphisms (SNPs) that distinguish the different
copies are shown in green (mother) and purple (father).
copy M1 from its mother and copy F2 from its father. Look of mutation rises with the mother’s age when controlling
more closely at the child’s two copies of the segment, and for the age of the father, the team found no evidence that it
you will notice a unique variant (red letter) that occurs in did. Older mothers did not pass on more new point muta-
the child but in neither of its parents. This is a de novo tions to their offspring than younger ones. (Older moth-
point mutation. It this case, it is a mutation from a guanine ers are known to produce more chromosomal aberrations
(G) to a thymine (T). We can see that the mutation arose in than younger mothers, such as an extra copy of the 21st
the father because it is on the F2 copy of the segment. chromosome that causes Down syndrome; see Chapter 17.)
Where and exactly when did the new mutation depicted Next, they examined the relationship between mutation
in Figure 1-16 arise? Most of our bodies are composed of and the age of the father when controlling for the age of
somatic cells that make up everything from our brain to the mother. Here, they found a powerful relationship. Older
our blood. However, we also have a special lineage of cells fathers produce more new point mutations than young ones
called the germline that divide to produce eggs in women (Figure 1-17). In fact, for each year of increase in his age,
and sperm in men. New mutations that arise in somatic a father will pass on two additional new mutations to his
cells as theytodivide
Introduction Geneticduring the
Analysis, 11egrowth and development of children. A 20-year-old father will pass on about 25 new
our bodies
Figure are not passed on to our offspring. However, a
01.17 #131 mutations to each of his children, but a 40-year-old father
04/03/14
new mutation that occurs in the germline can be transmit- will pass on about 65 new mutations. Weinberg’s observa-
05/01/14
ted to the offspring. The mutation depicted in Figure 1-16 tion made 100 years earlier was confirmed.
Dragonfly Media Group
arose in the germline of the father. Why does the age of the father matter, while that of the
With the genome sequence data for the trios, the Ice- mother seems to have no effect on the frequency of new
landic geneticists made some startling discoveries. First, point mutations? The answer lies in the different ways by
among the 78 children in the study, they observed a total which men and women form gametes. In women, as in the
of 4933 new point mutations. Each child carried about females of other mammals, the process of making eggs takes
63 unique mutations that did not exist in its parents. Most place largely before a woman is born. Thus, when a woman
of these occurred in parts of the genome where they have is born, she possesses in her ovaries a set of egg precursor
only a small chance to pose a health risk, but 62 of the cells that will mature into egg cells without further rounds of
4933 mutations caused potentially damaging changes to DNA replication. For a woman, from the point when she was
the genes such that they altered the amino acid sequence conceived until the formation of the egg cells in her ovaries,
of the protein encoded. Second, among the mutations that there are about 23 rounds of cell division with DNA replica-
could be assigned a parent of origin, there were on average tion and an opportunity for a copying error or mutation. All
55 from the father for every 14 from the mother. The chil- 23 of these rounds of chromosome replication occur before
dren were inheriting nearly four times as many new muta- a woman is born, so there are no additional rounds after her
tions from their fathers as their mothers. The Icelandic team birth and no chance for additional mutations as she ages.
had confirmed Haldane’s prediction made 90 years earlier. Thus, older mothers contribute no more new point muta-
The genome sequences also allowed the team to test tions to their children than younger mothers.
Weinberg’s prediction that the frequency of mutation rises Sperm production is altogether different. The cell divi-
with the age of the parents. For each trio, the researchers sions that produce sperm continue throughout a man’s
knew the ages of the mother and the father at the time of life, and there are many more rounds of cell division in
conception. When they investigated whether the frequency sperm formation than in egg formation. Sperm produced
80
60
40
15 20 25 30 35 40 45
Age of father at conception of child (years)
by 20-year-old men will have replicated their DNA about When rice gets its feet a little too wet
150 times since the man’s conception, compared to only
Among the cereal crops, rice is unique. Whereas wheat,
23 DNA replications for the eggs produced by 20-year-old
barley, maize, and the other grain crops grow solely in
women. By the time a man is age 40, his sperm will have a
dry fields, rice is commonly grown in flooded fields called
history that involves over 25 times as many rounds of DNA
paddies (Figure 1-18). The ability of rice to grow in flooded
replication as for eggs in a woman of the same age. Thus,
fields offers it an advantage: rice can survive modest flood-
there is much greater risk of new point mutations occurring
ing (up to 25 cm of standing water) in the paddies, but most
during these extra rounds of cell division and DNA replica-
weeds cannot. So rice farmers can use flooding to control
tion with the increase in the age of the father.
the weeds in their field while their rice thrives.
There is one final twist to the remarkable project per-
The strategy works well where farmers have irrigation
formed by the Icelandic geneticists. The 78 trios that they stud-
systems to control the water levels in their paddies and
ied were chosen because the children in most of the trios had
heavy rains do not exceed their capacity to control these
inherited disorders. These included 44 children with autism
Introduction to Genetic Analysis, 11e levels. If the water in the paddies gets too deep (greater
spectrum
Figure 01.18disorder
#132 and 21 with schizophrenia. For all these
than 50 cm) for a prolonged period, then the rice plants,
children,
04/02/14 there were no other cases of these disorders among
like the weeds, can suffer or even die.
05/01/14
their relatives, suggesting that their condition was due to a new
Dragonfly Media Group Paddy agriculture, as practiced in the lowlands of India,
mutation. As anticipated, the researchers observed a correlation
Southeast Asia, and West Africa, relies on natural rainfall,
between the father’s age and disease risk—older fathers were
rather than irrigation, to flood the fields. This circumstance
more likely to have children with autism and schizophrenia.
poses a risk. When the rains are heavy, water depth in the
Studies such as this can have important implications
paddies can exceed 50 cm and completely submerge the
for individuals and society. Some men who intend to delay
plants, causing rice plants to either suffer a loss in yield or
parenting until later in life might choose to freeze samples
simply die. Of the 60 million hectares of rain-fed lowland
of their sperm while still young. This study also informs
paddies, one-third experience damaging floods that reduce
us that changes in society can impact the number of new
yield on a regular basis. Since this loss is incurred mostly
mutations that enter the human gene pool. If men choose to
by the poorest farmers, it can lead to malnourishment and
delay fatherhood for postsecondary education or establish-
even starvation.
ing their careers, there will be an associated increase in the
In the early 1990s, David Mackill, a plant geneticist and
number of new mutations among their children. It is com-
breeder at the International Rice Research Institute, had an
mon knowledge that infertility rises with age for women—
idea about how to improve rice so that it could tolerate
as is often stated, a woman’s “biological clock” is ticking
being submerged in flood waters. He identified a remark-
once she is past puberty. This work by the Icelandic geneti-
able variety of rice called FR13A that could survive sub-
cists informs us that a clock is ticking for men as well.
mergence and even thrive after the plants remained fully
submerged in deep water for up to two weeks. Unfortu-
KEY CONCEPT Mutation is a random process that occurs
nately, FR13A had a low yield and the quality of its grain
during DNA replication.
was marginal. So Mackill set out to transfer FR13A’s
Rice growing in a flooded field or paddy The next question was, how does switching on SUB1
enable FR13A to survive complete submergence? To answer
this question, let’s review how ordinary rice plants respond
to submergence. When a plant is completely submerged, oxy-
gen levels in its cells drop, and the concentration of ethylene,
a plant hormone, in the cells increases. Ethylene signals the
plant to escape submergence by elongating its leaves and
stems to keep its “head” above water. This escape strategy
works fine as long as the water is not too deep. If the flood
waters are too deep, then the plant cannot grow enough
to escape. As a plant in such deeply flooded circumstances
grows, it uses up all its energy reserves (carbohydrates),
becomes spindly and weak, and eventually dies.
How does the FR13A variety manage to survive submer-
gence while many other types of rice cannot? FR13A has a
different strategy that could be called sit tight, and SUB1
FIGURE 1-18 Rice is grown in fields with standing water called
paddies. Rice is adapted to tolerate modest levels of standing water, acts as the master switch or regulatory gene to activate this
but the water suppresses the growth of weeds that could compete strategy. When the flood waters rise and the concentration
with the rice. [Debasish Banerjee/Dinodia Photo/AGE Fotostock.] of ethylene increases, the ethylene turns on SUB1, an eth-
ylene response factor. The ERF protein that SUB1 encodes
orchestrates the plant’s response by switching on (or off) a
genetic factor(s) for submergence tolerance into a rice vari- battery of genes involved in plant growth and metabolism.
ety with a higher yield and higher grain quality. He first In FR13A plants that become submerged, genes involved
crossed FR13A and a superior variety of rice and then for in stem and leaf elongation as part of the escape strategy
several generations crossed the hybrid plants back to the are switched off, as are genes involved in mobilizing the
superior variety until he had created an improved form of energy reserves (carbohydrates) needed to fuel the escape
rice that combined submergence tolerance and high yield. strategy. As a result, the plant prevents itself from burning
Mackill had achieved his initial goal of transferring sub- up all its reserve carbohydrates and becoming weak and
mergence tolerance into a superior variety, but the genetic spindly. Using the tools of molecular genetics and genom-
basis for why FR13A was submergence tolerant remained ics such as DNA microarrays (see Chapters 10 and 14),
obscure. Was FR13A’s submergence tolerance controlled by the rice team was able to decipher the extensive catalog of
many genes on multiple chromosomes, or might it be mostly genes controlling organ elongation, carbon metabolism,
controlled by just one gene? To delve into the genetic basis flowering, and photosynthesis that are regulated by SUB1
of submergence tolerance, Mackill and his team conducted to achieve the sit-tight response.
a form of genetic analysis called quantitative trait locus With the basic genetics of SUB1 elucidated, the rice
(QTL) mapping (see Chapter 19). A QTL is a genetic locus team could transfer it into a superior variety with surgi-
that contributes incrementally or quantitatively to varia- cal precision. This precision is important because it enabled
tion for a trait. Unlike in Mendel’s experiments, where one the team to avoid transferring other undesirable genes at
locus controlled one trait, a QTL is just one of multiple loci the same time. For this project, they worked with a sub-
that all affect the same trait. Using QTL mapping, Mackill mergence-intolerant, but superior, Indian variety called
learned that the submergence tolerance trait of FR13A was Swarna, which is widely grown and favored by farmers.
controlled by several QTL, but one of these had a particu- The new line they created, called Swarna-Sub1, has lived up
larly large effect. He named this large-effect QTL SUB1 for to expectations. Field trials showed a striking difference in
“submergence tolerant.” plant survival and yield between Swarna and Swarna-Sub1
To understand molecular nature of SUB1, molecular when there is complete submergence (Figure 1-19). As shown
geneticists Pamela Ronald at the University of California, in Figure 1-20, Swarna-Sub1 provides higher yield than the
Davis, and Julia Bailey-Serres at the University of California, original Swarna under all different levels of flooding. In
Riverside, joined Mackill’s team. This expanded team deter- various trials, the SUB1 improved yield between 1 and 3
mined that SUB1 is a member of a class of genes called eth- tons of grain per hectare.
ylene response factors (ERFs). ERF genes encode regulatory With the support of international research orga-
proteins that bind to regulatory elements in other genes and nizations, governmental agencies, and philanthropies,
thereby regulate their expression. Thus, SUB1 is a gene that Swarna-Sub1 and other superior varieties carrying the SUB1
regulates the expression of other genes. Moreover, they allele from FR13A have now been distributed to farmers.
determined that the allele of SUB1 in FR13A is switched By 2017, an estimated 10 million farmers were growing
on in response to submergence, while the allele of SUB1 SUB1-enhanced rice. Although precise data on how this has
found in submergence-sensitive varieties is not switched on reduced losses due to flooding are not available, the rapid
by submergence. adoption of SUB1-enhanced rice by farmers suggests it is
Flood-intolerant and flood-tolerant rice flooding. With this knowledge, it will be possible to manip-
ulate the genomes of other crop plants so that they too can
withstand getting their feet a little too wet.
condition caused by their inability to obtain enough oxygen associated with improved health at high elevation, its fre-
from the thin air of the mountains. quency would have risen among Tibetans over the many
Since these and other early observations, geneticists generations since they colonized the Tibetan Plateau,
have invested much effort into the study of human adap- because Tibetans with this variant would have been health-
tation to high altitude in South America, Tibet, and Ethi- ier and would have had more surviving children than those
opia. What enables the natives of these regions to flourish who lacked it. Charles Darwin’s natural selection would be
while lowlanders who move to high elevations suffer the at work.
grave health consequences of CMS? Let’s look at the case When the research team analyzed their SNP data, the
in Tibet, where the Tibetan highlanders live at altitudes SNPs in one gene stood out. The gene is called EPAS1, and
up to 4000 meters above sea level (Figure 1-21). The high some SNPs in it occur at very different frequencies in Tibet-
Tibetan Plateau was colonized thousands of years ago by ans (87 percent) and Han Chinese (9 percent). Their results
people who are closely related to the modern Han Chinese. are shown in Figure 1-22. In this figure, the human chro-
However, at high altitude, native Tibetans are far less likely mosomes, numbered 1 through 22, are along the x-axis,
than Han Chinese to experience CMS and conditions such and a measure of the difference in SNP variant frequency
as pulmonary hypertension and the associated formation of between Tibetans and Chinese is on the y-axis. Each dot
blood clots that underlie it. represents a SNP. SNPs that fall above the horizontal red
To understand the genetics of how Tibetans adapted to line are those for which the frequency difference between
life at high elevation, a research team led by Cynthia Beall Tibetans and Han Chinese is so large that the gene near
of Case Western Reserve University compared Tibetans to these SNPs likely provided some advantage to people who
Han Chinese at over 500,000 SNPs across the genome. colonized the Tibetan Plateau. The SNPs in EPAS1 fall
Because Tibetans and Chinese are closely related, she above this line.
expected each SNP variant to occur at about the same fre- These results suggest that Tibetans have a special vari-
quency in both groups. If the T variant of a SNP occurs at ant of EPAS1 that helps them adapt to life at high eleva-
a frequency of 10 percent in Han Chinese, it should also be tion. EPAS1 regulates the number of red blood cells (RBCs)
at about 10 percent in Tibetans. However, if the variant is that our bodies produce in response to the level of oxygen
in our tissues. When oxygen levels in our tissues are low,
EPAS1 signals the body to produce more RBCs. The EPAS1
Tibetans are genetically adapted response to low oxygen may be how our bodies normally
to life at high elevation respond to anemia (too few RBCs). People with low RBC
counts get too little oxygen in their tissues, and so EPAS1
could signal the body to make more RBCs to correct ane-
mia. This mechanism could explain why people who live at
low elevation need the EPAS1 gene.
China
Now, let’s think about how a person from low elevation
would respond if they move to high elevation. Because of
Tibet the thin air at high elevation, their tissues would get less
oxygen. If their bodies interpreted low oxygen due to thin
air as a sign of anemia, then EPAS1 would try to correct the
problem by signaling their body to make more RBCs. How-
ever, since they are not anemic and already have enough
RBCs, their blood would become overloaded with RBCs.
Too many RBCs can cause pulmonary hypertension and the
formation of blood clots, the conditions underlying CMS.
Finally, how does the Tibetan variant of EPAS1 help
them avoid CMS and adapt to high elevation? The Tibetan
version of EPAS1 is expressed at a lower level than the low-
land version, so their bodies are not as stimulated by EPAS1
to overproduce RBCs at high altitude, and thus they avoid
the associated blood clots and pulmonary hypertension.
Remarkably, a single SNP in a regulatory element for EPAS1
seems to be the key genetic variant for this adaptation.
9 EPAS1
6
Statistical test value
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Chromosome
FIGURE 1-22 Twenty-two human chromosomes are arrayed from left to right. The y-axis shows
results from a statistical test of whether there is a significant difference in SNP frequency between
Tibetans and Han Chinese. Each small dot represents one of the SNPs that was tested. SNPs
above the horizontal red line are significantly different. Only the SNPs in the EPAS1 gene show a
significant difference. [Proceedings of the National Academy of Sciences USA, 107, 25 2010,
11459–11464, Fig. 1.]
The complex genetics of color blindness The opsin gene for blue light detection is on one of our
non-sex chromosomes, or autosomes, for which both men and
Look at Figure 1-23a. Everyone should see the number “12.”
women have two copies—one copy from our mother and one
Now, look at Figure 1-23b. Do you see “74” or “21”? If you
copy from our father. Mutations in the blue opsin are rare, and
have normal color vision, you will see “74,” but if you have
only 0.01 percent of people carry mutations in both blue opsin
red-green color blindness, you will see “21.”
alleles, making them unable to distinguish blue and yellow.
Color blindness is an inherited disorder and a favor-
Far more common is red-green color blindness that
ite example among genetics instructors for several reasons.
occurs in about 5 percent of people, mostly men. Why is
(1) The color vision test can be administered in class. (2) Color
this form of color blindness so much more common? First,
blindness is common, so a class of 50 students may have at
the green and red opsin genes are on the X chromosome.
least one affected individual. (3) The genes for the major form
Since men have only one X chromosome, they will have
of color blindness are on the X chromosome, making it a
red-green color blindness if they have a single mutant allele
prime example
Introduction of sex-linked
to Genetic inheritance. To this list, we can
Analysis, 11e
Figure 01.23a #000 in either of these two genes. One in 12 men has this con-
now add fourth compelling reason to study color blindness:
04/02/14 dition. Since women carry two X chromosomes, they need
it has recently been corrected by gene therapy.
05/01/14 two defective alleles of one of these genes, which happens
To understand
Dragonfly Media Groupcolor vision, let’s start with some basics.
less frequently, so only about 0.6 percent of women are red-
We perceive light on our retinas where the surface has two
green color blind. Second, red-green color blindness can be
types of photoreceptor cells—rod cells, which work in low
caused by a mutation in either the red or green opsin gene.
light conditions, and cone cells, which work in high light
Thus, there are two targets for mutation, placing red-green
conditions and allow us to distinguish colors.
color vision in double jeopardy of being lost.
Cone cells come in three forms, depending on which one
But something else is up genetically. Notice that the fre-
of three opsin (light-sensitive protein) genes is expressed
quency of red-green color blindness in women (0.6 percent)
within them. The major opsin genes are short wavelength
is much higher than the frequency of blue-yellow color
(blue), medium wavelength (green), and long wavelength
blindness in women (0.01 percent). Why is this? The red
(red). Each cone cell expresses one of these three genes, cre-
and green opsin genes are neighbors on the X chromosome.
ating a mosaic of red, green, and blue sensory cells on the
retina (Figure 1-24). Deficiency of any one or more of the Red opsin Green opsin
opsin genes can cause a form of color blindness.
FIGURE 1-23 Plates for testing color vision based on the design of Dr. Shinobu Ishihara of the
University of Tokyo. See text for details. [PRISMA ARCHIVO/Alamy; Phanie/Alamy.]
Distribution of blue, red, and green cone Because the red and green opsin genes are neighbors,
cells on the retina. they can undergo a process called unequal crossing over
that can produce the chromosomes shown below, in which
either the green opsin is missing or a hybrid gene is formed.
Either of these causes red-green color blindness.
Red opsin
Hybrid opsin
FIGURE 1-25 Dalton the squirrel monkey received gene therapy for color vision. (a) How Dalton saw the
world through his original red-green color-blind eyes; (b) Dalton enjoying his new color sense after gene
therapy. [Dr. Jay Neitz, Neitz Lab, University of Washington.]
but after gene therapy, the monkeys could tell green and red gene therapy for color blindness sets the stage for treating
apart. Gene therapy worked. more severe eye diseases such as macular degeneration.
Why should we treat color blindness through gene With advances in technology, we can anticipate a day when
therapy? This condition does impair people’s lives. Red- gene therapy for visual impairments is no more costly or
green color-blind individuals have difficulty distinguishing complicated than putting on a pair of contact lenses.
cooked (brown) from partly cooked (pink) meat or ripe
from green fruits. They cannot be commercial pilots or air
KEY CONCEPT Genetic tools allow gene therapy to correct
traffic controllers, and they generally encounter difficulty
some disorders caused by mutant genes.
navigating a world designed for trichromats. Moreover,
SUMMARY
As you begin your study of genetics, imagine yourself as a • How do cells manage to seamlessly orchestrate the
person at halftime during an amazing journey of discovery. incredibly complex array of interacting genes and bio-
The last 100 years have witnessed a remarkable revolution chemical reactions that are found within them?
in human knowledge about how biological systems are put • How do genetic variants at hundreds or even thousands
together and how they work. Genetics has been at the epicen- of genes control the yield of crop plants?
ter of that revolution. Genetic analysis has answered many
• How can genetics guide both the prevention and treat-
fundamental questions about the transmission of genetic infor-
ment of cancer, autism, and other diseases?
mation within families, inside cells, and over the eons of evo-
lutionary time. Yet, as you will learn, the discovery process in • How do genes give humans the capacity for language
genetics has never been more dynamic and the pace of growth and consciousness?
in knowledge never greater. Unanswered questions abound. Genetic analysis over the next 100 years promises to help
• How do all the genes in the genome work together to answer many questions such as these.
transform a fertilized egg into an adult organism?
KEY TERMS
adenine (A) (p. 6) genetically modified organism point mutation (p. 14)
allele (p. 3) (GMO) (p. 11) quantitative trait locus (QTL)
codon (p. 8) genetics (p. 4) (p. 17)
complementary (base pairs) (p. 7) genomics (p. 11) regulatory element (p. 7)
cytosine (C) (p. 7) guanine (G) (p. 6) single nucleotide polymorphism
DNA polymerase (p. 10) ligase (p. 10) (SNP) (p. 13)
DNA replication (p. 8) messenger RNA (mRNA) (p. 8) thymine (T) (p. 6)
DNA sequencing (p. 11) model organism (p. 9) transcription (p. 8)
dominant (p. 3) multifactorial hypothesis (p. 6) transformation (p. 10)
gene (p. 3) nuclease (p. 10) translation (p. 8)
gene expression (p. 7) one-gene–one-enzyme hypothesis unequal crossing over (p. 21)
(p. 6)
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the www
WORKING WITH THE FIGURES 7. Figure 1-13b shows the human chromosomes in a
single cell. The yellow dots show the location of the
1. If the white-flowered parental variety in Figure 1-3
muscle glycogen phosphorylase gene. Is the cell in this
were crossed to the first-generation hybrid plant in
figure a sex cell (gamete)? Explain your answer.
that figure, what types of progeny would you expect to
see, and in what proportions? 8. Figure 1-15 shows the family tree, or pedigree, for
Louise Benge (Individual VI-1) who suffers from the
2. In Mendel’s 1866 publication as shown in Figure 1-4,
disease ACDC because she has two mutant copies of
he reports 705 purple-flowered (violet) offspring and
the CD73 gene. She has four siblings (VI-2, VI-3, VI-4,
224 white-flowered offspring. The ratio he obtained
and VI-5) who have this disease for the same reason.
is 3.15:1 for purple to white. How do you think he
Do all of the 10 children of Louise and her siblings
explained the fact that the ratio is not exactly 3:1?
have the same number of mutant copies of the CD73
3. Figure 1-7 shows a simplified pathway for arginine gene, or might this number be different for some of the
synthesis in Neurospora. Suppose you have a special 10 children?
strain of Neurospora that makes citrulline but not
arginine. Which gene(s) are likely mutant or missing
in your special strain? You have a second strain of BASIC PROBLEMS
Neurospora that makes neither citrulline nor arginine 9. State four questions about inheritance that arose after
but does make ornithine. Which gene(s) are mutant or Mendel’s rules of inheritance were rediscovered.
missing in this strain?
10. Name four tools or enzymes that molecular geneticists
4. Consider Figure 1-8a. can use to manipulate DNA or RNA molecules.
a. What do the small, blue spheres represent? 11. Below is the sequence of a single strand of a short
b. What do the brown slabs represent? DNA molecule. On a piece of paper, rewrite this
c. Do you agree with the analogy that DNA is struc- sequence and then write the sequence of the comple-
tured like a ladder? mentary strand below it.
a. Under blending theory, how long would you expect aqueous solutions by heating to yield two single
the stems of first and second hybrids to be? strands of DNA (see Figure 1-13a). How would you
b. Under Mendelian rules, and assuming stem length expect the relative amounts of GC versus AT base
is controlled by a single gene, what would you pairs in a DNA double helix to affect the amount of
expect to observe in the second-generation hybrids heat required to denature it? How would you expect
if all the first-generation hybrids were tall? the length of a DNA double helix in base pairs to
affect the amount of heat required to denature it?
13. If a DNA double helix that is 100 base pairs in length
has 32 adenines, how many cytosines, guanines, and 15. The figure below shows the DNA sequence of a portion
thymines must it have? of one of the chromosomes from a trio (mother, father,
and child). Can you spot any new point mutations in
14. The complementary strands of DNA in the double the child that are not in either parent? In which parent
helix are held together by hydrogen bonds: G ≡ C or did the mutation arise?
A = T. These bonds can be broken (denatured) in
Mother Father
Copy M1 • • C AGCAGA T TGCTGC T T TGT A TGAG • • Copy F1 • • C AGC TGA T TGCTGC T T TGT AGGAG • •
Copy M2 • • CAGC TGA T TGCTGC T T TGT A TGAG • • Copy F2 • • CA A C TGA T TGCTGC T T TGT A TGAG • •
Child
05/12/14
do women. How can you explain this discrepancy? www
Dragonfly Media Group
18. The human genome is approximately 3 billion base GENETICS AND SOCIETY
pairs in size. In this chapter, you read how older fathers contribute more
a. Using standard 8.5" × 11" paper with one-inch new point mutations to their offspring than younger fathers.
margins, a 12-point font size, and single-spaced Should this information be used to inform public policy or
lines, how many sheets of paper printed on one side individual decisions? What policy decisions might a society
would be required to print out the human genome? or an individual make and implement in this regard?
Core Principles in
Transmission Genetics
T
he subject of this book is the science of genetics, which is broadly defined as the CHAPTER 2
SINGLE-GENE INHERITANCE
study of heredity. Heredity in turn is defined as the transmission of properties from
generation to generation. Since the beginning of time, humans have no doubt won- CHAPTER 3
INDEPENDENT ASSORTMENT OF GENES
dered about how heredity is possible. Generally speaking, heredity has two equally myste-
rious components, constancy and variation. CHAPTER 4
Constancy is the simple fact that people have human babies, cats have kittens, dogs MAPPING EUKARYOTE CHROMOSOMES
BY RECOMBINATION
have puppies, and so on. Thus, the constancy of the species is transmitted down through
the generations. Of course, this general observation about heredity raises the question, how CHAPTER 5
GENE INTERACTION
is such constancy possible? In other words, what is the mechanism that allows faithful
reproduction of a species? If it is assumed that two parents are needed to beget offspring, CHAPTER 6
common sense suggests that some material must be transmitted from each parent to con- THE GENETICS OF BACTERIA AND THEIR
VIRUSES
tribute to the offspring. However, for most of the history of mankind, the nature of this
material was a mystery.
Variation is the common observation (using humans as an example) that although
people beget people, there is considerable difference among the members of a population
generally, and also within the progeny of a mating. These differences allow us to recognize
individuals in the general pool of the human population. Such variation is concerned with Kirk and Michael Douglas
essentially minor aspects of a species. Although the main characteristics that define the
species are held constant (we are bipedal, big-brained, have generally hairless bodies, etc.),
there are clear differences in properties such as hair color, nose shape, skin color, height,
predisposition to certain diseases, and a myriad of other qualities.
We can contrast two types of variation. The first is called continuous variation. A good
example is height or length, which tends to vary in the population from a low value to a
high value and all values between. The second is discontinuous variation, the existence in a
species of individuals with markedly distinct forms of a particular property. In humans, a
good example is the presence or absence of a chin dimple. This is not continuous variation;
Michael Douglas (right), son of Kirk Douglas
people either have a chin dimple or none. The movie actor Kirk Douglas has a prominent (left), has inherited his father’s chin dimple.
chin dimple, and furthermore his son Michael Douglas (also an actor) has clearly inherited [Sunset Boulevard/Getty Images; ScreenProd/
his chin dimple. Photononstop/Alamy.]
Another example from history is the Hapsburg lip, a striking discontinuous variant
that was handed down through generations of the Spanish Hapsburg royal family. We see
around us many other examples of the clear inheritance of such distinct properties in both The Hapsburg Lip
animals and plants. A mechanism is needed that explains not only how such variation
arises, but also how it is transmitted from one generation to the next, that is, a transmission
mechanism.
The inheritance of continuous variation is less obvious, and often no clear pattern of
inheritance is immediately obvious. An important complicating factor regarding continu-
ous variation is that environmental effects can have a profound influence. A simple exam-
ple is weight, which varies continuously. Although in some cases it can seem to be inherited,
the weight of an individual is obviously greatly influenced by the availability of food, a
crucial component of the environment.
The existence of variation and its inheritance also demands an explanation. How are Philip IV (left) and his son Charles II (right) were
the last of the Hapsburgs to rule Spain. Both
variants produced, and how are these variants passed on to subsequent generations?
ruled in the seventeenth century, and both had
In summary of the above preliminary ideas, we can set down several overarching obser- a Hapsburg lip. [DEA/G. NIMATALLAH/Getty
vations about heredity: Images; Heritage Images/Getty Images.]
25
PRINCIPLES OF HEREDITY
1. DNA is the genetic material that determines the basic properties of
an organism.
We now know that each cell of an organism has a fundamental and unique set of
DNA, called the genome, which has encoded in it the information for building that
organism. DNA is a long filamentous molecule made up of many thousands of func-
tional units called genes. The genome of eukaryotic organisms is composed of several
DNA molecules, each coiled up as a chromosome, with each chromosome bearing
many genes. In most cases, the information in each gene is translated into a protein,
and these many proteins are the essential units of form and function in an organism.
Hence, overall
3. During cell division, daughter DNA molecules are packaged into the
resulting cells.
Somatic cells in eukaryotes divide to produce more cells of the same organism, to
increase cell number during growth. During this cell division, the accompanying
nuclear division called mitosis ensures that each daughter cell receives one daughter
of each chromosome, the same number and identity as in the parent cell.
28
Single-Gene Inheritance
W
hat kinds of research do biologists do? One variants with the standard and, from these comparisons,
central area of research in the biology of all to make deductions about genetic function. It is similar to
organisms is the attempt to understand how an the way in which you could make inferences about how an
organism develops from a fertilized egg into an adult—in unfamiliar machine works by changing the composition or
other words, what makes an organism the way it is. Usu- positions of the working parts, or even by removing parts
ally, this overall goal is broken down into the study of one at a time. Each variant represents a “tweak” of the bio-
individual biological properties such as the development logical machine, from which its function can be deduced.
of plant flower color, or animal locomotion, or nutrient In genetics, the most common form of any trait of an
uptake, although biologists also study broader areas such organism is called the wild type, that which is found “in
as how cells work. How do geneticists analyze biological the wild,” or in nature. The heritable variants observed in a
properties? In genetics, individual biological properties of species that differ from wild type are called mutants, indi-
a species are referred to as characters or traits. The genetic vidual organisms having some abnormal form of the trait.
approach to understanding any biological trait is to find The alternative forms of a trait are called phenotypes, for
the subset of genes in the genome that influence it, a pro- example, the blue and white phenotypes of the flower color
cess sometimes referred to as gene discovery. After these trait. Figure 2-1 shows examples of the wild-type phenotype
genes have been identified, their cellular functions can be and several mutant phenotypes for a given trait in two dif-
elucidated through further research. ferent model organisms.
There are several different types of analytical Compared to wild type, mutants are rare. We know
approaches to gene discovery, but one widely used method that they arise from wild types by a process called
relies on the detection of single-gene inheritance patterns, mutation, which results in a heritable change in the DNA
and that is the topic of this chapter. of a gene. The changed form of the gene is also called a
All of genetics, in one aspect or another, is based on her- mutation. Mutations are not always detrimental to an
itable variants: individuals who inherit a trait that expresses organism; sometimes they can be advantageous, but most
differently from some standard form. For example, in often they have no observable effect. A great deal is known
regard to the flower color trait in some plant, it might about the mechanisms of mutation (see Chapter 15), but
express white in the variant instead of the normal blue. The generally it can be said that they arise from mistakes in cel-
basic approach of genetics is to compare the properties of lular processing of DNA.
WT lfy ap1 WT
ap2 ap3 ag
FIGURE 2-1 These photographs show the range of mutant phenotypes typical of those obtained
in the genetic dissection of traits. These examples are from the dissection of floral development in
Arabidopsis thaliana, a plant (a) and hyphal growth in Neurospora crassa, a mold (b). WT = wild type .
[(a) courtesy of George Haughn; (b) courtesy of Anthony Griffiths/Olivera Gavric.]
30
Simply stated, the general steps in analyzing a trait its underlying genetic program, not with a scalpel but with
through gene discovery are as follows: mutants. Each mutant potentially identifies a separate gene
affecting that trait.
1. Amass mutants affecting the trait of interest.
After a set of key genes has been defined in this way,
2. Cross (mate) mutant individuals to wild-type indi- several different molecular methods can be used to estab-
viduals to see if their descendants show ratios of lish the functions of each of the genes. These methods will
wild-type-to-mutant phenotypes that are characteristic be covered in later chapters.
of single-gene inheritance. This type of approach to gene discovery is sometimes
3. Deduce the functions of the gene at the molecular level. called forward genetics, a strategy to understanding bio-
4. Deduce how the gene interacts with other genes to pro- logical function starting with random single-gene mutants
duce the trait in question. and ending with their DNA sequence and biochemical
function. (In later chapters, we shall see reverse genetics at
Of these steps, only 1 and 2 will be covered in the present work. In brief, reverse genetics starts with genomic analysis
chapter. at the DNA level to identify a set of genes as candidates
Gene discovery starts with a “hunt” to amass mutants for encoding the biological trait of interest, then induces
in which the biological function under investigation is mutants targeted specifically to those genes, and then
altered or destroyed. Even though mutants are individu- examines the mutant phenotypes to see if they indeed affect
ally rare, there are ways of enhancing their recovery. One the trait under study.)
widely used method is to treat the organism with radiation
or chemicals that increase the mutation rate. After treat- KEY CONCEPT The genetic approach to understanding
ment, the most direct way to identify mutants is to visu- a biological trait is to discover the genes that control it.
ally screen a very large number of individuals, looking for One approach to gene discovery is to isolate mutants and
a chance occurrence of mutants in that population. Also, check each one for single-gene inheritance patterns (spe-
various selection methods can be devised to enrich for the cific ratios of wild-type and mutant expression of the trait in
types sought. descendants).
Armed with a set of mutants affecting the trait of inter-
est, one hopes that each mutant represents a mutation in
one of a set of genes that control the trait, and that a rea- Gene discovery is important not only in experimental
sonably complete gene pathway or network is represented. organisms, but also in applied studies. One crucial area is
However, not all mutants are caused by a single mutation agriculture, where gene discovery can be used to under-
within one gene (some have far more complex determina- stand a desirable commercial property of an organism,
tion), so first each mutant has to be tested to see if indeed it such as its protein content. We have already encountered
is caused by a single-gene mutation. an example of the power of genetics to affect agriculture in
The test for single-gene inheritance is to mate individu- Chapter 1, where we saw how genetic analysis facilitated
als showing the mutant phenotype with wild-type individ- the creation of a flood-resistant rice strain. Human genet-
uals, and then to analyze the first and second generations ics is another important area: to know which gene func-
of descendants. As an example, a mutant plant with white tions are involved in a specific disease or condition is useful
flowers would be crossed to a wild-type plant showing blue information in finding therapies, such as the gene therapy
flowers. The progeny of this cross are analyzed, and then technique being developed to treat color blindness, as dis-
they themselves are interbred to produce a second gener- cussed in Chapter 1.
ation of descendants. In each generation, the diagnostic The rules for single-gene inheritance were originally
ratios of plants with blue flowers to those with white flow- elucidated in the 1860s by the monk Gregor Mendel, who
ers will reveal whether a single gene controls white versus worked in a monastery in the town of Brno, now part of
blue flower color. If so, then by inference, the wild-type the Czech Republic. Mendel’s analysis is the prototype of
phenotype (blue flowers) would be encoded by the wild- the experimental approach to single-gene discovery still
type form of the gene, and the mutant phenotype (white used today. Indeed, Mendel was the first person to dis-
flowers) would be encoded by a form of the same gene in cover any gene! Mendel did not know what genes were,
which a mutation event has altered the DNA sequence in how they influenced traits, or how they were inherited at
some way. Other mutations affecting flower color (per- the cellular level. Now we know that genes either encode
haps mauve, blotched, striped, and so on) would be ana- proteins or RNA molecules that facilitate or regulate pro-
lyzed in the same way, resulting overall in a set of defined tein expression, a topic that we shall return to in later
“flower-color genes.” Through genetics, the set of gene chapters. We also know that single-gene inheritance pat-
functions that interact to produce the trait we call flower terns are produced because genes are parts of chromo-
color can be defined. The use of mutants in this way is somes, and chromosomes are partitioned very precisely
sometimes called genetic dissection, because the trait in down through the generations, as we shall see later in this
question (flower color in this case) is picked apart to reveal chapter.
Cross-pollination Selfing
Stigma
Progeny Progeny
FIGURE 2-3 In a cross of a pea plant (left), pollen from the anthers of one plant is transferred to the
stigma of another. In a self (right), pollen is transferred from the anthers to the stigmata of the same
plant.
Mendel grew F1 peas into plants, and he selfed these plants to To put it another way, the F2 was comprised of
obtain the second filial generation, or F2. The F2 was com- 1
pure-breeding greens
posed of 6022 yellow peas and 2001 green peas. In summary, 4
1
2
F1-like yellows (mixed progeny)
yellow F1 × yellow F1 → F2 comprised of 6022 yellow 1
4
pure-breeding yellows
2001 green
Thus, the 3:1 ratio of phenotypes observed in the F2 genera-
Total 8023
tion, at a more fundamental level, is a 1:2:1 ratio.
Mendel noted that this outcome was very close to a Mendel made another informative cross between the
mathematical ratio of three-fourths (75%) yellow and one- F1 yellow-seeded plants and any green-seeded plant. In this
fourth (25%) green. A simple calculation shows us that cross, the progeny showed the proportions of one-half yel-
6022/8023 = 0.751 or 75.1%, and 2001/8023 = 0.249 or low and one-half green. In summary,
24.9%. In other words, there was a 3:1 ratio of yellow to 1
F1 yellow × green → yellow
green. Interestingly, the green phenotype, which had dis- 2
1
appeared in the F1, had reappeared in one-fourth of the 2
green
F2 individuals, showing that the genetic determinants for
These two types of matings, the F1 self and the cross of
green must have been present in the yellow F1, although
the F1 with any green-seeded plant, both gave yellow and
unexpressed (not observed in the F1 phenotype).
green progeny, but in different ratios. These two ratios are
To further investigate the nature of the F2 plants, Mendel
represented in Figure 2-4. Notice that the ratios are seen
selfed plants grown from the F2 seeds. He found three dif-
only when the peas in several pods are combined.
ferent types of results. The plants grown from the F2 green
The 3:1 and 1:1 ratios found for pea color were also
seeds, when selfed, were found to bear only green peas. How-
found for comparable crosses for the other six characters
ever, plants grown from the F2 yellow seeds, when selfed,
that Mendel studied. The actual numbers for the 3:1 ratios
were found to be of two types: one-third of them were pure
for those characters are shown in Table 2-1.
breeding for yellow seeds, but two-thirds of them gave mixed
progeny: three-fourths yellow seeds and one-fourth green
seeds, just as the F1 plants had. In summary,
1
Mendel’s law of equal segregation
4
of the F2 were green, which when selfed gave all
greens Initially, the meaning of these precise and repeatable
mathematical ratios must have been unclear to Mendel,
3
4
of the F2 were yellow; but he was able to devise a brilliant model that not only
of these 13 when selfed gave all yellows accounted for all the results, but also represented the his-
3
23 when selfed gave 4
yellow and 1
4
green torical birth of the science of genetics. Mendel’s model for
either
F2 F2
Progeny Progeny
seeds seeds
Total 21 7 Total 11 11
FIGURE 2-4 Mendel obtained a 3:1 phenotypic ratio in his self-pollination of the F1 (left) and a 1:1
phenotypic ratio in his cross of F1 yellow with green (right). Sample sizes are arbitrary.
TABLE 2-1 Results of All Mendel’s Crosses in Which Parents Differed in One Character
Parental phenotypes F1 F2 F2 ratio
1. round × wrinkled seeds All round 5474 round; 1850 wrinkled 2.96:1
2. yellow × green seeds All yellow 6022 yellow; 2001 green 3.01:1
3. purple × white petals All purple 705 purple; 224 white 3.15:1
4. inflated × pinched pods All inflated 882 inflated; 299 pinched 2.95:1
5. green × yellow pods All green 428 green; 152 yellow 2.82:1
6. axial × terminal flowers All axial 651 axial; 207 terminal 3.14:1
7. long × short stems All long 787 long; 277 short 2.84:1
the pea-color example, translated into modern terms, was 4. A plant can be either Y /Y , y /y, or Y /y . The slash shows
as follows: that the alleles are a pair.
1. A hereditary factor called a gene is necessary for produc- 5. The phenotype of Y /y plants is always yellow, even
ing pea color. though an allele for the green phenotype is present. One
could say that the Y allele “dominates” over the y allele,
2. Each plant has a pair of this type of gene.
resulting in the yellow phenotype. The allele whose
3. The gene comes in two forms called alleles. If the gene phenotype is displayed in the Y /y plant, Y, is called
is phonetically called a “wye” gene, then the two alleles the dominant allele. The allele whose phenotype is not
can be represented by Y (standing for the yellow pheno- displayed in the Y /y plant, y, is known as the recessive
type) and y (standing for the green phenotype). allele.
6. In meiosis, the members of a gene pair separate equally only breed true. When crossed with each other, the Y /Y and
into the cells that become eggs and sperm, the gametes. the y /y lines produce an F1 generation composed of all het-
This equal partitioning has become known as Mendel’s erozygous individuals (Y /y) . Because Y is dominant, all F1
first law or as the law of equal segregation. Hence, a sin- individuals are yellow in phenotype. Selfing the F1 individu-
gle gamete contains only one member of the gene pair. als can be thought of as a cross of the type Y /y ×Y /y, which
7. At fertilization, gametes fuse randomly, regardless of is sometimes called a monohybrid cross. Equal segregation
which of the alleles they bear. of the Y and y alleles in the heterozygous F1 results in gam-
etes, both male and female, half of which are Y and half of
Here, we introduce some terminology. A fertilized egg, which are y. Male and female gametes fuse randomly at fer-
the first cell that develops into a progeny individual, is tilization, with the results shown in the grid in Figure 2-5.
called a zygote. A plant with a pair of identical alleles for a The composition of the F2 is three-fourths yellow seeds
given gene is called a homozygote (adjective homozygous), and one-fourth green, a 3:1 ratio. The one-fourth of the F2
and a plant in which the alleles of the gene pair differ is seeds that are green breed true as expected of the genotype
called a heterozygote (adjective heterozygous). Sometimes y /y . However, the yellow F2 seeds (totaling three-fourths)
a heterozygote for one gene is called a monohybrid. An are of two genotypes: two-thirds of them are heterozygotes
individual can be classified as either homozygous dominant Y /y , and one-third are homozygous dominant Y /Y . Under-
(such as Y /Y ), heterozygous (Y /y) , or homozygous reces- lying the 3:1 phenotypic ratio in the F2 is a 1:2:1 genotypic
sive (y /y). In genetics, allelic combinations underlying phe- ratio:
notypes are called genotypes. Hence, Y /Y , Y /y , and y /y are
}
1
all genotypes. 4
Y /Y yellow
3
4
yellow (Y /−)
2
4
Y /y yellow
KEY CONCEPT At meiosis, the members of a gene pair seg- 1
y /y green
regate equally into the product cells (often sperm or eggs). This 4
is known as Mendel’s first law or the law of equal segregation. The general representation of an individual expressing
the dominant allele is Y / −; the dash represents a slot that
Figure 2-5 shows how Mendel’s postulates explain the can be filled by either another Y or a y. Note that equal seg-
progeny ratios illustrated in Figure 2-4. The pure-breeding regation is detectable only in the meiosis of a heterozygote;
lines are homozygous, either Y /Y or y /y . Hence, each line Y /y produces one-half Y gametes and one-half y gametes.
produces only Y gametes or only y gametes and thus can Although equal segregation is taking place in homozygotes,
Pure Pure
P × P Y/Y × y /y
F1 F1 Y/y
Equal segregation
×
Selfed Crossed with
× y /y
green
F2
1 1
2 Y 2 y all y
1 1 1
4 Y /Y 4 Y/y 2 Y/y
F2 1 1
3 1 2 Y 2 Y
4 2
1 1 1
4 Y/y 4 y /y 2 y /y
1 1
1 1 2 y 2 y
4 2
FIGURE 2-5 Mendel’s results (left) are explained by a single-gene model (right) that postulates the
equal segregation of the members of a gene pair into gametes.
too, neither segregation 12 Y : 12 Y nor segregation 12 y : 12 y is are involved in a highly programmed set of moves that partition
meaningful or detectable at the genetic level. them into new cells. We must now turn to these movements.
We can now also explain results of the cross between There are two types of cell division in eukaryotes, each
the plants grown from F1 yellow seeds (Y /y) and the plants with unique consequences. Somatic cell division is division
grown from green seeds (y /y). In this case, equal segregation of cells of the main body, known as the soma. The prod-
in the yellow heterozygous F1 gives gametes with a 12 Y : 12 y ucts of somatic cell division are exact copies of the parent
ratio. The y /y parent can make only y gametes, however; so cell. Sexual cell division takes place in sex organs. Spe-
the phenotype of the progeny depends only on which allele cialized cells called meiocytes divide to produce sex cells
they inherit from the Y /y parent. Thus, the 12 Y : 12 y gametic such as sperm and eggs in plants and animals, or sexual
ratio from the heterozygote is converted into a 12 Y /y : 12 y /y spores in fungi or algae. When cells divide, so do their
genotypic ratio, which corresponds to a 1:1 phenotypic nuclei; somatic cell nuclear division is called mitosis, and
ratio of yellow-seeded to green-seeded plants. This is illus- sexual cell nuclear division is called meiosis. The life cycles
trated in the right-hand panel of Figure 2-5. of some well-known organisms, showing when mitosis and
meiosis occur, are found in Figure 2-6.
KEY CONCEPT Equal segregation of a gene pair during
eiosis is observable only in heterozygotes, making them
m KEY CONCEPT During somatic cell division, the accompany-
critical tools for genetic analysis. In a cross between a ing nuclear division is mitosis. During sexual cell division, the
heterozygote and a homozygous recessive, the gametic ratio accompanying nuclear division is meiosis.
produced by meiosis in the heterozygote is observed in the
phenotypic ratio of the progeny.
Notice that in Figure 2-6, in the animal and plant examples,
the cells of the adult body are labeled 2n. In this terminology,
Note that, in defining the allele pairs that underlay his
n = the number of chromosomes in the genome, and the num-
phenotypes, Mendel had identified a gene that radically
ber 2 indicates that there are two genomes (chromosome sets)
affects pea color. This identification was not his prime
per adult cell. 2n somatic cells are called diploid. A photograph
interest, but we can see how finding single-gene inheritance
of a diploid cell in a muntjac deer is shown in Figure 2-7. Note
patterns is a process of gene discovery, identifying individ-
that in a diploid cell, the chromosomes are in pairs (there are n
ual genes that influence a biological trait.
pairs); the two members of a pair are called homologous chro-
KEY CONCEPT All 1:1, 3:1, and 1:2:1 genetic ratios are mosomes, or homologs. The third example in Figure 2-6 shows
diagnostic of single-gene inheritance and are based on equal a haploid organism (in this case a fungus) whose somatic cells
segregation in a heterozygote. have just one chromosome set, n. A large proportion of organ-
isms on the planet are haploid.
Mendel’s research in the mid-nineteenth century was
KEY CONCEPT The somatic cells of diploid organisms con-
not noticed by the international scientific community until tain two copies of each chromosome. The somatic cells of
similar observations were independently published by haploid organisms contain one copy of each chromosome.
several other researchers in 1900. Soon research in many
species of plants, animals, fungi, and algae showed that
Mendel’s law of equal segregation was applicable to all sex- To understand chromosome segregation, we must first
ual eukaryotes and, in all cases, was based on the chromo- understand and contrast the two types of nuclear divisions
somal segregations taking place in meiosis, a topic that we that take place in eukaryotic cells. Somatic cell division and the
turn to in the next section. accompanying nuclear division (mitosis) is a programmed stage
of all eukaryotic cell-division cycles (Figure 2-8). Mitosis can
take place in diploid or haploid cells. As a result, one progenitor
2.2 GENES AND CHROMOSOMES cell becomes two genetically identical cells. Hence,
either 2n → 2n + 2n
LO 2.2 Explain single-gene inheritance ratios in terms of
or n → n + n
chromosome behavior at meiosis.
This “trick” of constancy is accomplished when each
Mendel explained his inheritance ratios by postulating hypo- chromosome replicates to make two identical copies of
thetical entities that he called heritable factors, but he did not itself, with underlying DNA replication. The two identical
know specifically what they were, or where they were located. copies of each chromosome are pulled to opposite ends of
Today we know that Mendel’s heritable factors are genes, and the cell. When the cell divides, each daughter cell has the
genes are located on chromosomes. In this section, we correlate same chromosomal set as its progenitor.
gene behavior with chromosome behavior and, in doing so, In sexual cell division two sequential divisions take place,
explain the chromosomal basis of gene inheritance. along with two nuclear divisions (meiosis). Because there are
In the normal day-to-day chemistry of the cell, genes carry two divisions, four cells are produced from each progenitor
out their activities on chromosomes that are relatively immobile cell. Meiosis takes place only in diploid cells, and the result-
within the nucleus. However, when cells divide, chromosomes ing gametes are haploid. Hence, the net result of meiosis is
2n Meiocytes 2n
Meiosis Meiosis
2n Meiocytes 2n n n
n n n n n ×
× n ×
n n
Tetrad Tetrad
Meiosis Meiosis
Mitosis Mitosis 2n Transient
diploid cell
n n n n n n ×
× n ×
n (meiocyte)
Tetrad Tetrad gp gp gp gp gp
Meiosis
sperm egg
(gamete) (gamete)
Sperm n n Egg
Sexual
2n Zygote n n n n spores
Tetrad
Mitosis
Animal Fungus
Plant
FIGURE 2-6 The life cycles of humans, plants, and fungi, showing the structure composed of haploid cells that will produce gametes. In many
points at which mitosis and meiosis take place. Note that in the females plants such as corn, a nucleus from the male gametophyte fuses with
of humans and many plants, three cells of the meiotic tetrad abort; only two nuclei from the female gametophyte, giving rise to a triploid (3n)
one develops into an egg. The abbreviation n indicates a haploid cell, cell, which then replicates to form the endosperm, a nutritive tissue that
2n a diploid cell; “gp” stands for gametophyte, the name of the small surrounds the embryo (which is derived from the 2n zygote).
FIGURE 2-7 The six visible chromosomes are from a cell caught in
the process of nuclear division. The three pairs of chromosomes have
been stained with chromosome-specific DNA probes, each tagged
with a different fluorescent dye (chromosome paint). A nucleus derived
from another cell is at the stage between divisions. [Republished with
permission of Annual Reviews, from Ferguson-Smith, Malcolm A., “Putting
Medical Genetics into Practice,” Annual Review of Genomics and Human
Genetics, 2011, September; 12: 1–23, Figure 2. Permission conveyed
through Copyright Clearance Center, Inc.]
FIGURE 2-8
A a
A a
A a A a
+
A a
A a A a
ANIMATED ART
Meiosis
2n
Replication
Meiosis
2n
Replication
Pairing
We see that mitosis faithfully produces two cells of the over two cell divisions. In the first division, the centromere
same genotype as the original cell: holding a pair of sister chromatids together does not divide.
One pair of chromatids is pulled into each daughter cell
Aa → Aa + Aa
by spindles that attach to the undivided centromeres. At
Meiosis is also preceded by replication and chromo- the second division of meiosis, the centromeres divide; and
some condensation, but a key difference is that in this case, now each chromatid is pulled into its own cell, which is
the two homologous chromosomes pair to form a group of now effectively haploid. Meiosis in a diploid heterozygote
four chromatids at the equatorial plane. Meiosis takes place Aa is shown here:
A
A +
A A
A A
A
A a A a A A
a +
A a a a
a a a
a
a +
a
a
2n
Segregation
Products of
Telophase II meiosis
n
Prophase II Metaphase II Anaphase II
Telophase I
Anaphase I n
Segregation
n
Segregation
We can see that a single diploid cell of genotype Aa pro- Demonstration of equal segregation within one
duces four haploid cells, two of genotype A and two of gen- meiocyte in the yeast S. cerevisiae
otype a. Hence, we now see the underlying chromosomal
mechanism that produces the 1:1 gametic ratio postulated by r + culture r culture
Mendel:
Aa → A + A + a + a
Aa → 1/2 A and 1/2 a
ratio = 1A : 1a
+ r
Figure 2-9 shows the named stages of mitosis and meio- n n
sis in a diploid organism. Note the difference in alignment
of homologous chromosomes in metaphase of mitosis and
metaphase I of meiosis, and the way the chromatids are Mix cells to make cross
partitioned during the subsequent anaphase.
Diploid
Single-gene inheritance in haploids
+
Mitosis in haploids progresses in much the same way, but 2n
in haploid organisms, each somatic cell bears only one r
chromosome set. If we choose to use the A and a terminol-
ogy to represent alleles of a gene, a cell can be either A or a.
Mitosis occurs as shown here: Chromosome replication
Meiocyte
A +
+
A r
A +
A r
A A
A
OR Products of
first division
a a + r
a a
+ r
a +
a
a
Ascus wall
+
Four products r
of meiosis:
Overall, 1:1 ratio +
of r + : r r
A → A + A, or
a → a+ a
In haploids, meiosis takes place at one special stage of
the life cycle when two haploid cells unite to form a tran- Inoculate cells to form
sient diploid meiocyte. This cell union is a type of sexual colonies, which
demonstrate + +
union, although haploids generally do not have true sexes. r
single-gene r
However they do have mating types, which can be thought segregation in
of as simple forms of sexes. The fungi we shall study in this one meiocyte.
book have two mating types, determined by two alleles of a
single gene. Meiocytes form only from the union of cells of FIGURE 2-10 One ascus isolated from the cross r + × r leads to two
different mating types. cultures of r + and two of r.
Let’s look at a cross in the haploid fungus baker’s yeast, level by using modern technology, they are generally found
involving a red mutant, which contrasts with the normal to be identical in most of their sequences and differ only at
white color of wild type. We will postulate a meiocyte one or several nucleotides of the hundreds or thousands of
formed by the union of a haploid cell of a red mutant, nucleotides that make up the gene. Therefore, we see that
r, and a haploid cell of wild type r + (the + superscript is the alleles are truly different versions of the same gene. The
commonly used to designate wild type alleles). The tran- following diagram represents the DNA of two alleles of one
sient diploid meiocyte is a heterozygote r + /r. As expected gene; the letter x represents a difference in the nucleotide
from our previous discussion of a diploid meiosis, the four sequence:
haploid cells produced are 1/2 r and 1/2 r + , also reflecting
Allele 1
Mendel’s first law of equal segregation.
It is noteworthy that in many haploid organisms, espe- Allele 2 x
cially fungi, the four cells that are the products of a single
meiosis remain together enclosed in a membranous sac. In If the nucleotide sequence of an allele changes as the
yeast, this sac is called an ascus. The four haploid nuclear result of a rare chemical “accident,” a new mutant allele is
products in the ascus represent the meiotic tetrad. Thus, it created. Such changes can occur anywhere along the nucle-
is possible to perform analyses of single meioses by using otide sequence of a gene. For example, a mutation could be
a needle to separate and isolate each of the contents of the a change in the identity of a single nucleotide, or the dele-
sac, and such tetrad analyses have made valuable contribu- tion of one or more nucleotides, or even the addition of one
tions to research on the analysis of meiosis itself, and on or more nucleotides.
single-gene inheritance. The yeast cross of r × r + is shown A gene can be changed by mutation in many ways. For
in detail in Figure 2-10. one thing, the mutational damage can occur at any one of
many different sites. We can represent the situation as fol-
KEY CONCEPT Mitotic division results in the original lows, where dark blue indicates the normal wild-type DNA
chromosome number in each of the two product cells. Meiotic sequence, and red with the letter x represents the altered
division results in half the original chromosome number in each sequence:
of the four product cells.
Wild-type allele A
Mutant allele a9 x
Mutant allele a0 x
MENDELIAN INHERITANCE
PATTERNS Molecular aspects of gene transmission
Replication of alleles during the S phase What hap-
LO 2.3 Propose reasonable hypotheses to explain
dominance and recessiveness of specific alleles pens to alleles at the molecular level during cell division?
at the molecular level. We know that the primary genomic component of each
chromosome is a DNA molecule. This DNA molecule is
Of course, Mendel had no idea of the molecular nature of replicated during the S phase, which precedes both mito-
the concepts he was working with. In this section, we can sis and meiosis. As we will see in Chapter 7, replication is
begin putting some of Mendel’s concepts into a molecular an accurate process, and so all the genetic information is
context. Let’s begin with alleles. We have used the concept duplicated, whether wild type or mutant. For example, if
of alleles without defining them at the molecular level. a mutation is the result of a change in a single nucleotide
What are the structural differences between wild-type and pair—say, from GC (wild type) to AT (mutant)—then in a
mutant alleles at the DNA level of a gene? What are the heterozygote, replication will be as follows:
functional differences at the protein level? Mutant alleles
chromatid GC
can be used to study single-gene inheritance without need- homolog GC → replication →
ing to understand their structural or functional nature. chromatid GC
However, because a primary reason for embarking on chromatid AT
homolog AT → replication →
single-gene inheritance is ultimately to investigate a gene’s chromatid AT
function, we must come to grips with the molecular nature
DNA replication before mitosis in a haploid and a dip-
of wild-type and mutant alleles at both the structural and
loid are shown in Figure 2-11. This type of illustration serves
the functional level.
to remind us that, in our considerations of the mechanisms
Structural differences between alleles of inheritance, it is essentially DNA molecules that are
being moved around in the dividing cells.
at the molecular level
Mendel proposed that genes come in different forms we Meiosis and mitosis at the molecular level The repli-
now call alleles. What are alleles at the molecular level? cation of DNA during the S phase produces two copies of
When alleles such as A and a are examined at the DNA each of the alleles A and a, that can now be segregated into
DNA molecules replicate to form identical chromatids separate cells. Nuclear division visualized at the DNA level
is shown in Figure 2-12.
Chromatid formation DNA replication
Demonstrating chromosome segregation at the
Homozygous diploid b+/b+ molecular level We have interpreted single-gene pheno-
b+ b+ G typic inheritance patterns in relation to the segregation of
b + G C
chromosomal DNA at meiosis. Is there any way to show
G DNA segregation directly (as opposed to phenotypic seg-
C
regation)? The most straightforward approach would be to
b+ C
sequence the alleles (say, A and a) in the parents and the
b+ G meiotic products: the result would be that one-half of the
G C products would have the A DNA sequence and one-half
b+ C G would have the a DNA sequence. The same would be true
b+ b+
for any DNA sequence that differed in the inherited chro-
C
mosomes, including regions of DNA found between genes
(i.e., not inside alleles correlated with known phenotypes
Heterozygous diploid b+/b such as red and white flowers). Thus, we see the rules of
b+ b+ G segregation enunciated by Mendel apply not only to genes,
b+ G C but to any stretch of DNA along a chromosome.
C G
KEY CONCEPT Mendelian inheritance is shown by any seg-
b+ C ment of DNA on a chromosome: by genes and their alleles
b A and by molecular markers not necessarily associated with any
A biological function.
T
b T A
b b T
Alleles at the molecular level
Homozygous diploid b/b At the molecular level, the primary phenotype of a gene is
b b A the protein it produces. What are the functional differences
b A between proteins that explain the different effects of wild-
T
type and mutant alleles on the traits of an organism?
T A
Let’s explore the topic by using the human disease
b T phenylketonuria (PKU). We shall see in a later section on
b A pedigree analysis that the PKU phenotype is inherited as a
A T Mendelian recessive. The disease is caused by a defective
b A allele of the gene that encodes the liver enzyme phenylal-
T
anine hydroxylase (PAH). This enzyme normally converts
b b T
phenylalanine in food into the amino acid tyrosine:
Haploid b+ phenylalanine
hydroxylase
b+ b+ G phenylalanine tyrosine
b+ G C
C G
b+ C
FIGURE 2-11 Each chromosome divides longitudinally into two
chromatids (left); at the molecular level (right), the single DNA molecule
of each chromosome replicates, producing two DNA molecules, one
Haploid b
for each chromatid (orange indicates the newly synthesized strand).
b b A
Also shown are various combinations of a gene with wild-type allele
b A T b+ and mutant form b, caused by the change in a single base pair
A from GC to AT. Notice that, at the DNA level, the two chromatids
T
produced when a chromosome replicates are always identical with
b T each other and with the original chromosome. Also, note that after
DNA replication, but before anaphase of mitosis, each pair of sister
chromatids is joined at the centromere (not shown).
A
a
A a
A
A
a A A
a A
A a A
Mitosis Mitosis Meiosis
a
Alignment Alignment Pairing of
on equator on equator homologs
at equator A
A A a (tetrad)
A A a
a
A A a
A End of
first division
a A a
and and
A A
A a
a
End of
and second division and
1 1
A Sex cells a
2 2
FIGURE 2-12 DNA and gene transmission in mitosis and meiosis in eukaryotes. The S phase and
the main stages of mitosis and meiosis are shown. Mitotic divisions (left and middle) conserve the
genotype of the original cell. At the right, the two successive meiotic divisions that take place during
the sexual stage of the life cycle have the net effect of halving the number of chromosomes. The
alleles A and a of one gene are used to show how genotypes are transmitted in cell division.
However, a mutation in the gene encoding this enzyme KEY CONCEPT Most mutations that alter phenotype alter the
may alter the amino acid sequence in the vicinity of the amino acid sequence of the gene’s protein product, resulting in
enzyme’s active site. In this case, the enzyme cannot bind reduced or absent function.
phenylalanine (its substrate) or convert it into tyrosine.
Therefore, phenylalanine builds up in the body and is con-
verted instead into phenylpyruvic acid. This compound Some of the general consequences of mutation at the
interferes with the development of the nervous system, protein level are shown in Figure 2-14. Many mutant alleles
leading to intellectual disabilities. are of a type generally called null alleles: the proteins
Babies are now routinely tested for this processing defi- encoded by them completely lack function. Other mutant
ciency at birth. If the deficiency is detected, phenylalanine alleles reduce the level of enzyme function; they are some-
can be withheld with the use of a special diet and the devel- times called leaky mutations, because some wild-type
opment of the disease arrested. function seems to “leak” into the mutant phenotype. DNA
The functional PAH enzyme is a single protein, encoded sequencing often detects changes within a gene that have
by a single gene. What changes have occurred in the DNA no functional impact at all, so these alleles, although they
of the mutant form of the PKU gene, and how can such have silent mutations, are functionally wild type. Hence,
change at the DNA level affect protein function and pro- we see that the terms wild type and mutant sometimes have
duce the disease phenotype? Sequencing of the mutant to be used carefully.
alleles from many PKU patients has revealed a plethora We have been pursuing the idea that finding a set of
of mutations at different sites along the gene; the results genes that impinge on the biological trait under investiga-
are summarized in Figure 2-13. They represent a range of tion is an important goal of genetics, because it defines the
DNA changes, but most are small changes affecting only components of the system. However, finding the precise
one nucleotide pair among the thousands that constitute way in which mutant alleles lead to mutant phenotypes is
the gene. What all these alleles have in common is that they often challenging, requiring not only the identification of
result in a defective protein that no longer has normal PAH the protein products of these genes, but also detailed cellu-
activity. Most of the mutant alleles contain mutations in the lar and physiological studies to measure the effects of the
regions of the PKU gene that encode the amino acids that mutations. Furthermore, finding how the set of genes inter-
make up the PAH enzyme. The protein-coding regions of acts is a second level of challenge and a topic that we will
a gene are called exons. By changing one or more amino pursue later, starting in Chapter 5.
acids, mutations within exons inactivate some essential part
of the protein encoded by the gene. The effect of the muta- Dominance and recessiveness With an understand-
tion on the function of the gene depends on where within ing of how genes function through their protein products,
the gene the mutation occurs. An important functional we can now better understand dominance and recessive-
region of the gene is that encoding an enzyme’s active site; ness. Dominance was defined earlier in this chapter as the
so this region is very sensitive to mutation. In addition, a phenotype shown by a heterozygote. Formally, it is the
minority of mutations are found to be in noncoding regions phenotype that is dominant or recessive; but, in practice,
of the gene in between the exons. These areas are known geneticists more often apply the term to alleles. This for-
as introns, and mutations within introns often prevent the mal definition has no molecular content, but both domi-
normal processing of the primary RNA transcript. (Exons, nance and recessiveness can have simple explanations at the
introns, and RNA processing will be explored further in molecular level. We introduce the topic here, to be revisited
Chapter 8.) in Chapter 5.
24 5 7 21
Exon 4 11 10 4 7 37 12 9 1
mutations
1 2 3 4 5 6 7 8 9 10 11 12 13
Intron
mutations 1 4 2 2 1 3 1 41 1
FIGURE 2-13 Many mutations of the human phenylalanine hydroxylase gene that cause enzyme
malfunction are known. The number of mutations in the exons, or protein-encoding regions (black),
are listed above the gene. The number of mutations in the intron regions (green, numbered 1 through
13) that alter RNA processing are listed below the gene. [Data from C. R. Scriver, Ann. Rev. Genet. 28,
1994, 141–165.]
m2: null
m3: null
m4: leaky
m5: silent
m6: null
m2 Active site
Protein m4
m3 m5
= mutant site
How can alleles be dominant? How can they be reces- heterozygote, the wild-type allele cannot mask this new
sive? Recessiveness is observed in null mutations in genes function.
that are functionally haplosufficient, loosely meaning that From the above brief considerations, we see that pheno-
one gene copy has enough function to produce a wild-type type, the description or measurement that we track during
phenotype. Although a wild-type diploid cell normally has Mendelian inheritance, is an emergent property based on
two fully functional copies of a gene, one copy of a haplo- the nature of alleles and the way in which the gene func-
sufficient gene provides enough gene product (generally a tions normally and abnormally. The same can be said for
protein) to carry out the normal transactions of the cell. In the descriptions “dominant” and “recessive” that we apply
a heterozygote (say, + /m , where m is a null allele), the single to a phenotype.
functional copy encoded by the + allele provides enough
protein product for normal cellular function. In a simple KEY CONCEPT As a general rule, a null mutation is recessive
example, assume a cell needs a minimum of 10 protein units in a haplosufficient gene, and a null mutation is dominant in a
to function normally. Each wild-type allele can produce 12 haploinsufficient gene.
units. Hence, a homozygous wild type + / + will produce
24 units. The heterozygote + /m will produce 12 units, in
excess of the 10-unit minimum, and hence the mutant allele
is recessive as it has no impact in the heterozygote.
Other genes are haploinsufficient. In such cases, a null
2.4 SOME GENES DISCOVERED
mutant allele will be dominant because, in a heterozygote BY OBSERVING SEGREGATION
(+ /P), the single wild-type allele cannot provide enough RATIOS
product for normal function. As another example, let’s
assume the cell needs a minimum of 20 units of this protein, LO 2.4 Predict phenotypic ratios among descendants from
and the wild-type allele produces only 12 units. A homozy- crosses of parents differing at a single gene.
gous wild type + / + makes 24 units, which is over the min-
imum. However, a heterozygote involving a null mutation Recall that one general aim of genetic analysis today is
(+ /P) produces only 12; hence, the presence of the mutant to dissect a biological trait by discovering the set of sin-
allele in the heterozygote results in an inadequate supply of gle genes that affect it. We learned that an important way
product, and a mutant phenotype ensues. In this situation, to identify these genes is by the phenotypic segregation
the mutant allele P is dominant. ratios generated by their mutations—most often 1:1 and
In some cases, mutation results in a new function for 3:1 ratios, both of which are based on equal segregation as
the gene. Such mutations can be dominant because, in a defined by Gregor Mendel.
A gene active in the development KEY CONCEPT The Punnett square is a graphical represen-
of flower color tation of parental gametes and shows how they randomly unite
To begin the process, the white-flowered plant is crossed to produce progeny genotypes, from which phenotypic ratios
with the normal wild-type red. All the F1 plants are red of the progeny can be deduced.
flowered, and, of 500 F2 plants sampled, 378 are red flow-
ered and 122 are white flowered. If we acknowledge the
existence of sampling error, these F2 numbers are very
close to a 43 : 41 , or 3:1, ratio. Because this ratio indicates A gene for wing development
single-gene inheritance, we can conclude that the mutant In the fruit-fly example, the cross of the mutant short-
Introduction to Genetic Analysis, 11e
is caused by a recessive alteration in a single gene. Accord- winged
Figurefly with #271
02UN09 wild-type long-winged stock yielded 788
ing to the general rules of gene nomenclature, the mutant progeny, classified
04/20/14 as follows:
allele for white petals might be called alb for albino and the Dragonfly Media Group
wild-type allele would be alb+ or just +. (The conventions 196 short-winged males
for allele nomenclature vary somewhat among organisms: 194 short-winged females
some of the variations are shown in Appendix A on nomen- 197 long-winged males
clature.) We surmise that the wild-type allele plays an
201 long-winged females
essential role in producing the colored petals of the plant,
a property that is almost certainly necessary for attracting In total, there are 390 short- and 398 long-winged prog-
pollinators to the flower. The gene might be implicated in eny, very close to a 1:1 ratio. The ratio is the same within
the biochemical synthesis of the pigment or in the part of males and females, again within the bounds of sampling
the signaling system that tells the cells of the flower to start error. Hence, from these results, the “short wings” mutant
making pigment or in a number of other possibilities that was very likely produced by a dominant mutation. Note
require further investigation. At the purely genetic level, the that, for a dominant mutation to be expressed, only a sin-
crosses made would be represented symbolically as gle “dose” of mutant allele is necessary; so, in most cases,
when the mutant first shows up in the population, it will
P +/+× alb /alb
be in the heterozygous state. (This is not true for a reces-
F1 all +/alb sive mutation such as that in the preceding plant example,
1 which must be homozygous to be expressed and must have
F2 4
+/+
1
+/alb come from the selfing of an unidentified heterozygous plant
2
1 in the preceding generation.)
alb /alb
4 When long-winged progeny were interbred, all of their
or graphically as in the grids in the next column (see also progeny were long winged, as expected of a recessive wild-
Figure 2-5). This type of grid showing gametes and gametic type allele. When the short-winged progeny were interbred,
fusions is called a Punnett square, named after an early their progeny showed a ratio of three-fourths short to one-
geneticist, Reginald C. Punnett. They are useful devices for fourth long.
explaining genetic ratios, and we shall encounter more in Dominant mutations are represented by uppercase let-
later discussions. ters or words: in the present example, the mutant allele
might be named SH, standing for “short.” Then the crosses A gene for hyphal branching
would be represented symbolically as
A hyperbranching fungal mutant (such as the button-like
P +/+ × SH /+ colony in Figure 2-1) was crossed with a wild-type fungus
F1 1
+/+ with normal sparse branching. In a sample of 300 progeny,
2
1 152 were wild type and 148 were hyperbranching, very
SH/+
2 close to a 1:1 ratio. We infer from this single-gene inher-
F1 +/+ × +/+ itance ratio that the hyperbranching mutation is of a sin-
gle gene. In haploids, assigning dominance is usually not
all +/+
possible, but, for convenience, we can call the hyperbranch-
F1 SH /+ × SH /+ ing allele hb and the wild type hb+ or +. The cross must
1 have been
4
SH /SH
1
SH/+ P +× hb
2
1
+/+ Diploid meiocyte +/hb
4
1
F1 2
+
or graphically as shown in the grids below.
1
This analysis of the fly mutant identifies a gene that is 2
hb
part of a subset of genes that, in wild-type form, are crucial The mutation and inheritance analysis has uncovered a
for the normal development of a wing. Such a result is the gene whose wild-type allele is essential for normal control
starting point of further studies that would focus on the pre- of branching, a key function in fungal dispersal and nutri-
cise developmental and cellular ways in which the growth of ent acquisition. Now the mutant needs to be investigated to
the wing is arrested, which, once identified, reveal the time of see the location in the normal developmental sequence at
action of the wild-type allele in the course of development. which the mutant produces a block. This information will
reveal the time and place in the cells at which the normal
P SH allele acts.
Sometimes, the severity of a mutant phenotype renders
the organism sterile, unable to go through the sexual cycle.
/ SH /
How can the single-gene inheritance of sterile mutants be
demonstrated? In a diploid organism, a sterile recessive
mutant can be propagated as a heterozygote, and then
/ SH /
the heterozygote can be selfed to produce the expected 25
percent homozygous recessive mutants for study. A sterile
dominant mutant is a genetic dead end and cannot be prop-
agated sexually, but, in plants and fungi, such a mutant can
F1 be easily propagated asexually.
What if a cross between a mutant and a wild type does
not produce a 3:1 or a 1:1 ratio as discussed here, but some
/ /
other ratio? Such a result can be due to the interactions of
several genes or to an environmental effect. Some of these
/ /
possibilities are discussed in Chapter 5, and environmental
effects on phenotype are also considered in Solved Problem
1 at the end of this chapter.
However, the same principle of inheritance (essentially, seen in progeny are often different from the a utosomal
Mendel’s law of equal segregation) can also be used to pre- ratios.
dict phenotypic ratios in the progeny of parents of known
genotypes. These parents would be from stocks maintained Sex chromosomes
by the researcher. The types and proportions of the prog-
Most animals and many plants show sexual dimorphism; in
eny of crosses such as A /A × A /a, A /A × a /a, A /a × A /a, and
other words, individuals are either male or female. In most
A /a × a /a can be easily predicted. In summary,
of these cases, sex is determined by a special pair of sex
Cross parents of known genotypes → chromosomes. Let’s look at humans as an example. Human
Predict phenotypic ratios in progeny body cells have 46 chromosomes: 22 homologous pairs of
autosomes plus 2 sex chromosomes. Females have a pair of
This type of analysis is used in general breeding to
identical sex chromosomes called the X chromosomes. Males
synthesize genotypes for research or for agriculture. It is
have a nonidentical pair, consisting of one X and one Y. The
also useful in predicting likelihoods of various outcomes
Y chromosome is considerably shorter than the X. Hence, if
in human matings in families with histories of single-gene
we let A represent autosomal chromosomes, we can write
diseases.
After single-gene inheritance has been established, females = 44A + XX
an individual showing the dominant phenotype but of males = 44A + XY
unknown genotype can be tested to see if the genotype is
homozygous or heterozygous. Such a test can be performed At meiosis in females, the two X chromosomes pair and
by crossing the individual (of phenotype A/?) with a reces- segregate like autosomes, and so each egg receives one X
sive tester strain a /a . If the individual is heterozygous, a 1:1 chromosome. Hence, with regard to sex chromosomes, the
ratio will result ( 12 A /a and 12 a /a); if the individual is homo- gametes are of only one type and the female is said to be
zygous, all progeny will show the dominant phenotype (all the homogametic sex. At meiosis in males, the X and the
A /a ). In general, the cross of an individual of unknown het- Y chromosomes pair over a short region, which ensures
erozygosity (for one gene or more) with a fully recessive that the X and Y segregate so that there are two types of
parent is called a testcross, and the recessive individual is sperm, half with a single X and the other half with a single
called a tester. We will encounter testcrosses many times Y. Therefore, the male is called the heterogametic sex.
throughout subsequent chapters; they are very useful in
KEY CONCEPT Human sex chromosomes, X and Y, contain
deducing the meiotic events taking place in more complex
different sets of genes. Females are the homogametic sex,
genotypes. The use of a fully recessive tester means that
with a pair of X chromosomes (XX). Males are the heteroga-
meiosis in the tester parent can be ignored because all of its metic sex, with a nonidentical pair of sex chromosomes (XY).
gametes are recessive and do not contribute to the pheno-
types of the progeny. An alternative test for heterozygosity
(useful if a recessive tester is not available and the organism The inheritance patterns of genes on the sex chro-
can be selfed) is simply to self the unknown: if the organism mosomes are different from those of autosomal genes.
being tested is heterozygous, a 3:1 ratio will be found in Sex-chromosome inheritance patterns were first investi-
the progeny. Such tests are useful and common in routine gated in the early 1900s in the laboratory of the great genet-
genetic analysis. icist Thomas Hunt Morgan, using the fruit fly Drosophila
melanogaster (see the Model Organism box on page 50).
KEY CONCEPT The principles of inheritance (such as the This insect has been one of the most important research
law of equal segregation) can be applied in two directions: organisms in genetics; its short, simple life cycle contrib-
(1) inferring genotypes from phenotypic ratios and (2) predict- utes to its usefulness in this regard. Fruit flies have three
ing phenotypic ratios from parents of known genotypes.
pairs of autosomes plus a pair of sex chromosomes, again
referred to as X and Y. As in mammals, Drosophila females
have the constitution XX and males are XY. However, the
2.5 SEX-LINKED SINGLE-GENE mechanism of sex determination in Drosophila differs from
INHERITANCE PATTERNS that in mammals. In Drosophila, the number of X chromo-
somes in relation to the autosomes determines sex: two X’s
result in a female, and one X results in a male. In mammals,
LO 2.5 In the progeny of controlled crosses, recognize
phenotypic ratios diagnostic of X-linked single-gene the presence of the Y chromosome determines maleness
inheritance. and the absence of a Y determines femaleness. However,
it is important to note that, despite this somewhat differ-
The chromosomes that we have analyzed so far are auto- ent basis for sex determination, the single-gene inheritance
somes, the “regular” chromosomes that form most of the patterns of genes on the sex chromosomes are remarkably
genomic set. However, many animals and plants have a spe- similar in Drosophila and mammals.
cial pair of chromosomes associated with sex. The sex chro- Vascular plants show a variety of sexual arrangements.
mosomes also segregate equally, but the phenotypic ratios Dioecious species are those showing animal-like sexual
White-eyed and red-eyed Drosophila the sexes, is explained in Figure 2-18. The basis of the inher-
itance pattern is that all the F1 flies receive a wild-type
allele from their mothers, but the F1 females also receive a
white-eye allele from their fathers. Hence, all F1 females are
heterozygous wild type (w+ /w) , and the F1 males are hemi-
zygous wild type (w+ ). The F1 females pass on the white-eye
allele to half their sons, who express it, and to half their
daughters, who do not express it, because they must inherit
the wild-type allele from their fathers.
FIGURE 2-17 The red-eyed fly is wild type, and the white-eyed fly is
a mutant. [Science Source/Getty Images.] The reciprocal cross gives a different result; that is,
the cross between white-eyed females and red-eyed males
is w for white eyes (the lowercase letter indicates that the gives an F1 in which all the females are red eyed but all the
allele is recessive), and the corresponding wild-type allele males are white eyed. In this case, every female inherited
is w+ . When white-eyed males are crossed with red-eyed the dominant w+ allele from the father’s X chromosome,
females, all the F1 progeny have red eyes, suggesting that whereas every male inherited the recessive w allele from its
the allele for white eyes is recessive. Crossing these red-eyed mother. The F2 consists of one-half red-eyed and one-half
F1 males and females produces a 3:1 F2 ratio of red-eyed to white-eyed flies of both sexes. Hence, in sex linkage, we see
white-eyed flies, but all the white-eyed flies are males. This examples not only of different ratios in different sexes, but
inheritance pattern, which shows a clear difference between also of differences between reciprocal crosses.
1 day
Sex is determined by X and Y sex chromosomes (XX = female,
XY = male), and males and females are easily distinguished. Pupa
Mutant phenotypes regularly arise in lab populations, and
First instar
their frequency can be increased by treatment with mutagenic
radiation or chemicals. It is a diploid organism, with four pairs
of homologous chromosomes (2n = 8). In salivary glands 1 day
and certain other tissues, multiple rounds of DNA replication 1
2 2 –3 days
without chromosomal division result in “giant chromosomes,”
each with a unique banding pattern that provides geneticists Second instar
with landmarks for the study of chromosome mapping and Third instar
rearrangement. It is also noteworthy that there are many spe-
cies and local races of Drosophila, which have been important 1 day
raw material for the study of evolution.
X X X Y X X X Y
w w+
1 1 1 1
2 2 2 2
w+ w+ w w+ w w+ w w
Female Female
gametes gametes
1 1 1 1
2 Red female 2 Red male 2 Red female 2 White male
w+ w
1 1 1 1
2 2 2 2
w+ w+ w+ w+ w+ w+ w w+
1 1
2 2
1 1 1 1
4 Red female 4 Red male 4 Red female 4 Red male
Female Female
gametes gametes
w w+ w w w w w w
1 1
2 2
1 1 1 1
4 Red female 4 White male 4 White female 4 White male
FIGURE 2-18 Reciprocal crosses between red-eyed (red) and white-eyed (white) Drosophila give
different results. The alleles are X linked, and the inheritance of the X chromosome explains the A N I M ATED A RT
phenotypic ratios observed, which are different from those of autosomal genes. (In Drosophila and X-linked inheritance in flies
many other experimental systems, a superscript plus sign is used to designate the normal, or wild-
type, allele. Here, w + encodes red eyes and w encodes white eyes.)
Note that Drosophila eye color has nothing to do humans, and so geneticists must resort to scrutinizing medical
with sex determination, and so we have an illustration of records in the hope that informative matings have been made
the principle that genes on the sex chromosomes are not (such as monohybrid crosses) that could be used to infer sin-
necessarily related to sexual function. The same is true in gle-gene inheritance. Such a scrutiny of records of matings
humans: in the discussion of pedigree analysis later in this is called pedigree analysis. A member of a family who first
chapter, we shall see many X-linked genes, yet few could be comes to the attention of a geneticist is called the proposi-
construed as being connected to sexual function. tus. Usually, the phenotype of the propositus is exceptional
The abnormal allele associated with white eye color in in some way; for example, the propositus might have some
Drosophila is recessive, but abnormal alleles of genes on type of medical disorder. The investigator then traces the his-
the X chromosome that are dominant also arise, such as tory of the phenotype through the history of the family and
the Drosophila mutant hairy wing (Hw). In such cases, the draws a family tree, or pedigree, by using the standard sym-
wild-type allele (Hw+ ) is recessive. The dominant abnormal bols given in Figure 2-19. This was exactly the approach taken
alleles show the inheritance pattern corresponding to that by researchers in the case of Louise Benge, who we learned
of the wild-type allele for red eyes in the preceding exam- about in Chapter 1, that ultimately led to the discovery of the
ple. The ratios obtained are the same. gene allele at the cause of her disease (see the pedigree analy-
sis of Benge’s family tree in Figure 1-15).
KEY CONCEPT Sex-linked inheritance is recognized by dif- To see single-gene inheritance, the patterns in the ped-
ferent phenotypic ratios in the two sexes of progeny, as well as igree have to be interpreted according to Mendel’s law of
different ratios in reciprocal crosses. equal segregation, but humans usually have few children
and so, because of this small progeny sample size, the
Historically, in the early decades of the twentieth cen- expected 3:1 and 1:1 ratios are usually not seen unless
tury, the demonstration by Morgan of X-linked inheritance many similar pedigrees are combined. The approach to
of white eyes in Drosophila was a key piece of evidence that pedigree analysis also depends on whether one of the con-
suggested that genes are indeed located on chromosomes, trasting phenotypes is a rare disorder or both phenotypes
because an inheritance pattern was correlated with one of a pair are common (in which case they are said to be
specific chromosome pair. The idea became known as “the
chromosome theory of inheritance.” At that period in history,
it had recently been shown that, in many organisms, sex is Pedigree symbols
determined by an X and a Y chromosome and that, in males,
Male
these chromosomes segregate equally at meiosis to regenerate 2 3 Number of children
of sex indicated
equal numbers of males and females in the next generation.
Female
Morgan recognized that the inheritance of alleles of the eye-
Affected individuals
color gene is exactly parallel to the inheritance of X chromo-
Mating
somes at meiosis; hence, the gene was likely to be on the X
Heterozygotes for
chromosome. The inheritance of white eyes was extended to autosomal recessive
Drosophila lines that had abnormal numbers of sex chromo- Parents and
children:
somes. With the use of this novel situation, it was still possi- 1 boy; 1 girl Carrier of sex-linked
ble to predict gene-inheritance patterns from the segregation (in order of birth) recessive
of the abnormal chromosomes. That these predictions proved
correct was a convincing test of the chromosome theory. Death
Other genetic analyses revealed that, in chickens and
moths, sex-linked inheritance could be explained only if the
female was the heterogametic sex. In these organisms, the Abortion or stillbirth
Dizygotic (sex unspecified)
female sex chromosomes were designated ZW and males (nonidentical twins)
were designated ZZ.
Propositus
“morphs” of a polymorphism). Most pedigrees are drawn Notice that, even though Mendelian rules are at work,
for medical reasons and therefore concern medical disor- Mendelian ratios are not necessarily observed in single fam-
ders that are almost by definition rare. In this case, we have ilies because of small sample size, as predicted earlier. In
two phenotypes: the presence and the absence of the disor- the preceding example, we observe a 1:1 phenotypic ratio
der. Four patterns of single-gene inheritance are revealed in in the progeny of a monohybrid cross. If the couple were
pedigrees. Let’s look, first, at recessive disorders caused by to have, say, 20 children, the ratio would be something like
recessive alleles of single autosomal genes. 15 unaffected children and 5 with PKU (a 3:1 ratio), but, in
a small sample of 4 children, any ratio is possible, and all
Autosomal recessive disorders ratios are commonly found.
The family pedigrees of autosomal recessive disorders
The affected phenotype of an autosomal recessive disorder tend to look rather bare, with few black symbols. A reces-
is inherited as a recessive allele; hence, the corresponding sive condition shows up in groups of affected siblings, and
unaffected phenotype must be inherited as the correspond- the people in earlier and later generations tend not to be
ing dominant allele. For example, the human disease phe- affected. To understand why this is so, it is important to
nylketonuria (PKU), discussed earlier, is inherited in a have some understanding of the genetic structure of popu-
simple Mendelian manner as a recessive phenotype, with lations underlying such rare conditions. By definition, if the
PKU determined by the allele p and the normal condition condition is rare, most people do not carry the abnormal
determined by P. Therefore, people with this disease are of allele. Furthermore, most of those people who do carry the
genotype p /p, and people who do not have the disease are abnormal allele are heterozygous for it rather than homo-
either P /P or P /p. Recall that the term wild type and its zygous. The basic reason why heterozygotes are much more
allele symbols are not used in human genetics because wild common than recessive homozygotes is that to be a reces-
type is impossible to define. sive homozygote, both parents must have the a allele, but to
What patterns in a pedigree would reveal autosomal be a heterozygote, only one parent must have it.
recessive inheritance? The two key points are that (1) gen- The birth of an affected person usually depends on
erally the disorder appears in the progeny of unaffected the rare chance union of unrelated heterozygous parents.
parents, and (2) the affected progeny include both males However, inbreeding (mating between relatives, some-
and females. When we know that both male and female times referred to as consanguinity in humans) increases the
progeny are affected, we can infer that we are most likely chance that two heterozygotes will mate. An example of
dealing with simple Mendelian inheritance of a gene on an mating between cousins is shown in Figure 2-20. Individuals
autosome, rather than a gene on a sex chromosome. The
following typical pedigree illustrates the key point that
Homozygous recessives from inbreeding
affected children are born to unaffected parents:
I
1 2
o n e A/a; on e A/A
II
From this pattern, we can deduce a simple monohybrid 1 2 3 4 5
cross, with the recessive allele responsible for the excep- A/A A/a A /– A /a A/A
tional phenotype (indicated in black). Both parents must be
heterozygotes—say, A /a; both must have an a allele because
each contributed an a allele to each affected child, and both III
must have an A allele because they are phenotypically nor- 1 2 3 4 5 6 7
mal. We can identify the genotypes of the children (shown A /– A /– A /– A /– A /a A /a A/–
left to right) as A /−, a /a, a /a, and A /−. The pedigree can be
rewritten as follows: IV
A/a A/a 1 2 3 4 5
A /– a /a A/– a /a A/–
Early-onset
Parkinson’s disease
(PARK7),
Male infertility (USP9Y), autosomal Ehlers-Danlos syndrome
Y-linked. recessive. type IV (COL3A1),
Defect of sperm cells. Neurodegeneration. autosomal dominant.
Stretchy collagen.
Hemophilia (F8), X-linked recessive.
Inactive blood clotting factor.
Alkaptonuria (HGD),
autosomal recessive. Black urine.
XY 1 2
autosomal recessive.
Metabolic disorder.
22 3
21 Chromosome 4
Pseudoachondroplasia (COMP), 20 pairs 5 Cystic fibrosis (CFTR),
autosomal recessive.
autosomal dominant. A type of dwarfism.
Abnormal chlorine and
19 6 sodium transport;
mucus in the lungs
17 8
Hereditary hemorrhagic telangiectasia
(MADH4), autosomal dominant. 16 9
Dilation of capillaries causing bleeding.
15 10
14 11
13 12
Werner syndrome
Canavan disease (ASPA), (WRN), autosomal
autosomal recessive. recessive. Premature aging.
Damage to nerve cells
and brain.
FIGURE 2-21 The positions of the genes mutated in some single-gene diseases, shown in the
23 pairs of chromosomes in a human being. Each chromosome has a characteristic banding pattern.
X and Y are the sex chromosomes (XX in women and XY in men). Genes associated with each
disease are shown in parentheses.
III-5 and III-6 are first cousins and produce two homo- matings between nonrelatives. For this reason, first-cousin
zygotes for the rare allele. You can see from Figure 2-20 matings contribute a large proportion of people with reces-
Introduction to Genetic
that an ancestor who Analysis, 11e
is a heterozygote may produce many sive diseases in the population.
Figure 02.20 #205
descendants who also are heterozygotes. Hence, two cous- Some other examples of human recessive disorders are
04/20/14
ins can carry the same rare recessive allele inherited from a
05/12/14 shown in Figure 2-21. Cystic fibrosis is a disease inherited
05/13/14
common ancestor. For two unrelated persons to be hetero- on chromosome 7 according to Mendelian rules as an auto-
Dragonfly Mediawould
zygous, they Group have to inherit the rare allele from both somal recessive phenotype. Its most important symptom
their families. Thus, matings between relatives generally is the secretion of large amounts of mucus into the lungs,
run a higher risk of producing recessive disorders than do resulting in death from a combination of effects but usually
precipitated by infection of the respiratory tract. The mucus A mutant gene causes albinism
can be dislodged by mechanical chest thumpers, and pul-
monary infection can be prevented by antibiotics; thus,
with treatment, cystic fibrosis patients can live to adult-
hood. The cystic fibrosis gene (a mutant allele of the wild-
type gene) was one of the first human disease genes to be
isolated at the DNA level, in 1989. This line of research
eventually revealed that the disorder is caused by a defec-
tive protein that normally transports chloride ions across
the cell membrane. The resultant alteration of the salt bal-
ance changes the constitution of the lung mucus. This new
understanding of gene function in affected and unaffected
persons has given hope for more effective treatment.
Human albinism also is inherited in the standard auto-
somal recessive manner. The mutant allele is of a gene that
normally synthesizes the brown or black pigment mela-
nin, normally found in skin, hair, and the retina of the eye FIGURE 2-22 A nonfunctional version of a skin-pigment gene results
(Figure 2-22). in lack of pigment. In this case, both members of the gene pair are
mutated. [Friedrich Stark/Alamy.]
KEY CONCEPT In human pedigrees, an autosomal recessive
disorder is generally revealed by the appearance of the disor- be in principle D /d or D /D. However, the two “doses” of
der in the male and female progeny of unaffected parents. the D allele in the D /D genotype are believed to produce
such a severe effect that this genotype is lethal. If this belief
is generally true, all dwarf individuals are heterozygotes.
Autosomal dominant disorders In pedigree analysis, the main clues for identifying an
What pedigree patterns are expected from autosomal dom- autosomal dominant disorder with Mendelian inheritance
inant disorders? Here, the normal allele is recessive, and are that the phenotype tends to appear in every generation
the defective allele is dominant. It may seem paradoxical of the pedigree and that affected fathers or mothers trans-
that a rare disorder can be dominant, but remember that mit the phenotype to both sons and daughters. Again, the
dominance and recessiveness are simply properties of how equal representation of both sexes among the affected off-
alleles act in heterozygotes and are not defined in reference spring rules out inheritance through the sex chromosomes.
to how common they are in the population. A good exam- The phenotype appears in every generation because, gen-
ple of a rare dominant phenotype that shows single-gene erally, the abnormal allele carried by a person must have
inheritance is pseudoachondroplasia, a type of dwarfism come from a parent in the preceding generation. (Abnormal
(Figure 2-23). In regard to this gene, people with normal stat- alleles can also arise de novo by mutation. This possibil-
ure are genotypically d /d , and the dwarf phenotype could ity must be kept in mind for disorders that interfere with
Inheritance of an autosomal dominant disorder The phenotype is one of neural degeneration, leading to con-
vulsions and premature death. Folk singer Woody Guthrie suf-
I fered from Huntington disease. The disease is rather unusual
1 2 in that it shows late onset, the symptoms generally not
A /a a /a appearing until after the person has reached reproductive age.
When the disease has been diagnosed in a parent, each child
already born knows that he or she has a 50 percent chance
II of inheriting the allele and the associated disease. This tragic
1 2 3 4 5 6 7 pattern has inspired a great effort to find ways of identifying
a/a a /a a /a A /a a /a A /a a /a people who carry the abnormal allele before they experience
the onset of the disease. Now there are molecular diagnostics
for identifying people who carry the Huntington allele.
III Some other rare dominant conditions are polydactyly
1 2 3 4 5 6 7 8 9 10 11 12 13 (extra digits), shown in Figure 2-25, and piebald spotting,
a/a a/a a /a a /a A /a a /a A /a a /a a /a a /a A /a a /a A /a
shown in Figure 2-26.
FIGURE 2-24 Pedigree of a dominant phenotype determined by
a dominant allele A. In this pedigree, all the genotypes have been KEY CONCEPT Pedigrees of Mendelian autosomal dominant
deduced. disorders show affected males and females in each generation;
they also show affected men and women transmitting the con-
dition to equal proportions of their sons and daughters.
reproduction because, here, the condition is unlikely to have
been inherited from an affected parent.) A typical pedigree
for a dominant disorder is shown in Figure 2-24. Once again,
notice that Mendelian ratios are not necessarily observed in Autosomal polymorphisms
families. As with recessive disorders, persons bearing one Most natural populations also show polymorphisms,
copy of the rare A allele (A /a) are much more common than defined as the coexistence of two or more reasonably
those bearing two copies (A /A); so most affected people are common phenotypes of a biological property, such as the
heterozygotes, and virtually all matings that produce prog- occurrence of both red- and orange-fruited plants in a pop-
eny with dominant disorders are A /a × a /a. Therefore, if the ulation of wild raspberries. The alternative phenotypes of a
progeny of such matings are totaled, a 1:1 ratio is expected polymorphism (the morphs) are often inherited as alleles of
of unaffected (a /a) to affected (A /a) persons. a single autosomal gene in the standard Mendelian manner.
Huntington disease is an example of a disease inherited as Among the many human examples are the following dimor-
a dominant phenotype determined by an allele of a single gene. phisms (with two morphs, the simplest polymorphisms):
Polydactyly
(a) (b)
I
II
5,5
6,6
III
5,5 5,5 5,5 6,6 6,6
5,5 5,5 6,6 5,5 5,5
IV
6 unaffected 7 unaffected
3 affected
5,5 5,5 5,5 5,5 5,6
6,6 6,6 6,6 6,6 6,7
V 12 unaffected
6,6
6,6
FIGURE 2-25 Polydactyly is a rare dominant phenotype of the human hands and feet. (a)
Polydactyly, characterized by extra fingers, toes, or both, is determined by an allele P. The numbers
in the pedigree (b) give the number of fingers in the upper lines and the number of toes in the lower.
(Note the variation in expression of the P allele.) [(a) Biophoto Associates/Science Source.]
(a)
(b) I
1 2
II 4 3
1– 4 5 6 7 8 9 10 11– 13
III
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
IV
1 2 3 4
FIGURE 2-26 Piebald spotting is a rare dominant human phenotype. Although the phenotype
is encountered sporadically in all races, the patterns show up best in those with dark skin. (a) The
photographs show front and back views of affected persons IV-1, IV-3, III-5, III-8, and III-9 from
(b) the family pedigree. Notice the variation in expression of the piebald gene among family members.
The patterns are believed to be caused by the dominant allele interfering with the migration of
melanocytes (melanin-producing cells) from the dorsal to the ventral surface in the course of
development. The white forehead blaze is particularly characteristic and is often accompanied by a
white forelock in the hair.
Piebaldism is not a form of albinism; the cells in the light patches have the genetic potential to
make melanin, but, because they are not melanocytes, they are not developmentally programmed
to do so. In true albinism, the cells lack the potential to make melanin. (Piebaldism is caused by
mutations in c-kit, a type of gene called a proto-oncogene; see Chapter 15.) [Winship, K. Young, R.
Martell, R. Ramesar, D. Curtis, and P. Beighton, “Piebaldism: An Autonomous Autosomal Dominant Entity,”
Clinical Genetics 39, 1991, 330. © Reproduced with permission of John Wiley & Sons, Inc.]
brown versus blue eyes, pigmented versus blond hair, abil- it as a foul, bitter taste or—to the great surprise and disbe-
ity to smell freesia flowers versus inability, widow’s peak lief of tasters—cannot taste it at all. From the pedigree in
versus none, sticky versus dry earwax, and attached versus Figure 2-27, we can see that two tasters sometimes produce
free earlobes. In each example, the morph determined by nontaster children, which makes it clear that the allele that
the dominant allele is written first. confers the ability to taste is dominant and that the allele for
The interpretation of pedigrees for polymorphisms is nontasting is recessive. Notice in Figure 2-27 that almost all
somewhat different from that of rare disorders because, by people who enter into this family carry the recessive allele
definition, the morphs are common. Let’s look at a ped- either in heterozygous or in homozygous condition. Such a
igree for an interesting human case. Most human popu- pedigree thus differs from those of rare recessive disorders,
lations are dimorphic for the ability to taste the chemical for which the conventional assumption is that all outsiders
phenylthiocarbamide (PTC); that is, people can either detect who enter into a family are homozygous normal. Because
I I XAXA XaY
1 2 3 4 1 2
III
1 2 3 4
III
XaY XAY XAXa XAXA
1 2 3 4 5 6 7 8 9 10 11
XI
William Catherine Henry Meghan
(Kate) (Harry)
XII
George Charlotte Louis Archie
(b)
FIGURE 2-29 A pedigree for the X-linked recessive condition hemophilia in the royal families of
Europe. A recessive allele causing hemophilia (failure of blood clotting) arose through mutation in the
reproductive cells of Queen Victoria or one of her parents. This hemophilia allele spread into other
royal families by intermarriage. (a) This partial pedigree shows affected males and carrier females
(heterozygotes). Most spouses marrying into the families have been omitted from the pedigree for
simplicity. Can you deduce the likelihood of the present British royal family’s harboring the recessive
allele? (b) A painting showing Queen Victoria surrounded by her numerous descendants. [(b) Lebrecht
Music and Arts Photo Library/Alamy.]
ultimately from Queen Victoria, who was the grandmother Inheritance of an X-linked dominant disorder
of his mother, Alexandra. Hemophilia can now be treated
medically, but it was formerly a potentially fatal condition. I
It is interesting to note that the Jewish Talmud contains Xa/ Xa XA/ Y
rules about exemptions to male circumcision clearly show-
ing that the mode of transmission of the disease through
unaffected carrier females was well understood in ancient II
times. For example, one exemption was for the sons of Xa/ Y XA/ Xa XA/ Xa Xa/ Y Xa / Y
women whose sisters’ sons had bled profusely when they
were circumcised. Hence, abnormal bleeding was known
III
to be transmitted through the females of the family but
Xa/ Xa XA/ Xa Xa/ Y XA/ Y
expressed only in their male children.
A rare X-linked recessive phenotype that is interesting FIGURE 2-31 All the daughters of a male expressing an X-linked
from the point of view of sexual differentiation is a con- dominant phenotype will show the phenotype. Females heterozygous
dition called androgen insensitivity syndrome (previously for an X-linked dominant allele will pass the condition on to half their
referred to as testicular feminization syndrome), which has sons and daughters.
a frequency of about 1 in 65,000 male births. People with
this syndrome are chromosomally males, having 44 auto-
somes plus an X and a Y chromosome, but they develop as 2. Affected heterozygous females mating with unaf-
females (Figure 2-30). They have female external genitalia, fected males pass the condition to half their sons and
a blind vagina, and no uterus. Testes may be present either daughters.
in the labia or in the abdomen. Afflicted persons are sterile. This mode of inheritance is not common. One example
The condition is not reversed by treatment with the male is hypophosphatemia, a type of vitamin D–resistant rickets.
hormone androgen. The reason for the insensitivity is that Some forms of hypertrichosis (excess body and facial hair)
a mutation in the androgen-receptor gene causes the recep- show X-linked dominant inheritance.
tor to malfunction, and so the male hormone can have no
effect on the target organs that contribute to maleness. In Y-linked inheritance
humans, femaleness results when the male-determining sys-
tem is not functional. Only males inherit genes in the differential region of the
human Y chromosome, with fathers transmitting the
genes to their sons. The gene that plays a primary role
X-linked dominant disorders
in maleness is the SRY gene, sometimes called the testis-
The inheritance patterns of X-linked dominant disorders determining factor. Genomic analysis has confirmed that,
have the following characteristics in pedigrees (Figure 2-31): indeed, the SRY gene is in the differential region of the Y
1. Affected males pass the condition to all their daughters chromosome. Hence, maleness itself is Y linked and shows
but to none of their sons. the expected pattern of exclusively male-to-male transmis-
sion. Some cases of male sterility have been shown to be
caused by deletions of Y-chromosome regions containing
Androgen insensitivity phenotype sperm-promoting genes. Male sterility is not heritable, but,
interestingly, the fathers of these men have normal Y chro-
mosomes, showing that the deletions are new.
There have been no convincing cases of nonsexual
phenotypic variants associated with the Y chromosome,
although there are cases in other animals.
autosomal recessive disease caused by malfunction of the 3. The overall probability of the man being a heterozy-
enzyme hexosaminidase A. The defect leads to the buildup gote must be calculated with the use of a statistical rule
of fatty deposits in nerve cells, causing paralysis followed called the product rule, which states that
by an early death. The pedigree is as follows:
The probability of two independent events both
occurring is the product of their individual
probabilities.
Because gene transmissions in different generations are
independent events, we can calculate that the probabil-
? ity of the man being a heterozygote is the probability of
his father being a heterozygote (2/3) times the probabil-
The probability of the couple’s first child having ity of his father having a heterozygous son (1/2), which
Tay-Sachs can be calculated in the following way. Because is 2/3 × 1/2 = 1/3.
neither of the couple has the disease, each can only be an
4. Likewise, the probability of the woman being heterozy-
unaffected homozygote or heterozygote. If both are hetero-
gous is also 1/3.
zygotes, then they each stand a chance of passing the reces-
sive allele on to a child, who would then have Tay-Sachs 5. If they are both heterozygous (T /t) , their mating would
disease. Hence, we must calculate the probability of their be a standard monohybrid cross, and so the probability
both being heterozygotes, and then, if so, the probability of of their having a t /t child is 1/4.
passing the deleterious allele on to a child. 6. Overall, the probability of the couple having an affected
child is the probability of them both being heterozygous
1. The man’s grandparents must have both been heterozy-
and then both transmitting the recessive allele to a child.
gotes (T /t) because they produced a t /t child (the uncle).
Again, these events are independent, and so we can cal-
Therefore, they effectively constituted a monohybrid
culate the overall probability as 1/3 × 1/3 × 1/4 = 1/36. In
cross. The man’s father could be T /T or T /t, but within
other words, there is a 1 in 36 chance of their having a
the 3/4 of unaffected progeny, we know that the rela-
child with Tay-Sachs disease.
tive probabilities of these genotypes must be 1/4 and 1/2,
In some Jewish communities, the Tay-Sachs allele is
respectively (the expected progeny ratio in a monohybrid
not as rare as it is in the general population. In such cases,
cross is 41 T /T, 12 T /t, 41 t /t). Therefore, there is a 2/3 prob-
unaffected people who enter into families with a history
ability that the man’s father is a heterozygote (two-thirds
of Tay-Sachs cannot be assumed to be T /T . If the fre-
is the proportion of unaffected progeny who are heterozy-
quency of T /t heterozygotes in the community is known,
gotes: that is, the ratio of 2/4 to 3/4).
this frequency can be factored into the product-rule cal-
2. The man’s mother is assumed to be T /T , because she culation. Nowadays, molecular diagnostic tests for Tay-
is an outsider who entered into the family and dis- Sachs alleles are available, and the judicious use of these
ease alleles are generally rare. Thus, if the father is T /t, tests has drastically reduced the frequency of the disease in
then the mating with the mother was a cross T /t × T /T some communities.
and the expected proportions in the progeny (which
includes the man in question) are 12 T /T, 12 T /t .
SUMMARY
In somatic cell division, the genome is transmitted by mito- An easy way to remember the main events of meiosis,
sis, a nuclear division. In this process, each chromosome by using your fingers to represent chromosomes, is shown
replicates into a pair of chromatids, and the chromatids are in Figure 2-32.
pulled apart to produce two identical daughter cells. (Mito- Genetic dissection of a biological trait begins with a col-
sis can take place in diploid or haploid cells.) At meiosis, lection of mutants. Each mutant has to be tested to see if it
which takes place in the sexual cycle in meiocytes, each is inherited as a single-gene change. The procedure followed
chromosome replicates to form a pair of chromatids; then, is essentially unchanged from the time of Mendel, who per-
homologous chromosomes pair up at the equatorial plane formed the prototypic analysis of this type. The analysis is
of the cell. The homologous chromosomes (each a pair of based on observing specific phenotypic ratios in the progeny
chromatids) segregate over the course of two cell divisions. of controlled crosses. In a typical case, a cross of A /A × a /a
The result is four haploid cells, or gametes. Meiosis can take produces an F1 that is all A /a. When the F1 is selfed or inter-
place only in a diploid cell; hence, haploid organisms must crossed, a genotypic ratio of 41 A /A : 12 A /a : 41 a /a is produced
temporarily unite to form a diploid meiocyte. in the F2. (At the phenotypic level, this ratio is 43 A /− : 41 a /a.)
The main events of mitosis and meiosis The three single-gene genotypes are homozygous dominant,
heterozygous (monohybrid), and homozygous recessive.
Mitosis Meiosis If an A /a individual is crossed with a /a (a testcross), a 1:1
Pair of homologous ratio is produced in the progeny. The 1:1, 3:1, and 1:2:1
chromosomes ratios stem from the principle of equal segregation, which
is that the haploid products of meiosis from A /a will be 12 A
and 12 a. The cellular basis of the equal segregation of alleles
is the segregation of homologous chromosomes at meiosis.
Chromatid Haploid fungi can be used to show equal segregation at the
formation
level of a single meiosis (a 1:1 ratio in an ascus).
The molecular basis for chromatid production in mei-
osis is DNA replication. Segregation at meiosis can be
observed directly at the molecular (DNA) level. The molec-
Alignment at equator Pairing at equator ular force of segregation is the depolymerization and subse-
quent shortening of spindle fibers that are attached to the
centromeres. Recessive mutations are generally in genes that
are haplosufficient, whereas dominant mutations are often
due to gene haploinsufficiency.
In many organisms, sex is determined chromosomally,
Anaphase Anaphase I
and, typically, XX is female and XY is male. Genes on the
X chromosome (X-linked genes) typically have no counter-
parts on the Y chromosome and show a single-gene inheri-
tance pattern that differs in the two sexes, often resulting in
different ratios in the male and female progeny.
Anaphase II Mendelian single-gene segregation is useful in identifying
mutant alleles underlying many human disorders. Analyses of
pedigrees can reveal autosomal or X-linked disorders of both
dominant and recessive types. The logic of Mendelian genetics
has to be used with caution, taking into account that human
progeny sizes are small and phenotypic ratios are not neces-
FIGURE 2-32 Using fingers to remember the main events of mitosis sarily typical of those expected from larger sample sizes. If a
and meiosis. known single-gene disorder is present in a pedigree, Mendelian
logic can be used to predict the likelihood of children inheriting
the disease.
KEY TERMS
allele (p. 34) heterozygote (p. 35) parental generation (P) (p. 32)
character (p. 30) heterozygous (p. 35) pedigree analysis (p. 52)
chromatid (p. 38) homogametic sex (p. 48) phenotype (p. 30)
cross (p. 31) homozygote (p. 35) polymorphism (p. 56)
dimorphism (p. 56) homozygous dominant (p. 35) product rule (p. 61)
dioecious species (p. 48) homozygous recessive (p. 35) propositus (p. 52)
diploid (p. 36) law of equal segregation (Mendel’s pseudoautosomal regions 1 and 2
dominant (p. 34) first law) (p. 35) (p. 49)
first filial generation (F1) (p. 32) leaky mutation (p. 44) pure line (p. 32)
forward genetics (p. 31) mating types (p. 40) recessive (p. 34)
gene (p. 34) meiocyte (p. 36) reverse genetics (p. 31)
gene discovery (p. 30) meiosis (p. 36) second filial generation (F2 ) (p. 33)
genetic dissection (p. 31) mitosis (p. 36) self (p. 32)
genotype (p. 35) monohybrid (p. 35) sex chromosome (p. 48)
haploid (p. 36) monohybrid cross (p. 35) sex linkage (p. 49)
haploinsufficient (p. 45) morph (p. 56) sexual cell division (p. 36)
haplosufficient (p. 45) mutant (p. 30) silent mutation (p. 44)
hemizygous (p. 49) mutation (p. 30) somatic cell division (p. 36)
heterogametic sex (p. 48) null allele (p. 44) SRY gene (p. 60)
testcross (p. 48) wild type (p. 30) Y chromosome (p. 48)
tester (p. 48) X chromosome (p. 48) Y linkage (p. 49)
trait (p. 30) X linkage (p. 49) zygote (p. 35)
SOLVED PROBLEMS
This section in each chapter contains a few solved problems must be dominant, and so we can represent the white allele
that show how to approach the problem sets that follow. as W and the yellow allele as w. The results can then be
The purpose of the problem sets is to challenge your under- expressed as follows:
standing of the genetic principles learned in the chapter. The
P W /W × w /w
best way to demonstrate an understanding of a subject is
to be able to use that knowledge in a real or simulated sit- F1 W /w
1
uation. Be forewarned that there is no machine-like way of F2 4
W /W
solving these problems. The three main resources at your 1
W /w
2
disposal are the genetic principles just learned, logic, and 1
w /w
trial and error. 4
Here is some general advice before beginning. First, for No doubt, if the parental rabbits had been sacrificed,
each problem, it is absolutely essential to read and under- one parent (we cannot tell which) would have been pre-
stand the entire problem. Most of the problems use data dicted to have white fat and the other yellow. Luckily, the
taken from research that somebody actually carried out: rabbits were not sacrificed, and the same animals were bred
ask yourself why the research might have been initiated and again, leading to a very interesting, different result. Often
what was the probable goal. Find out exactly what facts are in science, an unexpected observation can lead to a novel
provided, what assumptions have to be made, what clues principle, and, rather than moving on to something else, it is
are given in the problem, and what inferences can be made useful to try to explain the inconsistency. So why did the 3:1
from the available information. Second, be methodical. Star- ratio disappear? Here are some possible explanations.
ing at the problem rarely helps. Restate the information in First, perhaps the genotypes of the parental animals
the problem in your own way, preferably using a diagram- had changed. This type of spontaneous change affecting the
matic representation or flowchart to help you think out the whole animal, or at least its gonads, is very unlikely, because
problem. Good luck. even common experience tells us that organisms tend to be
stable to their type.
Second, in the repeat, the sample of 22 F2 animals did
SOLVED PROBLEM 1 not contain any yellow fat simply by chance (“bad luck”).
Crosses were made between two pure lines of rabbits that This explanation, again, seems unlikely, because the sample
we can call A and B. A male from line A was mated with was quite large, but it is a definite possibility.
a female from line B, and the F1 rabbits were subsequently A third explanation draws on the principle that genes do not
intercrossed to produce an F2 . Three-fourths of the F2 ani- act in a vacuum; they depend on the environment for their effects.
mals were discovered to have white subcutaneous fat, and Hence, the formula “genotype + environment = phenotype”
one-fourth had yellow subcutaneous fat. Later, the F1 was is a useful mnemonic. A corollary of this formula is that genes
examined and was found to have white fat. Several years can act differently in different environments; so
later, an attempt was made to repeat the experiment by genotype 1 + environment 1 = phenotype 1
using the same male from line A and the same female from
line B. This time, the F1 and all the F2 (22 animals) had white and
fat. The only difference between the original experiment genotype 1 + environment 2 = phenotype 2
and the repeat that seemed relevant was that, in the origi-
nal, all the animals were fed fresh vegetables, whereas in the In the present problem, the different diets constituted
repeat, they were fed commercial rabbit chow. Provide an different environments, and so a possible explanation of the
explanation for the difference and a test of your idea. results is that the homozygous recessive w /w produces yel-
low fat only when the diet contains fresh vegetables. This
SOLUTION explanation is testable. One way to test it is to repeat the
The first time that the experiment was done, the breeders experiment again and use vegetables as food, but the par-
would have been perfectly justified in proposing that a pair ents might be dead by this time. A more convincing way
of alleles determine white versus yellow body fat because is to breed several of the white-fatted F2 rabbits from the
the data clearly resemble Mendel’s results in peas. White second experiment. According to the original interpretation,
SOLVED PROBLEM 2
Phenylketonuria (PKU) is a human hereditary disease
resulting from the inability of the body to process the
chemical phenylalanine, which is contained in the protein
that we eat. PKU is manifested in early infancy and, if it
remains untreated, generally leads to intellectual disabilities.
PKU is caused by a recessive allele with simple Mendelian
inheritance.
A couple intends to have children but consult a genetic 1 2 3 4 5 6 7 8 9 10
counselor because the man has a sister with PKU and the
a. Deduce the most likely mode of inheritance.
woman has a brother with PKU. There are no other known
cases in their families. They ask the genetic counselor to b. What would be the outcomes of the cousin matings
determine the probability that their first child will have 1× 9, 1× 4, 2 × 3, and 2 × 8?
PKU. What is this probability?
SOLUTION
SOLUTION a. The most likely mode of inheritance is X-linked dom-
What can we deduce? If we let the allele causing the PKU inant. We assume that the disease phenotype is dom-
phenotype be p and the respective normal allele be P, then inant because, after it has been introduced into the
the sister and brother of the man and woman, respectively, pedigree by the male in generation II, it appears in
must have been p /p . To produce these affected persons, all every generation. We assume that the phenotype is
four grandparents must have been heterozygous normal. X linked because fathers do not transmit it to their
The pedigree can be summarized as follows: sons. If it were autosomal dominant, father-to-son
P/p P/p P/p P/p transmission would be common.
In theory, autosomal recessive could work, but it
... ... is improbable. In particular, note the matings between
affected members of the family and unaffected out-
siders. If the condition were autosomal recessive,
p/p P/ P/ p/p
the only way in which these matings could have
?
affected offspring is if each person entering into the
When these inferences have been made, the problem family were a heterozygote; then the matings would
is reduced to an application of the product rule. The only be a /a (affected) × A /a (unaffected) . However, we are
way in which the man and woman can have a PKU child told that the disease is rare; in such a case, heterozy-
is if both of them are heterozygotes (it is obvious that they gotes are highly unlikely to be so common. X-linked
themselves do not have the disease). Both the grandparental recessive inheritance is impossible, because a mating
matings are simple Mendelian monohybrid crosses expected of an affected woman with a normal man could not
to produce progeny in the following proportions: produce affected daughters. So we can let A represent
1× 4: Must be A /a ♀ × a /Y ♂.
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a Problem Solving Video. Problems with the icon have an Unpacking the Problem exercise.
www
www
WORKING WITH THE FIGURES 11. Considering Figure 2-14, if you had a homozygous
double mutant m3/m3 m5/m5, would you expect it to
(The first 25 questions require inspection of text figures.)
be mutant in phenotype? (Note: This line would have
1. In Figure 2-1, what seems to be abnormal about the two mutant sites in the same coding sequence.)
Arabidopsis mutants ap3 and ag?
12. In Figure 2-14, mutant m5, whose locus is not in an
2. In the left-hand part of Figure 2-4 , the red arrows active site (green area), is silent. Can you think of a
show selfing as pollination within single flowers of way in which a mutation in such a green area could
one F1 plant. Would the same F2 results be produced by have a phenotypic effect?
cross-pollinating two different F1 plants?
13. In which of the stages of the Drosophila life cycle (rep-
3. In the right-hand part of Figure 2-4, in the plant show- resented in the box on page 50) would you find the
ing an 11:11 ratio, do you think it would be possible products of meiosis?
to find a pod with all yellow peas? All green? Explain.
14. If you assume Figure 2-16 also applies to mice and you
4. In Table 2-1, state the recessive phenotype in each of irradiate male sperm with X rays (known to inactivate
the seven cases. genes), what phenotype would you look for in prog-
5. In Table 2-1, line 6, if the 651 axial F1 plants were eny in order to find cases of individuals with an inacti-
individually selfed, approximately how many of them vated SRY gene?
would have some terminal progeny? 15. In Figure 2-18, how does the 3:1 ratio in the bottom-
6. In Figure 2-6, what would you say was the “job” of left-hand grid differ from the 3:1 ratios obtained by
mitosis in each of the three life cycles? Mendel?
7. Considering Figure 2-9 , is the sequence “pairing → 16. In Figure 2-18, what progeny would you predict from
replication → segregation → segregation ” a good short- a cross of a red F2 male from the first cross and a red F2
hand description of meiosis? female from the second cross?
8. In assessing Figure 2-9, what would you say was the 17. In Figure 2-20, assume that the pedigree is for mice, in
main difference between the metaphase of mitosis and which any chosen cross can be made. If you bred IV-1
the metaphase II of meiosis? with IV-3, what is the probability that the first baby
will show the recessive phenotype?
9. In Figure 2-11, is there any case of a chromosome pro-
ducing one GC-bearing daughter chromatid and one 18. In Figure 2-20, (1) can you tell which of the generation
AT-bearing daughter chromatid? I parents is heterozygous in this pedigree? (2) Is gener-
ation IV an F2 Mendelian ratio? Explain.
10. In Figure 2-12, assume (as in corn plants) that allele
A encodes an allele that produces starch in pollen and 19. In Figure 2-21, list all the mutations that affect the
that allele a does not. Iodine solution stains starch human nervous system.
black. How would you demonstrate Mendel’s first law 20. Which part of the pedigree in Figure 2-24 in your
directly with such a system? opinion best demonstrates Mendel’s first law?
21. Considering all the individuals shown in Figure 2-24, 33. In a possible future scenario, male fertility drops to
would a cross between any male and female produce zero, but, luckily, scientists develop a way for women
progeny all of which are A /A ? to produce babies by virgin birth. Meiocytes are con-
22. In Figure 2-25b, the first progeny set in generation verted directly (without undergoing meiosis) into
IV contains three afflicted children. Can you explain zygotes, which implant in the usual way. What would
why? be the short- and long-term effects in such a society?
23. In Figure 2-27, what are the likely genotypes of II-1 34. In what ways does the second division of meiosis differ
and III-11? from mitosis?
24. Could the pedigree in Figure 2-31 be explained as an 35. Make up mnemonics for remembering the five stages
autosomal dominant disorder? Explain. of prophase I of meiosis and the four stages of mitosis.
25. Refer back to Figure 1-15 concerning the condition 36. In an attempt to simplify meiosis for the benefit of stu-
ACDC and the family tree of Louise Benge. dents, mad scientists develop a way of preventing pre-
meiotic S phase and making do with having just one
a. Assign Mendelian allelic symbols for the affected division, including pairing, crossing over, and segre-
and unaffected condition and apply them to the gation. Would this system work, and would the prod-
individuals in generations V, VI, and VII. Indicate ucts of such a system differ from those of the present
cases where the genotype is in doubt. system?
b. Discuss the siblings in generation VI in terms of 37. Theodor Boveri said, “The nucleus doesn’t divide; it is
Mendelian ratios covered in Chapter 2. divided.” What was he getting at?
c. In a family with five children born to parents who 38. Francis Galton, a geneticist of the pre-Mendelian era,
are both unaffected carriers of the mutant allele, devised the principle that half of our genetic makeup
what is the probability that all five children would is derived from each parent, one-quarter from each
be affected? grandparent, one-eighth from each great-grandparent,
and so forth. Was he right? Explain.
BASIC PROBLEMS 39. If children obtain half their genes from one parent
26. Make up a sentence including the words chromosome, and half from the other parent, why aren’t siblings
genes, and genome. identical?
27. Peas ( Pisum sativum ) are diploid and 2n = 14 . In 40. State where cells divide mitotically and where they
Neurospora, the haploid fungus, n = 7. If you were to divide meiotically in a fern, a moss, a flowering plant,
isolate genomic DNA from both species and use gel a pine tree, a mushroom, a frog, a butterfly, and a
electrophoresis to separate DNA molecules by size, snail.
how many distinct DNA bands would be visible in 41. Human cells normally have 46 chromosomes. For each
each species? (See Section 10.1 and Figure 10.4 for a of the following stages, state the number of nuclear
description of the gel electrophoresis technique.) DNA molecules present in a human cell:
28. The broad bean (Vicia faba) is diploid and 2n = 18. a. Metaphase of mitosis
Each haploid chromosome set contains approximately b. Metaphase I of meiosis
4 m of DNA. The average size of each chromosome
during metaphase of mitosis is 13 µm. What is the c. After telophase of mitosis
average packing ratio of DNA at metaphase? (Packing d. After telophase I of meiosis
ratio = length of chromosome/length of DNA mole- e. After telophase II of meiosis
cule therein.) How is this packing achieved?
42. Four of the following events are part of both meio-
29. If we call the amount of DNA per genome “x,” then sis and mitosis, but only one is meiotic. Which one?
name a situation or situations in diploid organisms in (1) Chromatid formation, (2) spindle formation,
which the amount of DNA per cell is (3) chromosome condensation, (4) chromosome move-
a. x ment to poles, (5) chromosome pairing.
b. 2x 43. In corn, the allele f ′ causes floury endosperm, and the
c. 4x allele f ′′ causes flinty endosperm. In the cross f ′ /f ′ ♀ ×
f ′′ /f ′′ ♂, all the progeny endosperms are floury, but,
30. Name the key function of mitosis. in the reciprocal cross, all the progeny endosperms are
31. Name two key functions of meiosis. flinty. What is a possible explanation? (Check the leg-
32. Design a different nuclear-division system that would end for Figure 2-6.)
achieve the same outcome as that of meiosis. 44. What is Mendel’s first law?
45. If you had a fruit fly (Drosophila melanogaster) that inheritance of three cotyledons? Invent gene symbols
was of phenotype A, what cross would you make to as part of your explanation.
determine if the fly’s genotype was A /A or A /a ? 52. In the plant Arabidopsis thaliana, a geneticist is inter-
46. In examining a large sample of yeast colonies on a ested in the development of trichomes (small pro-
petri dish, a geneticist finds an abnormal-looking col- jections). A large screen turns up two mutant plants
ony that is very small. This small colony was crossed (A and B) that have no trichomes, and these mutants
with wild type, and products of meiosis (ascospores) seem to be potentially useful in studying trichome
were spread on a plate to produce colonies. In total, development. (If they were determined by single-gene
there were 188 wild-type (normal-size) colonies and mutations, then finding the normal and abnormal
180 small ones. functions of these genes would be instructive.) Each
a. What can be deduced from these results regarding plant is crossed with wild type; in both cases, the next
the inheritance of the small-colony phenotype? generation (F1) had normal trichomes. When F1 plants
(Invent genetic symbols.) were selfed, the resulting F2 progeny were as follows:
b. What would an ascus from this cross look like? F2 from mutant A: 602 normal; 198 no trichomes
47. Two black guinea pigs were mated and over sev- F2 from mutant B: 267 normal; 93 no trichomes
eral years produced 29 black and 9 white offspring. a. What do these results show? Include proposed gen-
Explain these results, giving the genotypes of parents otypes of all plants in your answer.
and progeny. b. Under your explanation to part a, is it possible to
48. In a fungus with four ascospores, a mutant allele lys-5 confidently predict the F1 from crossing the original
causes the ascospores bearing that allele to be white, mutant A with the original mutant B?
whereas the wild-type allele lys-5+ results in black 53. You have three dice: one red (R), one green (G), and
ascospores. (Ascospores are the spores that constitute one blue (B). When all three dice are rolled at the
the four products of meiosis.) Draw an ascus from same time, calculate the probability of the following
each of the following crosses: outcomes:
a. lys-5 × lys-5+ a. 6 (R), 6 (G), 6 (B)
b. lys-5 × lys-5 b. 6 (R), 5 (G), 6 (B)
c. lys-5+ × lys-5+ c. 6 (R), 5 (G), 4 (B)
49. For a certain gene in a diploid organism, eight units of d. No sixes at all
protein product are needed for normal function. Each
wild-type allele produces five units. e. A different number on all dice
a. If a mutation creates a null allele, do you think this 54. In the pedigree below, the black symbols represent
allele will be recessive or dominant? individuals with a very rare blood disease.
56. John and Martha are contemplating having children, Objectives and Key Concepts of this chapter and ask
but John’s brother has galactosemia (an autosomal yourself which might be relevant to your difficulty.
recessive disease) and Martha’s great-grandmother
also had galactosemia. Martha has a sister who has 57. Holstein cattle are normally black and white. A superb
three children, none of whom have galactosemia. black-and-white bull, Charlie, was purchased by a
What is the probability that John and Martha’s first farmer for $100,000. All the progeny sired by Charlie
child will have galactosemia? were normal in appearance. However, certain pairs of
his progeny, when interbred, produced red-and-white
www
UNPACKING PROBLEM 56 progeny at a frequency of about 25 percent. Charlie
www
was soon removed from the stud lists of the Holstein
Before attempting a solution to this problem, try answer- breeders. Use symbols to explain precisely why.
ing the following questions:
58. Suppose that a man and a woman are both heterozy-
1. Can the problem be restated as a pedigree? If so, gous for a recessive allele for albinism. If they have
write one. dizygotic (two-egg) twins, what is the probability
2. Can parts of the problem be restated by using Pun- that both the twins will have the same phenotype for
nett squares? pigmentation?
3. Can parts of the problem be restated by using 59. The plant blue-eyed Mary grows on Vancouver Island
branch diagrams? and on the lower mainland of British Columbia. The
populations are dimorphic for purple blotches on
4. In the pedigree, identify a mating that illustrates the leaves—some plants have blotches and others do
Mendel’s first law. not. Near Nanaimo, one plant in nature had blotched
5. Define all the scientific terms in the problem, and look leaves. This plant, which had not yet flowered, was dug
up any other terms about which you are uncertain. up and taken to a laboratory, where it was allowed to
6. What assumptions need to be made in answering self. Seeds were collected and grown into progeny. One
this problem? randomly selected (but typical) leaf from each of the
progeny is shown in the accompanying illustration.
7. Which unmentioned family members must be con-
sidered? Why?
8. What statistical rules might be relevant, and in what
situations can they be applied? Do such situations
exist in this problem?
9. What are two generalities about autosomal reces-
sive diseases in human populations?
10. What is the relevance of the rareness of the pheno-
type under study in pedigree analysis generally, and
what can be inferred in this problem?
11. In this family, whose genotypes are certain, and
whose are uncertain?
12. In what way is John’s side of the pedigree different
from Martha’s side? How does this difference affect
your calculations?
13. Is there any irrelevant information in the problem
as stated?
14. In what way is solving this kind of problem similar
to solving problems that you have already success-
fully solved? In what way is it different?
15. Can you make up a short story based on the human
dilemma in this problem?
Now try to solve the problem. If you are unable to do so, a. Formulate a concise genetic hypothesis to explain
try to identify the obstacle and write a sentence or two these results. Explain all symbols and show all
describing your difficulty. Then go back to the expansion genotypic classes (and the genotype of the original
questions and see if any of them relate to your difficulty. plant).
If this approach does not work, inspect the Learning b. How would you test your hypothesis? Be specific.
60. Can it ever be proven that an animal is not a carrier of b. Give genotypes for as many individuals in the ped-
a recessive allele (that is, not a heterozygote for a given igree as possible. (Invent your own defined allele
gene)? Explain. symbols.)
61. In nature, the plant Plectritis congesta is dimorphic for c. Consider the four unaffected children of parents
fruit shape; that is, individual plants bear either wing- III-4 and III-5. In all four-child progenies from
less or winged fruits, as shown in the illustration. parents of these genotypes, what proportion is
expected to contain all unaffected children?
63. Four human pedigrees are shown in the accompanying
illustration. The black symbols represent an abnormal
phenotype inherited in a simple Mendelian manner.
1
Wingless (selfed) 4* 80
Winged × wingless 161 0
Winged × wingless 29 31
Winged × wingless 46 0
Winged × winged 44 0
*Phenotype probably has a nongenetic explanation. 4
III
IV
A B
marries a normal woman. If they have children, what d. Of all the children (sex unspecified) of these par-
proportion of their sons will have hypophosphatemia? ents, what proportion can be expected to have nor-
73. Duchenne muscular dystrophy is sex linked and usu- mal color vision?
ally affects only males. Victims of the disease become 78. Male house cats are either black or orange; females are
progressively weaker, starting early in life. black, orange, or calico.
a. What is the probability that a woman whose a. If these coat-color phenotypes are governed by a
brother has Duchenne’s disease will have an sex-linked gene, how can these observations be
affected child? explained?
b. If your mother’s brother (your uncle) had Duchenne’s b. Using appropriate symbols, determine the pheno-
disease, what is the probability that you have received types expected in the progeny of a cross between
the allele? an orange female and a black male.
c. If your father’s brother had the disease, what is the c. Half the females produced by a certain kind of
probability that you have received the allele? mating are calico, and half are black; half the males
74. A man and woman discover that each had an uncle are orange, and half are black. What colors are the
with alkaptonuria (black urine disease), a rare dis- parental males and females in this kind of mating?
ease caused by an autosomal recessive allele of a d. Another kind of mating produces progeny in
single gene. They are about to have their first baby. the following proportions: one-fourth orange
What is the probability that their child will have males, one-fourth orange females, one-fourth black
alkaptonuria? males, and one-fourth calico females. What colors
75. The accompanying pedigree concerns a rare inherited are the parental males and females in this kind of
dental abnormality, amelogenesis imperfecta. mating?
79. The pedigree below concerns a certain rare disease
that is incapacitating but not fatal.
a. What genotypes are possible for the mother of the 81. A plant geneticist has two pure lines, one with purple
color-blind man? petals and one with blue. She hypothesizes that the
phenotypic difference is due to two alleles of one gene.
b. If the couple has children, what are the chances To test this idea, she aims to look for a 3:1 ratio in
that their first child will be a color-blind boy? the F2 . She crosses the lines and finds that all the F1
c. Of any girls that may be born to these parents, progeny are purple. The F1 plants are selfed, and 400 F2
what proportion can be expected to be color blind? plants are obtained. Of these F2 plants, 320 are purple
and 80 are blue. Do these results fit her hypothesis example of a dimorphism. Its significance is a com-
well? If not, suggest why. plete mystery. In one family, a boy was unable to roll
82. A man’s grandfather has galactosemia, a rare auto- his tongue but, to his great chagrin, his sister could.
somal recessive disease caused by the inability to Furthermore, both his parents were rollers, and so
process galactose, leading to muscle, nerve, and kid- were both grandfathers, one paternal uncle, and one
ney malfunction. The man married a woman whose paternal aunt. One paternal aunt, one paternal uncle,
sister had galactosemia. The woman is now preg- and one maternal uncle could not roll their tongues.
nant with their first child. www a. Draw the pedigree for this family, defining your
a. Draw the pedigree as described.www symbols clearly, and deduce the genotypes of as
many individual members as possible.
b. What is the probability that this child will have
galactosemia? b. The pedigree that you drew is typical of the inher-
itance of tongue rolling and led geneticists to come
c. If the first child does have galactosemia, what is up with the inheritance mechanism that no doubt
the probability that a second child will have it? you came up with. However, in a study of 33 pairs
of identical twins, both members of 18 pairs could
CHALLENGING PROBLEMS roll, neither member of 8 pairs could roll, and one
83. A geneticist working on peas has a single plant mono- of the twins in 7 pairs could roll but the other
hybrid Y /y (yellow) plant and, from a self of this plant, could not. Because identical twins are derived from
wants to produce a plant of genotype y /y to use as a the splitting of one fertilized egg into two embryos,
tester. How many progeny plants need to be grown to the members of a pair must be genetically identi-
be 95 percent sure of obtaining at least one in the cal. How can the existence of the seven discordant
sample? pairs be reconciled with your genetic explanation
of the pedigree?
84. A curious polymorphism in human populations has to
do with the ability to curl up the sides of the tongue 85. Red hair runs in families, as the pedigree here shows.
to make a trough (“tongue rolling”). Some people can (Pedigree data from W. R. Singleton and B. Ellis, Jour-
do this trick, and others simply cannot. Hence, it is an nal of Heredity 55, 1964, 261.)
I ?
II
III
IV
II
III
IV
VI
Red hair Red beard and body hair
a. Does the inheritance pattern in this pedigree sug- “porcupine man” married and had six sons, all of
gest that red hair could be caused by a dominant whom had this condition, and several daughters, all
or a recessive allele of a gene that is inherited in a of whom were normal. For four generations, this con-
simple Mendelian manner? dition was passed from father to son. From this evi-
b. Do you think that the red-hair allele is common or dence, what can you postulate about the location of
rare in the population as a whole? the gene?
86. When many families were tested for the ability to taste 88. The wild-type (W) Abraxas moth has large spots
the chemical phenylthiocarbamide, the matings were on its wings, but the lacticolor (L) form of this spe-
grouped into three types and the progeny were totaled, cies has very small spots. Crosses were made between
with the results shown below: strains differing in this character, with the following
results:
Children
Number Non- Parents Progeny
Parents of families Tasters Parents
Cross ♀ ♂ F1 F2
Taster × taster 425 929 130
1 L W ♀ W ♀ 1
L, 1
W
2 2
Taster × nontaster 289 483 278
♂W ♂ W
Nontaster × nontaster 86 5 218
2 W L ♀ L ♀ 1
W, 1
L
2 2
With the assumption that PTC tasting is dominant (P) ♂W ♂ 1
W, 1
L
2 2
and nontasting is recessive (p), how can the progeny
ratios in each of the three types of mating be accounted
Provide a clear genetic explanation of the results in
for?
these two crosses, showing the genotypes of all indi-
87. A condition known as icthyosis hystrix gravior vidual moths.
appeared in a boy in the early eighteenth century. His
89. This pedigree shows the inheritance of a rare human
skin became very thick and formed loose spines that
disease.
were sloughed off at intervals. When he grew up, this
90. A certain type of deafness in humans is inherited as a. Concisely state exactly what is unusual about this
an X-linked recessive trait. An unaffected woman pedigree.
is expecting a child with a man who has this type of b. Can the pattern be explained by Mendelian
deafness. They find out that they are distantly related. inheritance?
Part of the family tree is shown here.
GENETICS AND SOCIETY
Spindle
Pole
4 Mitotic metaphase
The photographs show mitosis in the nuclei of root-tip cells of Lilium regale. [Republished A N I M ATED A RT
with permission of Springer Nature, after J. McLeish and B. Snoad, Looking at Chromosomes, Mitosis
copyright 1972, St. Martin’s, Macmillan, Red Globe Press, permission conveyed through Copyright
Clearance Center, Inc.] 75
15 The tetrad
Cells divide
Telophase II: The nuclei
re-form around the
chromosomes at the poles.
76
4 Diplotene 5 Diakinesis
Prophase I: Diplotene. Although the DNA has already Prophase I: Diakinesis. Further
replicated during the premeiotic S phase, this fact first chromosome contraction produces
becomes manifest only in diplotene as each chromosome compact units that are very
is seen to have become a pair of sister chromatids. The maneuverable.
synapsed structure now consists of a bundle of four
homologous chromatids. The paired homologs separate
slightly, and one or more cross-shaped structures called
chiasmata (singular, chiasma) appear between nonsister Metaphase I: The nuclear membrane
chromatids. has disappeared, and each pair of
homologs takes up a position in the
equatorial plane. At this stage of
meiosis, the centromeres do not divide;
this lack of division is a major difference
from mitosis. The two centromeres of a
homologous chromosome pair attach to 6 Metaphase I
spindle fibers from opposite poles.
7 Early anaphase I
Anaphase I: The members of each
homologous pair move to opposite poles.
Telophase I and interphase: The chromosomes elongate and become diffuse, the nuclear 8 Later anaphase I
membrane re-forms, and the cell divides. After telophase I, there is an interphase, called interkinesis.
In many organisms, telophase I and interkinesis do not exist or are brief in duration. In any case, there
is never DNA synthesis at this time, and the genetic state of the chromosomes does not change.
Cell divides
10 Interphase 9 Telophase I
77
Independent Assortment
of Genes
T
his chapter is about the principles at work when two To make a truly superior genotype, combining such
or more cases of single-gene inheritance are analyzed alleles into one line is clearly desirable. To achieve such a
simultaneously. Nowhere have these principles been combination, mutant lines must be intercrossed two at a
more important than in plant and animal breeding in agri- time. For instance, a plant geneticist might start by cross-
culture. For example, between the years 1960 and 2000, the ing a strain homozygous for sd1 to another homozygous for
world production of food plants doubled, marking a so-called Xa4. The F1 progeny of this cross would carry both muta-
Green Revolution. What made this Green Revolution possible? tions, but in a heterozygous state. However, most agriculture
In part, it was due to improved agricultural practice, but more uses pure lines, because they can be efficiently propagated
important was the development of superior crop genotypes by and distributed to farmers. To obtain a pure-breeding,
plant geneticists. These breeders are constantly on the lookout doubly mutant sd1/sd1 ⋅ Xa4/Xa4 line, the F1 would have
for the chance occurrence of single-gene mutations that signifi- to be bred further to allow the alleles to “assort” into the
cantly increase yield or nutrient value. However, such muta- desirable combination. Some products of such breeding are
tions arise in different lines in different parts of the world. For shown in Figure 3-1. What must take place during gamete
example, in rice, one of the world’s main food crops, the fol- formation and fertilization in order for the offspring to
lowing mutations have been crucial in the Green Revolution: obtain the desired combination of alleles? It depends very
much on whether the two genes are on the same chromo-
sd1. This recessive allele results in short stature, making the
some pair or on different chromosome pairs. In the latter
plant more resistant to “lodging,” or falling over, in wind
case, the chromosome pairs act independently at meiosis,
and rain; it also increases the relative amount of the plant’s
and the alleles of two heterozygous gene pairs are said to
energy that is routed into the seed, the part that we eat.
show independent assortment.
se1. This recessive allele alters the plant’s requirement This chapter explains how we can recognize indepen-
for a specific daylength, enabling it to be grown at dif- dent assortment and how the principle of independent
ferent latitudes. assortment can be used in strain construction, both in agri-
Xa4. This dominant allele confers resistance to the dis- culture and in basic genetic research. (Chapter 4 covers the
ease bacterial blight. analogous principles applicable to heterozygous gene pairs
bph2. This allele confers resistance to brown plant hop- located on the same chromosome pair.)
pers (a type of insect). We shall also see that independent assortment of an
array of genes is also useful in providing a basic mechanism
Snb1. This allele confers tolerance to plant submersion
of inheritance for traits that display continuous phenotypes.
after heavy rains.
80
These are traits such as height or weight where phenotypes KEY CONCEPT Dihybrids, organisms heterozygous for two
do not fall into distinct categories but are nevertheless often genes ( A /a ⋅ B /b) are the key genotypes for the analysis of
heavily influenced by multiple genes collectively called poly independent assortment in this chapter, and departures from it
genes. We shall examine the role of independent assortment in Chapter 4.
in the inheritance of continuous phenotypes influenced by
such polygenes. We will see that independent assortment of
polygenes can produce a continuous phenotypic distribu- The pair of characters that Mendel began working with
tion among progeny. were seed shape and seed color. We have already followed the
Lastly, we will introduce a different type of independent monohybrid cross for seed color (Y /y × Y /y ), which gave a
inheritance, that of genes in the organelles mitochondria progeny ratio of 3 yellow:1 green (see Figure 2-5). The seed
and chloroplasts. Unlike nuclear chromosomes, these genes shape phenotypes (Figure 3-2) were round (determined by
are inherited cytoplasmically and result in different pat- allele R) and wrinkled (determined by allele r). The mono-
terns than observed for nuclear genes and chromosomes. hybrid cross R /r × R /r gave a progeny ratio of 3 round:1
However, this cytoplasmic inheritance pattern is indepen- wrinkled as expected (see Table 2-1, page 34). To perform a
dent of genes showing nuclear inheritance, which is why dihybrid cross, Mendel started with two pure parental lines.
they are included in this chapter. One line had wrinkled, yellow seeds. Because Mendel had
To begin, we examine the analytical procedures that no concept of the chromosomal location of genes, we must
pertain to the independent assortment of nuclear genes. initially use the dot representation to write the combined
These were first developed by the father of genetics, Gregor genotype as r /r ⋅ Y /Y . The other line had round, green seeds,
Mendel. So, again, we turn to his work as a prototypic with genotype R /R ⋅ y /y. When these two lines were crossed,
example. they must have produced gametes that were r ⋅ Y and R ⋅ y,
respectively. Hence, the F1 seeds had to be dihybrid, of gen-
otype R /r ⋅ Y /y. Mendel discovered that the F1 seeds were
round and yellow. This result showed that the dominance of
R over r and of Y over y was unaffected by the condition of
3.1 MENDEL’S LAW OF the other gene pair in the R /r ⋅ Y /y dihybrid. In other words,
INDEPENDENT ASSORTMENT
LO 3.1 Using standard genetic symbolism, diagram how
Round and wrinkled phenotypes
a dihybrid organism can be constructed starting
from two pure parental lines; use the F1 dihybrid in a
self- and testcross to show the expected phenotype
frequencies that would result if the two genes are
assorting independently.
R remained dominant over r, regardless of seed color, and Y (101 + 32). This result is close to a 3 : 1 ratio (actually
remained dominant over y, regardless of seed shape. 3.2 : 1). Next, in regard to seed color, there are 416 yellow
Next, Mendel selfed the dihybrid F1 to obtain the F2 seeds (315 + 101) and 140 green seeds (108 + 32), almost
generation. The F2 seeds were of four different types in the exactly a 3 : 1 ratio. The presence of these two 3 : 1 ratios
following proportions: hidden in the 9 : 3 : 3 : 1 ratio was undoubtedly a source of
9 the insight that Mendel needed to explain the 9 : 3 : 3 : 1
16
round, yellow
ratio, because he realized that it was simply two different
3
16
round, green 3 : 1 ratios combined at random. One way of visualizing the
3
wrinkled, yellow random combination of these two ratios is with a branch
16
diagram, as follows:
1 wrinkled, green
16
3
4
of these round seeds
The result is illustrated in Figure 3-3 with the actual
3 will be yellow
numbers obtained by Mendel. This initially unexpected 4 of the F2 is round
9 : 3 : 3 : 1 ratio for these two characters seems a lot more 1
will be green
4
complex than the simple 3 : 1 ratios of the monohybrid
3
crosses. Nevertheless, the 9 : 3 : 3 : 1 ratio proved to be a 4
of these wrinkled seeds
consistent inheritance pattern in peas. As evidence, Mendel 1 will be yellow
4 of the F2 is wrinkled
also made dihybrid crosses that included several other com- 1
binations of characters and found that all of the dihybrid F1 4 will be green
individuals produced 9 : 3 : 3 : 1 ratios in the F2. This charac-
teristic ratio was another inheritance pattern that required The probabilities of the four possible outcomes are cal-
the development of a new idea to explain it. culated by using the product rule, to which we were intro-
First, let’s check the actual numbers obtained by M endel duced in Chapter 2 (the probability of two independent
in Figure 3-3 to determine if the monohybrid 3 : 1 ratios events occurring together is the product of their individual
can still be found in the F2 . In regard to seed shape, there probabilities). Hence, we multiply along the branches in the
are 423 round seeds (315 + 108) and 133 wrinkled seeds diagram. For example, 3/4 of all seeds will be round, and
3/4 of the round seeds will be yellow, so the probability of a
seed being both round and yellow is calculated as 3/4 × 3/4,
Mendel’s breeding program that produced which equals 9/16. These multiplications give the following
a 9 : 3 : 3 : 1 ratio four proportions:
3
4
× 43 = 9
16
round, yellow
3 3
P R /R • y/y r /r • Y/Y 4
× 41 = 16
round, green
(round, green) (wrinkled, yellow)
1
4
× 43 = 3
16
wrinkled, yellow
1
4
× 41 = 1
16
wrinkled, green
Gametes R •y r •Y These proportions constitute the 9 : 3 : 3 : 1 ratio that
we are trying to explain. However, is this exercise not
F1 merely number juggling? What could the combination of
R/r • Y/y the two 3 : 1 ratios mean biologically? The way that Mendel
(round, yellow) phrased his explanation does in fact amount to a biological
mechanism. In what is now known as the law of indepen-
dent assortment (Mendel’s second law), he concluded that
F1 F1 different gene pairs assort independently during gamete for
mation. The consequence is that, for two heterozygous gene
pairs A/a and B/b, the b allele is just as likely to end up in
F2 315 round, yellow 9 Ratio a gamete with an a allele as with an A allele, and likewise
for the B allele. In hindsight, we now know that, for the
108 round, green 3 most part, this “law” applies to genes on different chromo-
somes. Genes on the same chromosome generally do not
101 wrinkled, yellow 3
assort independently because they are held together by the
chromosome itself.
32 wrinkled, green 1
556 seeds 16 KEY CONCEPT Mendel’s second law (the law of indepen-
dent assortment) states that the alleles of gene pairs on differ-
FIGURE 3-3 Mendel created a dihybrid that, when selfed, produced ent chromosome pairs assort independently at meiosis.
F2 progeny in the ratio 9 : 3 : 3 : 1.
KEY CONCEPT The 9 : 3 : 3 : 1 phenotypic ratio observed in Punnett square illustrating the genotypes
the progeny of a dihybrid self results from two 3 : 1 ratios com- underlying a 9 : 3 : 3 : 1 ratio
bining at random and is diagnostic of independent assortment
of the two genes.
P R /R ; y/y r /r ; Y/Y
(round, green) (wrinkled, yellow)
1
2 of these r gametes will be Y
1
2 of the gametes will be r R /r ; Y/y R /r ; y/y r/r ; y/y r/r ; Y/ y
r ;y 1 1 1 1
1 16 16 16 16
2 will be y 1
4
contains a variety of genotypes, but there are only four phe- As an example of the universal applicability of the
notypes, and their proportions are in the 9 : 3 : 3 : 1 ratio. So principle of independent assortment, we can examine its
we see that, when we calculate progeny frequencies directly action in haploids. If the principle of equal segregation is
through gamete frequencies, we still arrive at the 9 : 3 : 3 : 1 valid across the board, then we should be able to observe
ratio. Hence, Mendel’s laws explain not only the F2 pheno- it in haploids, given that haploids undergo meiosis. Indeed,
types, but also the genotypes of gametes and progeny that independent assortment can be observed in a cross of the
underly the F2 phenotypic ratio. type A ; B × a ; b. Fusion of parental cells results in a tran-
sient diploid meiocyte that is a dihybrid A/a ; B/b, and the
KEY CONCEPT Both male and female dihybrids exhibiting randomly sampled products of meiosis (sexual spores such
independent assortment will generate four types of gametes, in as ascospores in fungi) will be
a 1 : 1 : 1 : 1 ratio. The Punnett square is a useful tool for graph- 1
A;B
4
ically displaying the genotypic and phenotypic outcomes of
1
their random union. 4
A;b
1
4
a;B
Mendel went on to test his principle of independent assort- 1
a;b
4
ment, experimentally confirming it in a number of ways. The
most direct way focused on the 1 : 1 : 1 : 1 gametic ratio hypoth- Hence, we see the same ratio as in the dihybrid testcross
esized to be produced by the F1 dihybrid R/r ; Y/y. This ratio in a diploid organism; again, the ratio is a random combi-
sprang directly from his principle of independent assortment nation of two monohybrid 1 : 1 ratios because of indepen-
and was the biological basis of the 9 : 3 : 3 : 1 ratio in the F2, as dent assortment.
shown by the Punnett square. To verify the 1 : 1 : 1 : 1 gametic
ratio, Mendel testcrossed the F1 dihybrid with a tester of gen- KEY CONCEPT In haploids, the genotypes of the products of
otype r/r ; y/y, which produces only gametes with recessive meiosis (spores) are directly observable in the haploid organ-
isms that arise from each spore through mitotic cell division. A
alleles (genotype r ; y). He reasoned that, if there were in fact
ratio of 1 : 1 : 1 : 1 is diagnostic of independent assortment in a
a 1 : 1 : 1 : 1 ratio of R ; Y, R ; y, r ; Y, and r ; y gametes, the prog- dihybrid meiocyte.
eny proportions of this cross should directly correspond to
the gametic proportions produced by the dihybrid; in other
words,
We should pause to compare Mendel’s work on one-
1
4
R /r ; Y /y → round, yellow and two-gene systems. In working with several genes one
1 at a time, he was able to demonstrate the principle of equal
4
R /r ; y /y → round, green
segregation of alleles of each gene at meiosis. In two-gene
1
4
r /r ; Y /y → wrinkled, yellow systems, he showed that for the genes at his disposal, the
1
r /r ; y /y → wrinkled, green equal segregation principle for each gene was taking place
4
independently of the other genes.
These proportions were the result that he obtained, per-
fectly consistent with his expectations. He obtained similar
results for all the other dihybrid crosses that he made, and
these tests and other types of tests all showed that he had, 3.2 WORKING WITH
in fact, devised a robust model to explain the inheritance
patterns observed in his various pea crosses.
INDEPENDENT ASSORTMENT
KEY CONCEPT In a testcross, the ratio of phenotypes in LO 3.2 In crosses involving independently assorting
the progeny reflects the gametic genotypes of the nontester dihybrids, predict the genotypic ratios in meiotic
parent. For a dihybrid exhibiting independent assortment, this products, genotypic ratios in progeny, and
ratio is 1 : 1 : 1 : 1. The testcross is a useful tool for diploids, phenotypic ratios in progeny.
where the phenotypes and genotypes of gametes cannot be
observed directly.
LO 3.3 Use chi-square analysis to test whether observed
phenotypic ratios are an acceptable fit to those
predicted by independent assortment.
In the early 1900s, both of Mendel’s laws were tested
in a wide spectrum of eukaryotic organisms. The results LO 3.4 In diploids, outline how pure lines homozygous for
of these tests showed that Mendelian principles were gen- two or more gene mutations can be created starting
erally applicable. Mendelian ratios (such as 3 : 1, 1 : 1, from parental lines homozygous for single-gene
9 : 3 : 3 : 1, and 1 : 1 : 1 : 1) were extensively reported, sug- mutations.
gesting that equal segregation and independent assortment
are fundamental hereditary processes found throughout In this section, we will examine several analytical proce-
nature. Mendel’s laws are not merely laws about peas; dures that are part of everyday genetic research and are
they are laws about the genetics of eukaryotic organisms all based on the concept of independent assortment. These
in general. procedures are all used to analyze phenotypic ratios.
Predicting progeny ratios two dice follow the product rule because the outcome on
one die is independent of the other. As an example, let us
Genetics can work in either of two directions: (1) predict-
calculate the probability, p, of rolling a pair of 4’s. The
ing the unknown genotypes of parents by using phenotype
probability of a 4 on one die is 1/6 because the die has six
ratios of progeny, or (2) predicting progeny phenotype
sides and only one side carries the number 4. This probabil-
ratios from parents of known genotype. The latter is an
ity is written as follows:
important part of genetics concerned with predicting the
types of progeny that emerge from a cross and calculating p ( rolling one 4) = 1
6
their expected frequencies—in other words, their probabil-
Therefore, with the use of the product rule, the prob-
ities. This is useful not only in research on model organ-
ability of a 4 appearing on both dice is 1/6 × 1/6 = 1/36,
isms but also in predicting outcomes of matings in human
which is written:
genetics; for example, in genetic counseling, people appre-
ciate specific risk estimates. We have already examined p ( rolling two 4’s) = 61 × 61 = 1
36
two methods for prediction: Punnett squares and branch
diagrams. Punnett squares can be used to show hereditary KEY CONCEPT The product rule states that the probability of
patterns based on one gene pair, two gene pairs, or more. independent events occurring together is the product of their
Such grids are good graphic devices for representing prog- individual probabilities.
eny, but drawing them can be time consuming. Even the
16-compartment Punnett square that we used to analyze Now, we turn to the sum rule. The sum rule states that
a dihybrid cross takes a long time to write out; but for a the probability of either one or the other of two mutually
trihybrid cross, there are 23, or 8, different gamete types, exclusive events occurring is the sum of their individual
and the Punnett square has 64 compartments. The branch probabilities. Dice can also be used to illustrate the sum
diagram (shown below) is easier to create and is adaptable rule. We have already calculated that the probability of
for phenotypic, genotypic, or gametic proportions, as illus- two 4’s is 1/36; clearly, with the use of the same type of
trated for the dihybrid A/a ; B/b. calculation, the probability of two 5’s will be the same, or
Progeny Progeny 1/36. Now, we can calculate the probability of either two
genotypes phenotypes 4’s or two 5’s. Because these outcomes are mutually exclu-
from a self from a self Gametes sive, the sum rule can be used to tell us that the answer is
1/36 + 1/36, which is 1/18. This probability can be written
3 1
1
B /B B/− B as follows:
› 4 ›4 › 2
1
A /A ›
1
2
B /b 3
4
A /− 1
2
A p ( rolling two 4’s or rolling two 5’s) = 1
36
1
+ 36 = 1
18
4
› › ›
1 1 1
4
b /b 4
b /b 2
b
KEY CONCEPT The sum rule states that the probability of
1
B /B 3
B /− 1
B either one or the other of two mutually exclusive events occur-
4 4 › 2
› › ring is the sum of their individual probabilities.
1 1 1 1
2
A/a › 2
B/b 4
a /a 2
a
› › ›
1
b /b 1
b /b 1
b KEY CONCEPT The product rule is used to determine the
4 4 2
probability of observing both outcome A and outcome B. The
1
4
B /B sum rule is used to determine the probability of observing
› either outcome A or outcome B.
1 1
4
a /a › 2
B /b
›
1
b /b What proportion of progeny will be of a specific
4
g enotype? Now we can turn to a genetic example.
Note, however, that the “tree” of branches for genotypes Assume that we have two plants of genotypes
is quite unwieldy even in this simple case, which uses two
A /a ; b /b ; C /c ; D /d ; E /e
gene pairs, because there are 32 = 9 genotypes. For three
gene pairs, there are 33 , or 27, possible genotypes. To sim- and
plify this problem, we can use a statistical approach, which
A /a ; B /b ; C /c ; d /d ; E /e
constitutes a third method for calculating the probabilities
(expected frequencies) of specific phenotypes or genotypes From a cross between these plants, we want to recover
coming from a cross. The two statistical rules needed are a progeny plant of genotype a /a ; b /b ; c /c ; d /d ; e /e (perhaps
the product rule (introduced earlier, and also in Chapter 2) for the purpose of acting as the tester strain in a testcross).
and the sum rule, which we will now consider together. What proportion of the progeny should we expect to be of
The product rule states that the probability of indepen- that genotype? If we assume that all the gene pairs assort
dent events occurring together is the product of their indi- independently, then we can do this calculation easily by
vidual probabilities. The possible outcomes from rolling using the product rule. The five different gene pairs are
considered individually, as if five separate crosses, and then one success. So to satisfy the 95 percent confidence level,
the individual probabilities of obtaining each genotype are we must put this expression equal to 0.95 (the equivalent
multiplied together to arrive at the answer: of 95 percent).
Therefore,
From A /a × A /a, one-fourth of the progeny will be a/a.
From b /b × B /b, one-half of the progeny will be b/b. 1 − (255/256)n = 0.95
From C /c × C /c, one-fourth of the progeny will be c/c. Solving this equation for n gives us a value of 765, the
From D /d × d /d , one-half of the progeny will be d/d. number of progeny needed to virtually guarantee success.
Notice how different this number is from the naïve expec-
From E /e × E /e, one-fourth of the progeny will be e/e.
tation of success in 256 progeny. This type of calculation is
Therefore, the overall probability (or expected frequency) useful in many applications in genetics and in other situa-
of obtaining progeny of genotype a /a ; b /b ; c /c ; d /d ; e /e will tions in which a successful outcome is needed from many
be 1/4 × 1/2 × 1/4 × 1/2 × 1/4 = 1/256. This probability calcu- trials.
lation can be extended to predict phenotypic frequencies or
gametic frequencies. Indeed, there are many other uses for KEY CONCEPT To calculate the progeny sample size needed
this method in genetic analysis, and we will encounter some to be 95 percent certain of obtaining at least one individual of
in later chapters. the desired genotype, start by calculating the probability of
no successes in a sample size of n. The resulting sample size
KEY CONCEPT For independently assorting genes, the prob-
is always much larger than one calculated using hypothetical
ability of a multigene genotype or phenotype can be obtained expectations.
by multiplying the probabilities of the genotype or phenotype
for each of the individual genes.
you obtain constitute the expected 1 : 1 ratio. If there is a resulting value, χ2, will provide a numerical value that
close match, then the hypothesis is deemed consistent with estimates the degree of agreement between the expected
the result, whereas if there is a poor match, the hypothe- (hypothesized) and observed (actual) results, with the num-
sis is rejected. As part of this process, a judgment has to ber growing larger as the agreement increases.
be made about whether the observed numbers are close The calculation is most simply performed by using a
enough to those expected. Very close matches and blatant table:
mismatches generally present no problem, but, inevitably,
there are gray areas in which the match is not obvious. Class O E (O − E2 ) (O − E2 )/E
The χ2 test is simply a way of quantifying the various
Red 55 60 25 25/60 = 0.42
deviations expected by chance if a hypothesis is true. Take
the preceding simple hypothesis predicting a 1 : 1 ratio, for White 65 60 25 25/60 = 0.42
example. Even if the hypothesis were true, we can only Total = χ2 = 0.84
rarely expect an exact 1 : 1 ratio. We can model this idea
with a barrelful of equal numbers of red and white marbles. Now we must look up this χ2 value in Table 3-1, which
If we blindly remove samples of 100 marbles, on the basis will give us the probability value (p) we seek. The rows in
of chance we would expect samples to show small devia- Table 3-1 list different values of degrees of freedom (df).
tions such as 52 red : 48 white quite commonly and to show The number of degrees of freedom is the number of inde-
larger deviations such as 60 red : 40 white less commonly. pendent variables in the data. In the present context, the
Even 100 red marbles is a possible outcome, at a very low number of independent variables is simply the number of
probability of (1/2)100 . However, if any result is possible at phenotypic classes minus 1. In this case, df = 2 − 1 = 1. So
some level of probability even if the hypothesis is true, how we look only at the 1 df line. We see that our χ2 value of
can we ever reject a hypothesis? A general scientific con- 0.84 lies somewhere between the columns marked 0.5 and
vention is that a hypothesis will be rejected as false if there 0.1—in other words, a probability of between 50 percent
is a probability of less than 5 percent of observing a devia- and 10 percent. This probability value is much greater than
tion from expectations at least as large as the one actually the cutoff value of 5 percent, and so we accept the observed
observed. The hypothesis might still be true, but we have results as being compatible with the hypothesis.
to make a decision somewhere, and 5 percent is the con- Some important notes on the application of this test
ventional decision line. The implication is that, although follow:
results this far from expectations are expected 5 percent of
1. What does the probability value actually mean? It is the
the time even when the hypothesis is true, we will mistak-
probability of observing a deviation from the expected
enly reject the hypothesis in only 5 percent of cases, and we
results at least as large (not exactly this deviation) on the
are willing to take this chance of error. (This 5 percent is
basis of chance if the hypothesis is correct.
the converse of the 95 percent confidence level used earlier.)
2. The fact that our results have “passed” the chi-square
2
KEY CONCEPT The χ test quantifies the probability of vari- test because p > 0.05, does not mean that the hypothe-
ous deviations expected by chance if a hypothesis is true. It is sis is true; it merely means that the results are compati-
used to decide whether or not an observed experimental devi- ble with that hypothesis. However, if we had obtained a
ation is reasonably compatible with a working hypothesis. value of p < 0.05, we would have been forced to reject
the hypothesis. Science is all about falsifiable hypothe-
Let’s look at some real data. We will test our earlier ses, not “truth.”
hypothesis that a plant is a heterozygote. We will let A stand 3. We must be careful about the wording of the hypothe-
for red petals and a stand for white. Scientists test a hypoth- sis because tacit assumptions are often buried within it.
esis by making predictions based on the hypothesis. In the The present hypothesis is a case in point; if we were to
present situation, one possibility is to predict the results of a state it carefully, we would have to say that the “indi-
testcross. Assume that we testcross the presumed heterozy- vidual under test is a heterozygote A/a, these alleles
gote. On the basis of the hypothesis, Mendel’s law of equal show equal segregation at meiosis, and the A/a and a/a
segregation predicts that we should have 50 percent A/a and progeny are of equal viability.” We will investigate allele
50 percent a/a. Assume that, in reality, we obtain 120 prog- effects on viability in Chapter 5, but, for the time being,
eny and find that 55 are red and 65 are white. These num- we must keep them in mind as a possible complication
bers differ from the precise expectations, which would have because differences in survival would affect the sizes
been 60 red and 60 white. The result seems a bit far off the of the various classes. The problem is that, if we reject
expected ratio, which raises uncertainty; so we need to use a hypothesis that has hidden components, we do not
the χ2 test. We calculate χ2 by using the following formula: know which of the components we are rejecting. For
example, in the present case, if we were forced to reject
χ2 = ∑(O − E)2 /E for all classes
the hypothesis as a result of the χ2 test, we would not
in which E is the expected number in a class, O is the know if we were rejecting equal segregation or equal
observed number in a class, and Σ means “sum of.” The viability, or both.
4. The outcome of the χ2 test depends heavily on sample Pure lines of plants or animals are made through
sizes (numbers in the classes). Hence, the test must use repeated generations of selfing. (In animals, selfing is
actual numbers, not proportions or percentages. Addi- accomplished by mating animals of identical genotype.)
tionally, the larger the samples, the more powerful is the Selfing a monohybrid plant shows the principle at work.
test. Suppose we start with a population of individuals that
are all A/a and allow them to self. We can apply Mendel’s
Any of the familiar Mendelian ratios considered in this first law to predict that, in the next generation, there will
chapter or in Chapter 2 can be tested by using the χ2 test— be 41 A/A, 12 A/a, and 41 a/a. Note that the heterozygosity
for example, 3 : 1 (1 df), 1 : 2 : 1 (2 df), 9 : 3 : 3 : 1 (3 df), and (the proportion of heterozygotes) has halved, from 1 to 12 .
1 : 1 : 1 : 1 (3 df). We will return to more applications of the If we repeat this process of selfing for another generation,
χ2 test in Chapter 4. all descendants of homozygotes will be homozygous, but,
again, the heterozygotes will halve their proportion to a
KEY CONCEPT In genetics, the χ 2 test is commonly used quarter. The process is shown in the following display:
to assess whether or not the number of observed individuals
with certain phenotypes are an acceptable fit to an expected All A/a
Mendelian ratio.
1 1 1
4
A/A 2
A/a 4
a/a
Synthesizing pure lines
Pure lines are among the essential tools of genetics. For 1
A/A 1
A/A 14 A/a 1
a/a 1
a/a
4 8 8 4
one thing, only these fully homozygous lines will express
recessive alleles, but the main need for pure lines is in the After, say, eight generations of selfing, examination of
maintenance of stocks for research. The members of a pure one gene pair reveals the proportion of heterozygotes in
8
line can be left to interbreed over time and thereby act as the progeny is reduced to (1/2) , which is 1/256, or about
a constant source of the genotype for use in experiments. 0.4 percent. Let’s look at this probability value in a slightly
Hence, for most model organisms, there are international different way. Assume that we start such a program with
stock centers that are repositories of pure lines for use in a genotype that is heterozygous at 256 gene pairs. If we
research. Similar stock centers provide lines of plants and also assume independent assortment, then, after selfing for
animals for use in agriculture. eight generations, we would end up with an average of only
one heterozygous gene (that is, 1/256), and the rest will be
KEY CONCEPT Homozygous pure lines are important homozygous. In other words, we are well on our way to
research tools that allow geneticists to maintain a source of creating a pure line.
any given genotype. Recessive alleles can only be expressed
Let’s apply this principle to the selection of agricul-
in pure lines.
tural lines, the topic with which we began the chapter.
We can use as our example the selection of Marquis wheat selections that eventually led to a pure line that had the com-
by Charles Saunders in the early part of the twentieth cen- bination of favorable properties needed—good-quality grain
tury. Saunders’s goal was to develop a productive wheat and early maturation. This line was called Marquis. It was
line that would have a shorter growing season and hence rapidly adopted in many parts of the world.
open up large areas of terrain in northern countries such as A similar approach can be applied to the rice lines with
Canada and Russia for growing wheat, another of the which we began the chapter. All the single-gene mutations
world’s staple foods. He crossed a line having excellent grain are crossed in pairs, and then their F1 plants are selfed or
quality called Red Fife with a line called Hard Red Calcutta, intercrossed with other F1 plants. As a demonstration, let’s
which, although its yield and quality were poor, matured consider just four mutations, 1 through 4. A breeding pro-
20 days earlier than Red Fife. The F1 produced by the cross gram might be as follows, in which the mutant alleles and
was presumably heterozygous for multiple genes controlling their wild-type counterparts are always listed in the same
the wheat qualities. From this F1, Saunders made selfings and order ( recall that the + sign designates wild type):
1/1 ; +/+ ; +/+ ; +/+ 3 +/+ ; 2/2 ; +/+ ; +/+ +/+ ; +/+ ; 3/3 ; +/+ 3 +/+ ; +/+ ; +/+ ; 4/4
Self Self
Select the homozygote 1/1 ; 2/2 ; +/+ ; +/+ Select the homozygote +/+ ; +/+ ; 3/3 ; 4/4
Self
This type of breeding has been applied to many other KEY CONCEPT Pure lines are generated through repeated
crop species. The colorful and diverse pure lines of toma- selfing, which reduces the proportion of heterozygotes by half
toes used in commerce are and results in an increased proportion of homozygotes with
shown in Figure 3-5. each generation.
Representatives of Note that, in general
many tomato lines when a multiple heterozy-
gote is selfed, a range of Hybrid vigor
different homozygotes is
produced. For example, We have been considering the synthesis of superior pure
from A/a ; B/b ; C/c, there lines for research and for agriculture. Pure lines are conve-
are two homozygotes for nient in that propagation of the genotype from year to year
each gene pair (that is, for is fairly easy. However, a large proportion of commercial
the first gene, the homozy- seed that farmers (and gardeners) use is called hybrid seed.
gotes are A/A and a/a), and Curiously, in many cases in which two disparate homozy-
so there are 23 = 8 different gous lines of plants (and animals) are united in an F1 hybrid
homozygotes possible: (presumed heterozygote), the hybrid shows greater size
and vigor than do the two contributing lines (Figure 3-6).
A /A ; B /B ; C /C a /a ; b /b ; c /c This general superiority of multiple heterozygotes is called
A /A ; b /b ; C /C a /a ; B /B ; c /c hybrid vigor. Seed companies must develop pure lines by
A /A ; B /B ; c /c a /a ; b /b ; C /C the assortment through selfing methods we saw previously,
and then cross them every season to generate hybrid seeds
FIGURE 3-5 Tomato breeding A /A ; b /b ; c /c a /a ; B /B ; C /C
for the commercial market.
has resulted in a wide range
of lines of different genotypes Each distinct homozy- The molecular reasons for hybrid vigor are mostly
and phenotypes. [barmalini/ gote can be the start of a unknown and still hotly debated, but the phenomenon is
Shutterstock.] new pure line. undeniable and has made large contributions to agriculture.
FIGURE 3-6 Multiple heterozygous hybrid flanked by the two pure lines crossed to make it. (a) The
plants. (b) Cobs from the same plants. [(a) Photo courtesy of Jun Cao, Schnable Laboratory, Iowa State
University; (b) Deana Namuth-Covert, PhD, Univ of Nebraska, Plant and Soil Sciences eLibrary (http://
passel.unl.edu) hosted at the University of Nebraska, Institute of Agriculture and Natural Resources.]
A negative aspect of using hybrids is that, every season, the 3.3 THE CHROMOSOMAL
two parental lines must be grown separately and then inter-
crossed to make hybrid seed for sale. This process is much
BASIS OF INDEPENDENT
more inconvenient than maintaining pure lines, which ASSORTMENT
requires only letting plants self; consequently, hybrid seed is
more expensive than seed from pure lines. LO 3.5 Explain two-gene independent assortment ratios
From the user’s perspective, there is another negative in terms of chromosome behavior at meiosis, in
aspect of using hybrids. After a hybrid plant has grown and haploids and in diploids.
produced its crop for sale, it is not realistic to keep some of the
seeds that it produces and expect this seed to be equally vigor- Like equal segregation, the independent assortment of gene
ous the next year. The reason is that, when the hybrid under- pairs on different chromosomes is explained by the behav-
goes meiosis, independent assortment of the various mixed ior of chromosomes during meiosis. Consider a chromo-
gene pairs will form many different allelic combinations, and some that we might call number 1; its two homologs could
very few of these combinations will be that of the original be named 1′ and 1″. If the chromosomes pair and align on
hybrid. For example, a tetrahybrid, when selfed, produces 81 either side of the equator, then 1′ might go “north” and 1″
different genotypes, of which only a minority will be tetrahy- “south,” or vice versa. Similarly, for a chromosome 2 with
brid. If we assume independent assortment, then, for each gene homologs 2′ and 2″, 2′ might go north and 2″ south, or vice
pair, selfing will produce one-half heterozygotes A /a → 41 A/A, versa. Hence, chromosome 1′ could end up packaged with
1
A /a , and 41 a/a. Because there are four gene pairs in this tet- either chromosome 2′ or 2″ , depending on which chromo-
2
rahybrid, the proportion of progeny that will be like the origi- somes were pulled in the same direction.
nal hybrid A/a ; B/b ; C/c ; D/d will be (1/2)4 = 1/16. Independent assortment is not easy to demonstrate by
observing segregating chromosomes under the microscope
KEY CONCEPT Some hybrids between genetically different because homologs such as 1′ and 1″ do not usually look
pure lines show hybrid vigor. However, gene assortment when different, although they might carry minor sequence varia-
the hybrid undergoes meiosis breaks up the favorable allelic tion. However, independent assortment can be observed in
combination, and thus few members of the next generation certain specialized cases. One case was instrumental in the
have it. historical development of the chromosome theory.
Different chromosomes segregate independently pulled together into one cell, and the a/a and b/b are pulled
into the other cell. In the other case, the alleles A/A and b/b
are united in the same cell and the alleles a/a and B/B are
united in the other cell. The two patterns result from two
equally frequent spindle attachments to the centromeres
in the first anaphase. Meiosis then produces four cells of
= the indicated genotypes from each of these segregation
patterns. Because segregation patterns 4 and 4′ are equally
common, the meiotic product cells of genotypes A ; B, a ; b,
A ; b, and a ; B are produced in equal frequencies. In other
words, the frequency of each of the four genotypes is 1/4.
FIGURE 3-7 Carothers observed these two equally frequent patterns This gametic distribution is that postulated by Mendel for
by which a heteromorphic pair (brown) and an unpaired chromosome
a dihybrid, and it is the one that we inserted along one edge
(purple) move into gametes at meiosis.
of the Punnett square in Figure 3-4. The random fusion of
these gametes results in the 9 : 3 : 3 : 1 F2 phenotypic ratio.
In 1913, Elinor Carothers found an unusual chromo-
KEY CONCEPT The mechanical basis of equal segregation
somal situation in a certain species of grasshopper—a
and independent assortment of alleles is the anaphase
situation that permitted a direct test of whether different
segregation of chromosomes at meiosis. Segregation of a pair
chromosome pairs do indeed segregate independently. of homologs by spindle attachment from each pole accounts
Studying meioses in the testes of grasshoppers, she found for Mendel’s first law. The randomness of spindle attachment
a grasshopper in which one chromosome “pair” had non- throughout the chromosome set accounts for Mendel’s second
identical members. Such a pair is called a heteromorphic law.
pair; presumably, the chromosomes show only partial
homology. In addition, the same grasshopper had another
chromosome (unrelated to the heteromorphic pair) that
had no pairing partner at all. Carothers was able to use Independent assortment in haploid
these unusual chromosomes as visible cytological markers organisms
of the behavior of chromosomes during meiosis. She visu-
ally screened many meioses and found that there were two In the ascomycete fungi, we can actually inspect the prod-
distinct patterns, which are shown in Figure 3-7. In addition, ucts of a single meiocyte to show independent assortment
she found that the two patterns were equally frequent. To directly. Let’s use the filamentous fungus Neurospora crassa
summarize, if we hold the segregation of the heteromorphic to illustrate this point. As we have seen from earlier fun-
pair constant (brown in the figure), then the unpaired (pur- gal examples in Chapter 2, a cross in Neurospora is made
ple) chromosome was found to go to either pole equally by mixing two parental haploid strains of opposite mating
frequently, half the time with the long form and half the type. In a manner similar to that of yeast, mating type is
time with the short form. In other words, the purple and determined by two “alleles” of one gene—in this species,
brown sets were segregating independently. Although these called MAT-A and MAT-a.
are obviously not typical chromosomes, Carothers’s results The products of meiosis in fungi are sexual spores.
do strongly suggest that different chromosomes assort inde- Recall that the ascomycetes (which include Neurospora and
pendently at the first division of meiosis. Saccharomyces) are unique in that, for any given meiocyte,
the spores are held together in a membranous sac called an
ascus. Thus, for these organisms, the products of a single
Independent assortment in diploid meiosis can be recovered and tested. In the orange bread
organisms mold Neurospora, the nuclear divisions of meioses I and
The chromosomal basis of the law of independent assort- II take place along the linear axis of the ascus and do not
ment is formally diagrammed in Figure 3-8, which illustrates overlap, and so the four products of a single meiocyte lie in
how the separate behavior of two different chromosome a straight row (Figure 3-9a). Furthermore, for some reason
pairs gives rise to the 1 : 1 : 1 : 1 Mendelian ratios of gametic not understood, there is a postmeiotic mitosis, which also
types expected from independent assortment. The hypo- shows no spindle overlap. Hence, meiosis and the extra
thetical cell has four chromosomes: a pair of homologous mitosis result in a linear ascus containing eight ascospores,
long chromosomes (yellow) and a pair of homologous or an octad. In a heterozygous meiocyte A/a, if there are
short ones (blue). The genotype of the meiocytes is A/a ; B/b, no crossovers between the gene and its centromere (a pos-
and the two allelic pairs, A/a and B/b, are shown on two sibility we will explore in Chapter 4), then there will be
different chromosome pairs. Parts 4 and 4′ of Figure 3-8 two adjacent blocks of ascospores, four of A and four of a
show the key step in independent assortment: there are two (Figure 3-9b).
equally frequent allelic segregation patterns, one shown in 4 Now we can examine a dihybrid. Let’s make a cross
and the other in 4′. In one case, the A/A and B/B alleles are between two distinct mutants having mutations in different
Prophase. Chromosomes A
and centromeres have a
replicated, but centromeres a
have not split. b b
B
B
2
A B
Prophase. A B
Homologs synapse. a b
a b
3
A B The other, A b
Anaphase. B equally b
A A
Centromeres attach to a b frequent, a B
spindle and are pulled alignment
to poles of cell. a b a B
4 4´
A B A b
Telophase. A B A b
Two cells form. a b a B
a b a B
5 5´
A B A b
Second anaphase.
New spindles form, A B A b
and centromeres a b a B
finally divide.
6 a b 6´ a B
A B A b
1 1
4 4
A B A b
End of meiosis.
Four cells produced
from each meiosis. a b a B
1 1
4 4
a b a B
7 7´
genes on different chromosomes. By assuming that the loci flat, spreading colony of wild type (b+ ). We will assume that
of the mutated genes are both very close to their respective the two mutants are of opposite mating type. Hence, the
centromeres, we avoid complications due to crossing over cross is
between the loci and the centromeres (again, we will see
a ; b+ × a+ ; b
examples of such scenarios in Chapter 4). The first mutant
is albino (a), contrasting with the normal pink wild type Because of random spindle attachment, the two octad
(a+ ). The second mutant is biscuit (b), which has a very types will be produced with equal frequency. (Inspect
compact colony shaped like a biscuit in contrast with the Figure 3-8, which shows the mechanisms behind this result.)
Recombination
Ascus The independent assortment of genes at meiosis is one
First Second Postmeiotic Development of the main ways by which an organism produces new
meiotic meiotic mitotic of sexual combinations of alleles. The production of new allele
division division division spores combinations is formally called recombination.
(ascospores)
around There is general agreement that the evolutionary
nuclei advantage of producing new combinations of alleles is
that it provides variation as the raw material for nat-
ural selection. Recombination is a crucial principle in
(b) Allele segregation Octad genetics, partly because of its relevance to evolution
but also because of its use in genetic analysis. It is par-
ticularly useful for analyzing inheritance patterns of
A
multigene genotypes. In this section, we define recom-
Tetrad bination in such a way that we would recognize it in
A experimental results, and we lay out the way in which
Meiocyte after A
recombination is analyzed and interpreted.
chromatid
formation A A Recombination is observed in a variety of biological
situations, but, for the present, we define it in relation to
A
A A meiosis.
A
A eiotic recombination is any meiotic process
M
a a that generates a haploid product with new
a a combinations of the alleles carried by the
a a haploid genotypes that united to form the
a meiocyte.
a a This seemingly wordy definition is actually quite
First simple; it makes the important point that we detect
meiotic
division Second
a recombination by comparing the inputs into meiosis
meiotic with the outputs (Figure 3-10). The inputs are the two
division Mitosis haploid genotypes that combine to form the meiocyte,
the diploid cell that undergoes meiosis. For humans,
the inputs are the parental egg and sperm. They unite
FIGURE 3-9 Neurospora is an ideal model system for studying
allelic segregation at meiosis. (a) The four products of meiosis (tetrad) to form a diploid zygote, which divides to yield all the
undergo mitosis to produce an octad. The products are contained body cells, including the meiocytes that are set aside
within an ascus. (b) An A/a meiocyte undergoes meiosis followed within the gonads. The output genotypes are the hap-
by mitosis, resulting in equal numbers of A and a products and loid products of meiosis. In humans, these haploid
demonstrating the principle of equal segregation.
products are a person’s own eggs or sperm. Any mei-
otic product that has a new combination of the alleles
provided by the two input genotypes is by definition a
recombinant.
Recombinants are meiotic output KEY CONCEPT Meiosis generates recombinants, which are hap-
different from meiotic input loid meiotic products with new combinations of the alleles carried
by the haploid genotypes that united to form the meiocyte.
n n
Input A .B a .b
First, let us look at how recombinants are detected exper-
imentally. The detection of recombinants in organisms with
2n
haploid life cycles such as fungi or algae is straightforward. The
Meiotic A / a B/ b . input and output types in haploid life cycles are the genotypes
diploid of individuals rather than gametes and may thus be inferred
directly from phenotypes. Figure 3-10 can be viewed as sum-
marizing the simple detection of recombinants in organisms
Meiosis with haploid life cycles. Detecting recombinants in organisms
with diploid life cycles is trickier. The input and output types
n A .B Parental (input) type in diploid cycles are gametes. Thus, we must know the geno-
Output types of both input and output gametes to detect recombinants
n a .b Parental (input) type
in an organism with a diploid cycle. Though we cannot detect
the genotypes of input or output gametes directly, we can infer
these genotypes by using the appropriate techniques:
n A .b Recombinant
• To know the input gametes, we use pure-breeding diploid
n a .B Recombinant
parents because they can produce only one gametic type.
• To detect recombinant output gametes, we testcross the
FIGURE 3-10 Recombinants (blue) are those products of meiosis diploid individual and observe its progeny (Figure 3-11).
with allele combinations different from those of the haploid cells that
formed the meiotic diploid (yellow). Note that genes A/a and B/b
A testcross offspring that arises from a recombinant
are shown separated by a dot because they may be on the same product of meiosis also is called a recombinant. Notice,
chromosome or on different chromosomes. again, that the testcross allows us to concentrate on one
meiosis and prevent ambiguity (the recessive tester produces
only one type of gamete in meiosis and cannot generate
P A /A B/B . .
a /a b/ b
Input n A .B n a .b
2n 2n
Meiotic
diploid (F1)
.
A / a B/ b .
a /a b/ b
Tester
Meiosis Meiosis
Progeny (2n)
Output Parental-type
gamete
n
A .B + a .b n Fertilization
A /a . B/ b Parental type
Parental-type
gamete
n
a .b + a .b n
a /a . b/ b Parental type
Recombinant
gamete
n A .b + a .b n
A /a . b/ b Recombinant
Recombinant
gamete
n
a .B + a .b n
a /a . B/ b Recombinant
FIGURE 3-11 Recombinant products of a diploid meiosis are most readily detected in a cross of a
heterozygote and a recessive tester. Note that Figure 3-10 is repeated as part of this diagram.
recombinant gametes). From a self of the F1 in Figure 3-11, 50 percent of the total progeny. Hence, we see that independent
for example, a recombinant A /A ⋅ B /b offspring could not assortment at meiosis produces a recombinant frequency of
be distinguished from A /A ⋅ B /B without further crosses. 50 percent. If we observe a recombinant frequency of 50 per-
A central part of recombination analysis is recombinant cent in a testcross, we can infer that the two genes under study
frequency. One reason for focusing on recombinant frequency assort independently. The simplest and most likely interpreta-
is that its numerical value is a convenient test for whether tion of independent assortment is that the two genes are on sep-
two genes are on different chromosomes. Recombinants are arate chromosome pairs. (However, we must note that genes
produced by two different cellular processes: the independent that are very far apart on the same chromosome pair can assort
assortment of genes on different chromosomes (this chapter) virtually independently and produce the same result, as we will
and crossing over between genes on the same chromosome see in Chapter 4.)
(discussed in Chapter 4). The proportion of recombinants
is the key idea here because the diagnostic value can tell us KEY CONCEPT A recombinant frequency of 50 percent indi-
whether genes are on different chromosomes. We will deal cates that the genes are independently assorting and are most
with independent assortment here. likely on different chromosomes.
For genes on separate chromosomes, recombinants are
produced by independent assortment, as shown in Figure 3-12.
Again, we see the 1 : 1 : 1 : 1 ratio that we have seen before, but 3.4 POLYGENIC INHERITANCE
now the progeny of the testcross are classified as either recom-
binant or resembling the P (parental) input types. Set up in this LO 3.6 Extend the principle of independent assortment to
way, the proportion of recombinants is clearly 41 + 41 = 12 , or multiple genes that each contribute to a phenotype
showing continuous distribution.
Independent assortment produces So far, our analysis in this book has focused on single-gene
50 percent recombinants differences, with the use of sharply contrasting phenotypes
such as red versus white petals, smooth versus wrinkled
A B a b
seeds, and long- versus vestigial-winged Drosophila. How-
×
A B a b ever, much of the variation found in nature is continuous, in
P which a phenotype can take any measurable value between
two extremes. Height, weight, and color intensity are
A B a b examples of such metric, or quantitative, phenotypes (see
Figure 1-6 ). Typically, when the metric value is plotted
Gametes against frequency in a natural population, the distribution
A B a b
curve is shaped like a bell (Figure 3-13). The bell shape is due
a b × a b to the fact that average values in the middle are the most
common, whereas extreme values are rare. At first, it is dif-
Meiotic diploid (F1) (Tester)
ficult to see how continuous distributions can be influenced
A B by genes inherited in a Mendelian manner; after all, Men-
1 delian analysis is facilitated by using clearly distinguishable
Parental type
4
a b
a B
1
Recombinant
4
a b
categories. However, it soon became clear that the interac- genome; in many cases, they are on different chromosomes
tion of several genes affecting a metric trait could produce and show independent assortment.
a bell-shaped curve, a concept that became known as the
multifactorial hypothesis. In this section, we shall see how KEY CONCEPT Both environment and genotype can contrib-
the multifactorial hypothesis works. The topic is explored ute to continuous variation.
at length in Chapter 19.
Let’s see how the inheritance of several heterozygous
KEY CONCEPT Natural populations may show continuous polygenes (even as few as two) can generate a bell-shaped
variation of metric traits such as height or weight. Often the distribution curve. We can consider a simple model that was
distribution of measurements is in the form of a bell-shaped originally used to explain continuous variation in the degree
curve. of redness in wheat seeds. The work was done by Hermann
Nilsson-Ehle in the early twentieth century. We will assume
Of course, many cases of continuous variation have a two independently assorting gene pairs R1 /r1 and R2 /r2 . Both
purely environmental basis, little affected by genetics. For R1 and R2 contribute to wheat-seed redness. Each “dose” of
example, a population of genetically homozygous plants an R allele of either gene is additive, meaning that it increases
grown in a plot of ground often show a bell-shaped curve the degree of redness proportionately. An illustrative cross is
for height, with the smaller plants around the edges of the a self of a dihybrid R1 /r1 ; R2 /r2 . Both male and female gam-
plot and the larger plants in the middle. This variation can etes will show the genotypic proportions as follows:
be explained only by environmental factors such as moisture
and amount of fertilizer applied. However, many cases of con- R1 ; R2 2 doses of redness
tinuous variation do have a genetic basis. Human skin color R1 ; r2 1 dose of redness
is an example: all degrees of skin darkness can be observed in r1 ; R2 1 dose of redness
populations from different parts of the world, and this vari-
ation clearly has a genetic component. In such cases, from r1 ; r2 0 doses of redness
several to many alleles interact with a more or less additive Overall, in this gamete population, one-fourth have two
effect. The interacting genes underlying hereditary continu- doses, one-half have one dose, and one-fourth have zero
ous variation are called polygenes or quantitative trait loci doses. The union of male and female gametes both showing
(QTLs). (The term quantitative trait locus needs some defi- this array of R doses is illustrated in Figure 3-14. The num-
nition: quantitative is more or less synonymous with con- ber of doses in the progeny ranges from four (R1 /R1 ; R2 /R2 )
tinuous; trait is more or less synonymous with character or down to zero (r1 /r2 ; r2 /r2 ), with all values between.
property; locus, which literally means place on a chromo- The proportions in the grid of Figure 3-14 can be drawn
some, is more or less synonymous with gene.) The polygenes, as a histogram, as shown in Figure 3-15. The shape of the
or QTLs, for the same trait are distributed throughout the histogram can be thought of as a scaffold that could be the
1 2 4 2
2 16 16 16 3
4 4
0 doses 2 doses 1 dose 0 doses 2
1 1 2 1
4 16 16 16 1
1 1
0
0 1 2 3 4
4 doses 3 doses 2 doses 1 dose 0 doses Number of contributing
Overall in progeny 1 4 6 4 1 polygenic alleles, or “doses”
16 16 16 16 16
underlying basis for a bell-shaped distribution curve. When Chapter 19. Note that polygenes are not considered a special
this analysis of redness in wheat seeds was originally done, functional class of genes. They are identified as a group only
variation was found within each class that allegedly repre- in the sense that they have alleles that contribute to continu-
sented one polygene “dose” level. Presumably, this variation ous variation of a trait.
within a class is the result of environmental differences.
Hence, the environment can be seen to contribute in a way
that rounds off the sharp shoulders of the histogram bars,
3.5 ORGANELLE GENES:
resulting in a smooth bell-shaped curve (the red line in the INHERITANCE INDEPENDENT
histogram). If the number of polygenes is increased, the his- OF THE NUCLEUS
togram more closely approximates a smooth continuous
distribution. For example, for a characteristic determined LO 3.7 Apply the diagnostic criteria for assessing whether a
by three polygenes, the histogram is as shown in Figure 3-16. gene of interest resides on a nuclear chromosome or
on an organelle chromosome.
KEY CONCEPT The interaction of several additive heterozy-
gous genes can by themselves result in a bell-shaped curve, So far, we have considered how nuclear genes assort inde-
their alleles acting as metric “doses.” pendently by virtue of their loci on different chromosomes.
However, although the nucleus contains most of a eukary-
In our illustration, we used a dihybrid self to show how otic organism’s genes, a distinct and specialized subset of the
the histogram is produced. But how is our example relevant genome is found in the mitochondria, and, in plants, also in
to what is going on in natural populations? After all, not all the chloroplasts. These subsets are inherited independently of
crosses could be of this type. Nevertheless, if the alleles at the nuclear genome, and so they constitute a special case
each gene pair are approximately equal in frequency in the of independent inheritance, sometimes called extranuclear
population (for example, R1 is about as common as r1), then inheritance.
the dihybrid cross can be said to represent an average cross Mitochondria and chloroplasts are specialized organ-
for a population in which two polygenes are segregating. elles located in the cytoplasm. They contain small circular
Identifying polygenes and understanding how they act chromosomes that carry a defined subset of the total cell
and interact are important challenges for geneticists in the genome. Mitochondrial genes are concerned with the mito-
twenty-first century. Identifying polygenes will be especially chondrion’s task of energy production, whereas chloroplast
important in medicine. Many common human diseases such genes are needed for the chloroplast to carry out its function
as atherosclerosis (hardening of the arteries) and hypertension of photosynthesis. However, neither organelle is functionally
(high blood pressure) are thought to have a polygenic com- autonomous because each relies to a large extent on nuclear
ponent. If so, a full understanding of these conditions, which genes for its function. Why some of the necessary genes are in
affect large proportions of human populations, requires an the organelles themselves and others are in the nucleus is still
understanding of these polygenes, their inheritance, and their something of a mystery, which will not be addressed here.
function. Today, several molecular approaches can be applied Another peculiarity of organelle genes is the large num-
to the job of finding polygenes, and we will consider some in ber of copies present in a cell. Each organelle is present
in many copies per cell, and, furthermore, each organelle
contains many copies of its chromosome. Hence, a eukary-
Histogram of polygenes from a trihybrid self
otic cell can contain hundreds or thousands of organelle
chromosomes. Consider chloroplasts, for example. Any
Possible effects of
environmental variation green cell of a plant has many chloroplasts, and each chlo-
20 roplast contains many identical circular DNA molecules,
the so-called chloroplast chromosomes. Hence, the number
of chloroplast chromosomes per cell can be in the thou-
15
sands, and the number can even vary somewhat from cell
ths
20
64
Cell showing nucleoids within mitochondria in Figure 3-18. Organelle genes are very closely spaced, and,
in some organisms, organelle genes can contain untranslated
segments called introns. Note how most genes concern the
chemical reactions taking place within the organelle itself: pho-
tosynthesis in chloroplasts and oxidative phosphorylation in
mitochondria.
Organelle genomes
(a) Yeast mitochondrial DNA (~78 kb) (b) Liverwort chloroplast DNA (121 kb)
IR
IR
A
B
(ad +) (ad – )
Poky, ad – Normal, ad –
2n 2n
Meiosis Meiosis
(ad – ) (ad +)
FIGURE 3-19 Reciprocal crosses of poky and wild-type Neurospora produce different results
because a different parent contributes the cytoplasm. The female parent contributes most of the
cytoplasm of the progeny cells. Brown shading represents cytoplasm with mitochondria containing the
poky mutation, and green shading represents cytoplasm with wild-type mitochondria. Note that all the
progeny in part a are poky, whereas all the progeny in part b are normal. Hence, both crosses show
maternal inheritance. The nuclear gene with the alleles ad + (black) and ad − (red) is used to illustrate the
segregation of the nuclear genes in the 1 : 1 Mendelian ratio expected for this haploid organism.
Variegated leaves caused by a mutation in cpDNA lethality, white descendants would not live beyond the seed-
ling stage.)
The variegated zygotes (bottom of Figure 3-21) demon-
strate cytoplasmic segregation. These variegated progeny
come from eggs that are cytohets. Interestingly, when such
a zygote divides, the white and green chloroplasts often seg-
regate over successive cell divisions; that is, they sort them-
selves into separate cells, yielding the distinct green and white
sectors that cause the variegation in the branches. Here, then,
is a direct demonstration of cytoplasmic segregation.
Given that a cell is a population of organelle mole-
All-white branch cules, how is it ever possible to obtain a “pure” mutant cell,
containing only mutant chromosomes? Most likely, pure
All-green branch
mutants are created in asexual cells as follows. The variants
arise by mutation of a single gene in a single chromosome.
Then, in some cases, the mutation-bearing chromosome
may by chance increase in frequency in the population
within the cell. This process is called random genetic drift
Main shoot is variegated and is discussed more fully in Chapter 18. A cell that is
a cytohet may have, say, 60 percent A chromosomes and
40 percent a chromosomes. When this cell divides, some-
FIGURE 3-20 Leaf variegation in Mirabilis jalapa, the four-o’clock times all the A chromosomes go into one daughter, and all
plant. Flowers can form on any branch (variegated, green, or white), the a chromosomes into the other (again, by chance). More
and these flowers can be used in crosses.
often, this partitioning requires several subsequent genera-
tions of cell division to be complete. Hence, as a result of
Egg
type + White
1
Egg
type + Green
2
Egg
type +
3 Successive
Variegated
cell divisions
these chance events, the two alleles are expressed in differ- Cytoplasmic mutations in humans
ent daughter cells, and this separation will continue through
Are there cytoplasmic mutations in humans? Some human
the descendants of these cells. Note that cytoplasmic segre-
pedigrees show the transmission of rare disorders only
gation is not a mitotic process; it does take place in dividing
through females and never through males. This pattern
asexual cells, but it is unrelated to mitosis. In chloroplasts,
strongly suggests cytoplasmic inheritance and points to a
cytoplasmic segregation is a common mechanism for pro-
mutation in mtDNA as the reason for the phenotype. The
ducing variegated (green-and-white) plants, as already
disease MERRF (myoclonic epilepsy with ragged red fibers)
mentioned. In fungal mutants such as the poky mutant of
is such a phenotype, resulting from a single base change in
Neurospora, the original mutation in one mtDNA molecule
mtDNA. It is a disease that affects muscles, but the symp-
must have accumulated and undergone cytoplasmic segre-
toms also include eye and hearing disorders. Another exam-
gation to produce the strain expressing the poky symptoms.
ple is Kearns–Sayre syndrome, a constellation of symptoms
affecting the eyes, heart, muscles, and brain that is caused
KEY CONCEPT Organelle populations that contain mixtures
of two genetically distinct chromosomes often show segrega-
by the loss of part of the mtDNA. In some of these cases,
tion of the two types into the daughter cells after one or more the cells of those affected contain mixtures of normal and
cell divisions. This process is called cytoplasmic segregation. mutant chromosomes, and the proportions of each passed
on to progeny can vary as a result of cytoplasmic segrega-
KEY CONCEPT Alleles on organelle chromosomes tion. The proportions in one person can also vary in differ-
ent tissues or over time. The accumulation of certain types
1. in sexual crosses are inherited from one parent only (gen-
erally the maternal parent) and hence show no segregation of mitochondrial mutations over time has been proposed as
ratios of the type nuclear genes do. a possible cause of aging.
Figure 3-22 shows some of the mutations in human mito-
2. in asexual cells can show cytoplasmic segregation.
chondrial genes that can lead to disease when, by random
II
III
FIGURE 3-23 This pedigree shows that a human mitochondrial disease is inherited only from the
mother.
drift and cytoplasmic segregation, they rise in frequency to to construct evolutionary trees. Furthermore, it has been
such an extent that cell function is impaired. The inheritance possible to introduce some extinct organisms into evo-
of a human mitochondrial disease is shown in Figure 3-23. lutionary trees using mtDNA sequences obtained from
Note that the condition is always passed to offspring by their remains, such as skins and bones in museums.
mothers and never fathers. Occasionally, a mother will pro- mtDNA evolves relatively rapidly, so this approach has
duce an unaffected child (not shown), probably owing to been most useful in plotting recent evolution such as the
cytoplasmic segregation in the gamete-forming tissue. evolution of humans and other primates. One key find-
ing is that the “root” of the human mtDNA tree is in
mtDNA in evolutionary studies Africa, suggesting that Homo sapiens originated in Africa
and from there dispersed throughout the world (see
Differences and similarities of homologous mtDNA
Chapter 18).
sequences between species have been used extensively
SUMMARY
Genetic research and plant and animal breeding often neces- rule—that is, by multiplying the proportions relevant to the
sitate the synthesis of genotypes that are complex combi- individual genes. The probability of the occurrence of any
nations of alleles from different genes. Such genes can be of several categories of progeny is calculated by applying
on the same chromosome or on different chromosomes; the the sum rule—that is, by adding their individual probabil-
latter is the main subject of this chapter. ities. In mnemonic form, the product rule deals with “A
In the simplest case—a dihybrid for which the two gene AND B,” whereas the sum rule deals with “A′ OR A ″.” The
pairs are on different chromosome pairs—each individual χ2 test can be used to test whether the observed proportions
gene pair shows equal segregation at meiosis as predicted of classes in genetic analysis conform to the expectations
by Mendel’s first law. Because nuclear spindle fibers attach of a genetic hypothesis, such as a hypothesis of single- or
randomly to centromeres at meiosis, the two gene pairs are two-gene inheritance. If a probability value of less than
partitioned independently into the meiotic products. This 5 percent is calculated, the hypothesis must be rejected.
principle of independent assortment is called Mendel’s sec- Sequential generations of selfing increase the propor-
ond law because Mendel was the first to observe it. From tions of homozygotes, according to the principles of equal
a dihybrid A/a ; B/b, four genotypes of meiotic products segregation and independent assortment (if the genes are
are produced, A ; B, A ; b, a ; B, and a ; b, all at an equal fre- on different chromosomes). Hence, selfing is used to cre-
quency of 25 percent each. Hence, in a testcross of a dihy- ate complex pure lines with combinations of desirable
brid with a double recessive, the phenotypic proportions of mutations.
the progeny also are 25 percent (a 1 : 1 : 1 : 1 ratio). If such a The independent assortment of chromosomes at meiosis
dihybrid is selfed, the phenotypic classes in the progeny are can be observed cytologically by using heteromorphic chro-
9 3 3 1
16
A /− ; B /−, 16 A /− ; b /b, 16 a /a ; B /−, and 16 a /a ; b /b. The mosome pairs (those that show a structural difference). The
1 : 1 : 1 : 1 and 9 : 3 : 3 : 1 ratios are both diagnostic of inde- X and Y chromosomes are one such case, but other, rarer
pendent assortment. cases can be found and used for this demonstration. The
More complex genotypes composed of independently independent assortment of genes at the level of single meio-
assorting genes can be treated as extensions of the case cytes can be observed in the ascomycete fungi, because the
for single-gene segregation. Overall genotypic, phenotypic, asci show the two alternative types of segregations at equal
or gametic ratios are calculated by applying the product frequencies.
One of the main functions of meiosis is to produce histogram showing the proportion of each phenotype
recombinants, new combinations of alleles of the haploid approximates a bell-shaped curve typical of continuous
genotypes that united to form the meiocyte. Independent variation.
assortment is the main source of recombinants. In a dihy- The small subsets of the genome found in mitochondria
brid testcross showing independent assortment, the recom- and chloroplasts are inherited independently of the nuclear
binant frequency will be 50 percent. genome. Mutants in these organelle genes often show mater-
Metric characters such as color intensity show a con- nal inheritance, along with the cytoplasm, which is the loca-
tinuous distribution in a population. Continuous distribu- tion of these organelles. In genetically mixed cytoplasms
tions can be based on environmental variation or on variant (cytohets), the two genotypes (say, wild type and mutant)
alleles of multiple genes, or on a combination of both. A often sort themselves out into different daughter cells by a
simple genetic model proposes that the active alleles of sev- poorly understood process called cytoplasmic segregation.
eral genes (called polygenes) contribute more or less addi- Mitochondrial mutation in humans results in diseases that
tively to the metric character. In an analysis of the progeny show cytoplasmic segregation in body tissues and maternal
from the self of a multiply heterozygous individual, the inheritance in a mating.
KEY TERMS
chi-square test (p. 86) law of independent assortment product rule (p. 85)
chloroplast DNA (cpDNA) (p. 97) (Mendel’s second law) (p. 81) quantitative trait locus (QTL) (p. 96)
cytoplasmic segregation (p. 99) maternal inheritance (p. 98) recombinant (p. 93)
dihybrid (p. 81) meiotic recombination (p. 93) recombination (p. 93)
dihybrid cross (p. 81) mitochondrial DNA (mtDNA) sum rule (p. 85)
hybrid vigor (p. 89) (p. 97) uniparental inheritance (p. 98)
independent assortment (p. 80) polygene (quantitative trait locus)
(p. 96)
SOLVED PROBLEMS
and females. This ratio can be explained if both par- From this cross, we predict
ents are heterozygous for an autosomal gene; we can Females
represent them as L/l, where L stands for long and l
stands for clipped. › 12 L /l › 41 D /d ; L /l
1
D /d
Having done this partial analysis, we see that only 2 › 1
l /l › 41 D /d ; l /l
2
the inheritance of wing transparency is associated with
1 › 41 d /d ; L /l
sex. The most obvious possibility is that the alleles for › 2 L /l
1
transparent ( D) and dusky ( d) are on the X chromo- 2
d /d
› 1
l /l › 41 d /d ; l /l
some, because we have seen in Chapter 2 that gene 2
› 3
L /− › 3
d ; L /− dusky, long What were the genotypes of A, B, and C? (Use gene symbols
1 4 8
d of your own choosing; be sure to define each one.)
2 › 1
l /l › 1
d ; l /l dusky, clipped
4 8
SOLUTION
b. Generally, a good way to test such a model is to make
Notice that each of the crosses is
a cross and predict the outcome. But which cross? We
have to predict some kind of ratio in the progeny, yellow, round × green, wrinkled → progeny
and so it is important to make a cross from which a
Because A, B, and C were all crossed with the same
unique phenotypic ratio can be expected. Notice that
plant, all the differences between the three progeny popu-
using one of the female progeny as a parent would
lations must be attributable to differences in the underlying
not serve our needs: we cannot say from observing
genotypes of A, B, and C.
the phenotype of any one of these females what her
You might remember a lot about these analyses from the
genotype is. A female with transparent wings could
chapter, which is fine, but let’s see how much we can deduce
be D/D or D/d, and one with long wings could be
from the data. What about dominance? The key cross for
L/L or L/l. It would be good to cross the parental
deducing dominance is B. Here, the inheritance pattern is
female of the original cross with a dusky, clipped son,
because the full genotypes of both are specified under yellow, round × green, wrinkled → all yellow, round
the model that we have created. According to our
So yellow and round must be dominant phenotypes
model, this cross is
because dominance is literally defined by the phenotype of
D /d ; L /l × d ; l /l a hybrid. Now we know that the green, wrinkled parent
used in each cross must be fully recessive; we have a very Now cross B becomes crystal clear and must have been
convenient situation because it means that each cross is a test-
Y /Y ; R /R × y /y ; r /r → Y /y ; r /r
cross, which is generally the most informative type of cross.
Turning to the progeny of A, we see a 1 : 1 ratio for because any heterozygosity in pea B would have given rise
yellow to green. This ratio is a demonstration of Mendel’s first to several progeny phenotypes, not just one.
law (equal segregation) and shows that, for the character of What about C? Here, we see a ratio of 50 yellow : 50
color, the cross must have been heterozygote × homozygous green (1 : 1) and a ratio of 49 round : 51 wrinkled (also 1 : 1).
recessive. Letting Y represent yellow and y represent green, So both genes in pea C must have been heterozygous, and
we have cross C was
1 1
Y /y × y /y → 2
Y /y (yellow) → 2
y /y (green) Y /y ; R /r × y /y ; r /r
For the character of shape, because all the progeny are round, the which is a good demonstration of Mendel’s second law
cross must have been homozygous dominant × homozygous (independent assortment of different genes).
recessive. Letting R represent round and r represent wrinkled, How would a geneticist have analyzed these crosses?
we have Basically, the same way that we just did but with
fewer intervening steps. Possibly something like this: “yel-
R /R × r /r → R /r (round)
low and round dominant; single-gene segregation in A ; B
Combining the two characters, we have homozygous dominant; independent two-gene segregation
1 1 in C.”
Y /y ; R /R × y /y ; r /r → 2
Y /y ; R /r 2
y /y ; R /r
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a problem Solving Video.
b. Which organelle has the largest proportion of non- Which cells could have been used for these measure-
genic DNA? ments? (Note: In plants, the endosperm part of the
c. What do you think is the main reason for the differ- seed is often triploid, 3n.)
ence in size of yeast and human mtDNA? 28. Draw a haploid mitosis of the genotype a+ ; b.
20. Regarding Figure 3-19: 29. In moss, the genes A and B are expressed only in the
a. Do you think it possible that some paternal mito- gametophyte. A sporophyte of genotype A/a ; B/b is
chondria leak in the meiocyte? allowed to produce gametophytes.
b. How would you test that possibility? a. What proportion of the gametophytes will be A ; B?
c. What color is used to denote cytoplasm containing b. If fertilization is random, what proportion of spo-
wild-type mitochondria? rophytes in the next generation will be A/a ; B/b?
21. In Figure 3-20 , what would be the leaf types of 30. When a cell of genotype A/a ; B/b ; C/c having all the
progeny of the apical (top) flower? genes on separate chromosome pairs divides mitoti-
cally, what are the genotypes of the daughter cells?
22. In Figure 3-22, count the sites where mutations cause
myopathy. The word myopathy means disease of the 31. In the haploid yeast Saccharomyces cerevisiae, the two
muscles; why should mitochondrial mutation so often mating types are known as a and α. Mating type is
cause this phenotype? determined by two alleles of a single gene, MATa and
MATα. You cross a purple (ad − ) strain of mating type
23. From the pedigree in Figure 3-23, what principle can a and a white (ad + ) strain of mating type α . If ad − and
you deduce about the inheritance of mitochondrial ad + are alleles of one gene, and MATa and MATα are
disease from affected fathers? alleles of an independently inherited gene on a sepa-
rate chromosome pair, what progeny do you expect to
BASIC PROBLEMS obtain? In what proportions?
24. Assume independent assortment and start with a plant 32. In mice, dwarfism is caused by an X-linked recessive
that is dihybrid A/a ; B/b: allele, and pink coat is caused by an autosomal dom-
inant allele (coats are normally brownish). If a dwarf
a. What phenotypic ratio is produced from selfing it?
female from a pure line is crossed with a pink male
b. What genotypic ratio is produced from selfing it? from a pure line, what will be the phenotypic ratios in
c. What phenotypic ratio is produced from testcross- the F1 and F2 in each sex? (Invent and define your own
ing it? gene symbols.)
d. What genotypic ratio is produced from testcrossing 33. Suppose you discover two interesting rare cytological
it? abnormalities in the karyotype of a human male. (A
karyotype is the total visible chromosome comple-
25. Normal mitosis takes place in a diploid cell of geno-
ment.) There is an extra piece (satellite) on one of the
type A/a ; B/b. Which of the following genotypes might
chromosomes of pair 4, and there is an abnormal pat-
represent possible daughter cells?
tern of staining on one of the chromosomes of pair 7.
a. A ; B b. a ; b c. A ; b With the assumption that all the gametes of this male
d. a ; B e. A/A ; B/B f. A/a ; B/b are equally viable, what proportion of his children will
have the same karyotype that he has?
g. a/a ; b/b
34. Suppose that meiosis occurs in the transient diploid
26. In a diploid organism of 2n = 10, assume that you
stage of the cycle of a haploid organism of chromo-
can label all the centromeres derived from its female
some number n. What is the probability that an indi-
parent and all the centromeres derived from its male
vidual haploid cell resulting from the meiotic division
parent. When this organism produces gametes, how
will have a complete parental set of centromeres (that
many male- and female-labeled centromere combina-
is, a set all from one parent or all from the other
tions are possible in the gametes?
parent)?
27. It has been shown that when a thin beam of light is
35. Pretend that the year is 1868. You are a skilled young
aimed at a nucleus, the amount of light absorbed
lens maker working in Vienna. With your superior new
is proportional to the cell’s DNA content. Using this
lenses, you have just built a microscope that has better
method, the DNA in the nuclei of several different
resolution than any others available. In your testing of
types of cells in a corn plant were compared. The
this microscope, you have been observing the cells in
following numbers represent the relative amounts of
the testes of grasshoppers and have been fascinated by
DNA in these different types of cells:
the behavior of strange elongated structures that you
0.7, 1.4, 2.1, 2.8, and 4.2 have seen within the dividing cells. One day, in the
library, you read a recent journal paper by G. Mendel a. What do these results show? Include proposed
on hypothetical “factors” that he claims explain the genotypes of all plants in your answer.
results of certain crosses in peas. In a flash of reve- b. Assume that the genes are located on separate chro-
lation, you are struck by the parallels between your mosomes. An F1 is produced by crossing the original
grasshopper studies and Mendel’s pea studies, and mutant A with the original mutant B. This F1 is test-
you resolve to write him a letter. What do you write? crossed: What proportion of testcross progeny will
(Problem 35 is based on an idea by Ernest Kroeker.) have no trichomes?
36. From a presumed testcross A /a × a /a , in which A rep- 40. In dogs, dark coat color is dominant over albino, and
resents red and a represents white, use the χ2 test to short hair is dominant over long hair. Assume that
find out which of the following possible results would these effects are caused by two independently assorting
fit the expectations: genes. Seven crosses were done as shown below, in
a. 120 red, 100 white which D and A stand for the dark and albino pheno-
b. 5000 red, 5400 white types, respectively, and S and L stand for the short-hair
and long-hair phenotypes.
c. 500 red, 540 white
d. 50 red, 54 white Number of progeny
37. Look at the Punnett square in Figure 3-4. Parental phenotypes D, S D, L A, S A, L
39. Note: The first part of this problem was introduced in Mating phenotypes P, C P, Po G, C G, Po
Chapter 2. The line of logic is extended here. 1 P, C × G, C 323 102 309 106
In the plant Arabidopsis thaliana, a geneticist is 2 P, C × P, Po 220 206 65 72
interested in the development of trichomes (small pro-
3 P, C × G, C 723 229 0 0
jections) on the leaves. A large screen turns up two
mutant plants (A and B) that have no trichomes, and 4 P, C × G, Po 405 0 389 0
these mutants seem to be potentially useful in studying 5 P, Po × G, C 71 90 85 78
trichome development. (If they are determined by
a. Which alleles are dominant?
single-gene mutations, then finding the normal and
abnormal function of these genes will be instructive.) b. What are the most probable genotypes for the par-
Each plant was crossed with wild type; in both cases, ents in each cross?
the next generation ( F1 ) had normal trichomes. When 42. A mutant allele in mice causes a bent tail. Six pairs of
F1 plants were selfed, the resulting F2’s were as follows: mice were crossed. Their phenotypes and those of their
F2 from mutant A : 602 normal ; 198 no trichomes progeny are given in the following table. N is normal
phenotype; B is bent phenotype. Deduce the mode of
F2 from mutant B : 267 normal ; 93 no trichomes
inheritance of bent tail.
Parents Progeny 8. List the F2 phenotypic ratios for each character that
you came up with in answer to question 4.
Cross ♀ ♂ ♀ ♂
9. What do the F2 phenotypic ratios tell you?
1 N B All B All N
10. What major inheritance pattern distinguishes sex-
2 B N 1
B, 1
N 1
B, 1
N
2 2 2 2 linked inheritance from autosomal inheritance?
3 B N All B All B 11. Do the F2 data show such a distinguishing criterion?
4 N N All N All N 12. Do the F1 data show such a distinguishing criterion?
5 B B All B All B
13. What can you learn about dominance in the F1? The
6 B B All B 1
B, 1
N F2?
2 2
a. Is it recessive or dominant? 14. What rules about wild-type symbolism can you use
in deciding which allelic symbols to invent for these
b. Is it autosomal or sex-linked?
crosses?
c. What are the genotypes of all parents and progeny?
15. What does “deduce the inheritance of these pheno-
43. The normal eye color of Drosophila is red, but strains types” mean?
in which all flies have brown eyes are available. Sim-
Now try to solve the problem. If you are unable to do so,
ilarly, wings are normally long, but there are strains
try to identify the obstacle and write a sentence or two
with short wings. A female from a pure line with
describing your difficulty. Then go back to the expansion
brown eyes and short wings is crossed with a male
questions and see if any of them relate to your difficulty. If
from a normal pure line. The F1 consists of normal
this approach doesn’t work, inspect the Learning Objec-
females and short-winged males. An F2 is then pro-
tives and Key Concepts of this chapter and ask yourself
duced by intercrossing the F1. Both sexes of F2 flies
which might be relevant to your difficulty.
show phenotypes as follows:
3 44. In a natural population of annual plants, a single plant
8
red eyes, long wings
is found that is sickly looking and has yellowish leaves.
3
8
red eyes, short wings The plant is dug up and brought back to the labora-
1
brown eyes, long wings tory. Photosynthesis rates are found to be very low.
8
Pollen from a normal dark-green-leaved plant is used
1
8
brown eyes, short wings to fertilize emasculated flowers of the yellowish plant.
A hundred seeds result, of which only 60 germinate.
Deduce the inheritance of these phenotypes; use clearly
All the resulting plants are sickly yellow in appearance.
defined genetic symbols of your own invention. State
the genotypes of all three generations and the geno- a. Propose a genetic explanation for the inheritance
typic proportions of the F1 and F2 . pattern.
www
b. Suggest a simple test for your model.
UNPACKING PROBLEM 43
www c. Account for the reduced photosynthesis, sickliness,
Before attempting a solution to this problem, try answer- and yellowish appearance.
ing the following questions: 45. What is the basis for the green-and-white color varie-
1. What does the word normal mean in this problem? gation in the leaves of Mirabilis? If the following cross
is made,
2. The words line and strain are used in this problem.
What do they mean, and are they interchangeable? variegated ♀ × green ♂
3. Draw a simple sketch of the two parental flies what progeny types can be predicted? What about the
showing their eyes, wings, and sexual differences. reciprocal cross?
4. How many different characters are there in this 46. In Neurospora, the mutant stp exhibits erratic stop-
problem? and-start growth. The mutant site is known to be in the
5. How many phenotypes are there in this problem, mtDNA. If an stp strain is used as the female parent
and which phenotypes go with which characters? in a cross with a normal strain acting as the male,
what type of progeny can be expected? What about
6. What is the full phenotype of the F1 females called the progeny from the reciprocal cross?
“normal”?
47. Two corn plants are studied. One is resistant (R) and
7. What is the full phenotype of the F1 males called the other is susceptible (S) to a certain pathogenic fun-
“short winged”? gus. The following crosses are made, with the results
shown:
black, forked 230 brown, forked 240 55. Reciprocal crosses and selfs were performed between
the two moss species Funaria mediterranea and F.
black, unforked 210 brown, unforked 250 hygrometrica. The sporophytes and the leaves of the
2
Use the χ test to determine if these results fit the gametophytes are shown in the accompanying diagram.
results expected from testcrossing the hypothesized
dihybrid.
49. A plant geneticist has two pure lines, one with purple
petals and one with blue. She hypothesizes that the
phenotypic difference is due to two alleles of one gene.
To test this idea, she aims to look for a 3 : 1 ratio in
the F2 . She crosses the lines and finds that all the F1
progeny are purple. The F1 plants are selfed, and 400
F2 plants are obtained. Of these F2 plants, 320 are pur-
ple and 80 are blue. Do these results fit her hypothesis
well? If not, suggest why.
50. Are the following progeny numbers consistent with the
results expected from selfing a plant presumed to be a
dihybrid of two independently assorting genes, H/h ; R/r?
(H = hairy leaves; h = smooth leaves; R = round ovary;
r = elongated ovary.) Explain your answer.
hairy, round 178 smooth, round 56 The crosses are written with the female parent first.
hairy, elongated 62 smooth, elongated 24 3
Progeny
51. A dark female moth is crossed with a dark male. All
the male progeny are dark, but half the female progeny
are light and the rest are dark. Propose an explanation
for this pattern of inheritance.
52. In Neurospora, a mutant strain called stopper (stp)
arose spontaneously. Stopper showed erratic “stop
and start” growth, compared with the uninterrupted
growth of wild-type strains. In crosses, the following
results were found:
♀ stopper × ♂ wild type → progeny all stopper
♀ wild type × ♂ stopper → progeny all wild type 3
Progeny
a. What do these results suggest regarding the location
of the stopper mutation in the genome?
b. According to your model for part a, what progeny
and proportions are predicted in octads from the
following cross, including a mutation nic3 located
on chromosome VI?
♀ stp ⋅ nic3 × wild type ♂
a. Describe the results presented, summarizing the produce a new pure line for trade that is yellow, two-
main findings. loculed, and tall. How exactly should he go about doing
b. Propose an explanation of the results. so? Show not only which crosses to make, but also how
many progeny should be sampled in each case.
c. Show how you would test your explanation; be
sure to show how it could be distinguished from 60. We have dealt mainly with only two genes, but the
other explanations. same principles hold for more than two genes. Con-
sider the following cross:
56. Assume that diploid plant A has a cytoplasm
genetically different from that of plant B. To study A /a ; B /b ; C /c ; D /d ; E /e × a /a ; B /b ; c /c ; D /d ; e /e
nuclear–cytoplasmic relations, you wish to obtain a a. What proportion of progeny will phenotypically
plant with the cytoplasm of plant A and the nuclear resemble (1) the first parent, (2) the second parent,
genome predominantly of plant B. How would you go (3) either parent, and (4) neither parent?
about producing such a plant? b. What proportion of progeny will be genotypically
57. You are studying a plant with tissue comprising both the same as (1) the first parent, (2) the second
green and white sectors. You wish to decide whether parent, (3) either parent, and (4) neither parent?
this phenomenon is due (1) to a chloroplast muta- Assume independent assortment.
tion of the type considered in this chapter or (2) to a
dominant nuclear mutation that inhibits chlorophyll 61. The accompanying pedigree shows the pattern of trans-
production and is present only in certain tissue layers mission of two rare human phenotypes: cataract and
of the plant as a mosaic. Outline the experimental pituitary dwarfism. Family members with cataract are
approach that you would use to resolve this problem. shown with a solid left half of the symbol; those with
pituitary dwarfism are indicated by a solid right half.
CHALLENGING PROBLEMS I
1 2
58. You have three jars containing marbles, as follows:
jar 1 600 red and 400 white II
jar 2 900 blue and 100 white 1 2 3 4 5 6 7
(3) a red, a green, and a white. a. What is the most likely mode of inheritance of each
(4) a red and two whites. of these phenotypes? Explain.
(5) a color and two whites. b. List the genotypes of all members in generation III
as far as possible.
(6) at least one white.
c. If a hypothetical mating took place between IV-1
b. In a certain plant, R = red and r = white. You self and IV-5, what is the probability of the first child’s
a red R/r heterozygote with the express purpose of being a dwarf with cataracts? A phenotypically nor-
obtaining a white plant for an experiment. What mal child?
minimum number of seeds do you have to grow to
be at least 95 percent certain of obtaining at least (Problem 61 is adapted from J. Kuspira and R. Bham-
one white individual? bhani, Compendium of Problems in Genetics. Copy-
right 1994 by Wm. C. Brown.)
c. When a woman is injected with an egg fertilized in
vitro, the probability of its implanting successfully 62. A corn geneticist has three pure lines of genotypes
is 20 percent. If a woman is injected with five eggs a/a ; B/B ; C/C, A/A ; b/b ; C/C, and A/A ; B/B ; c/c. All the
simultaneously, what is the probability that she will phenotypes determined by a, b, and c will increase the
become pregnant? (Part c is from Margaret Holm.) market value of the corn; so, naturally, he wants to com-
bine them all in one pure line of genotype a/a ; b/b ; c/c.
59. In tomatoes, red fruit is dominant over yellow, two-
loculed fruit is dominant over many-loculed fruit, and a. Outline an effective crossing program that can be
tall vine is dominant over dwarf. A breeder has two used to obtain the a/a ; b/b ; c/c pure line.
pure lines: (1) red, two-loculed, dwarf and (2) yellow, b. At each stage, state exactly which phenotypes will
many-loculed, tall. From these two lines, he wants to be selected and give their expected frequencies.
c. Is there more than one way to obtain the desired S phase, a radioactive nucleotide was added and was
genotype? Which is the best way? incorporated into newly synthesized DNA. The cells
Assume independent assortment of the three gene were then removed from the radioactivity, washed,
pairs. (Note: Corn will self- or cross-pollinate easily.) and allowed to proceed through mitosis. Radioactive
chromosomes or chromatids can be detected by plac-
63. In humans, color vision depends on genes encoding ing photographic emulsion on the cells; radioactive
three pigments. The R (red pigment) and G (green chromosomes or chromatids appeared covered with
pigment) genes are close together on the X chromo- spots of silver from the emulsion. (The chromosomes
some, whereas the B (blue pigment) gene is autosomal. “take their own photograph.”) Draw the chromo-
A recessive mutation in any one of these genes can somes at prophase and telophase of the first and sec-
cause color blindness. Suppose that a color-blind man ond mitotic divisions after the radioactive treatment. If
married a woman with normal color vision. The four they are radioactive, show it in your diagram. If there
sons from this marriage were color-blind, and the five are several possibilities, show them, too.
daughters were normal. Specify the most likely gen-
otypes of both parents and their children, explaining 66. In the species of Problem 65, you can introduce
your reasoning. (A pedigree drawing will probably be radioactivity by injection into the anthers at the S
helpful.) (Problem 63 is by Rosemary Redfield.) phase before meiosis. Draw the four products of mei-
osis with their chromosomes, and show which are
64. Consider the accompanying pedigree for a rare human radioactive.
muscle disease.
67. The plant Haplopappus gracilis is diploid and
2n = 4. There are one long pair and one short pair
of chromosomes. The diagrams below (numbered
1 through 12) represent anaphases (“pulling apart”
stages) of individual cells in meiosis or mitosis in a
a. What unusual feature distinguishes this pedigree plant that is genetically a dihybrid (A/a ; B/b) for genes
from those studied earlier in this chapter? on different chromosomes. The lines represent chro-
mosomes or chromatids, and the points of the V’s
b. Where do you think the mutant DNA responsible represent centromeres. In each case, indicate if the
for this phenotype resides in the cell? diagram represents a cell in meiosis I, meiosis II, or
65. The plant Haplopappus gracilis has a 2n of 4. A dip- mitosis. If a diagram shows an impossible situation,
loid cell culture was established and, at premitotic say so.
A 1 A a 7
a B b
A a B b
a A b B
B b a A b B
A 2 A A 8
b B B
a a b b
a a b b
A b B B
A A
A 3 A a 9
B B b
A a B b
A a B b
A B A a B b
a 4 A 10
B b
A b
a B
a B a B
A a 5 a 11
b B B
a B
a B
A a b B a B
A a 6 A 12
B b a
A a
B b
a A b B B b
68. The pedigree below shows the recurrence of a rare What are the results of this testcross? Give geno-
neurological disease (large black symbols) and spon- types and phenotypes, and designate the kind of
taneous fetal abortion (small black symbols) in one cytoplasm.
family. (A slash means that the individual is deceased.) d. The restorer allele already described can be called
Provide an explanation for this pedigree in regard to Rf1. Another dominant restorer, Rf2, has been
the cytoplasmic segregation of defective mitochondria. found. Rf1 and Rf2 are located on different chro-
mosomes. Either or both of the restorer alleles will
give pollen fertility. With the use of a male-sterile
plant as a tester, what will be the result of a cross in
which the male parent is
(1) heterozygous at both restorer loci?
69. A man is brachydactylous (very short fingers; rare (2) homozygous dominant at one restorer locus and
autosomal dominant), and his wife is not. Both can homozygous recessive at the other?
taste the chemical phenylthiocarbamide (autosomal
dominant; common allele), but their mothers could (3) heterozygous at one restorer locus and homozy-
not. gous recessive at the other?
a. Give the genotypes of the couple. (4) heterozygous at one restorer locus and homozy-
gous dominant at the other?
If the genes assort independently and the couple has
four children, what is the probability of GENETICS AND SOCIETY
b. all of them being brachydactylous?
1. We have seen in this chapter that independent assort-
c. none being brachydactylous? ment can be used by plant and animal breeders to
d. all of them being tasters? develop new lines that have a combination of favor-
able mutant alleles making them suitable for com-
e. all of them being nontasters?
merce. This type of “genetic engineering” by making
f. all of them being brachydactylous tasters? crosses and selfs is deemed to be appropriate by the
g. none being brachydactylous tasters? general public, whereas genetic engineering by DNA
manipulation (details in later chapters) is not. Why do
h. at least one being a brachydactylous taster?
you think there is this perceived distinction, and is it
70. One form of male sterility in corn is maternally trans- logical?
mitted. Plants of a male-sterile line crossed with nor-
2. Most of the dog breeds familiar to us today were
mal pollen give male-sterile plants. In addition, some
developed by dog fanciers in Victorian times (nine-
lines of corn are known to carry a dominant nuclear
teenth century) before the discovery of Mendel’s Laws.
restorer allele ( Rf ) that restores pollen fertility in
Similarly, many breeds of agricultural animals were
male-sterile lines.
developed pre-Mendel. In what way do you think that
a. Research shows that the introduction of restorer type of animal breeding might have been different
alleles into male-sterile lines does not alter or affect from that which is used today?
the maintenance of the cytoplasmic factors for male
3. Human skin shows great variation in the amount of
sterility. What kind of research results would lead
the dark brown pigment called melanin, resulting in a
to such a conclusion?
wide range of skin colors. It is a common observation
b. A male-sterile plant is crossed with pollen from a that skin color is heritable. Devise a genetic model for
plant homozygous for Rf. What is the genotype of the inheritance of skin color, which explains this wide
the F1? The phenotype? range. Does your model shine any light on the popular
c. The F1 plants from part b are used as females in a perceptions of race and racism?
testcross with pollen from a normal plant ( rf/rf).
Mapping Eukaryote
Chromosomes by Recombination
Yellow body
0.0 Scute bristles
1.5 White eyes
3.0 Facet eyes
5.5 Echinus eyes
7.5 Ruby eyes
13.7 Crossveinless wings
20.0 Cut wings
21.0 Singed bristles
27.7 Lozenge eyes
33.0 Vermillion eyes
36.1 Miniature wings
43.0 Sable body
S
ome of the questions that geneticists want to answer and spaghetti-like that, in 1933, an electrical circuit engi-
about the genome are: What genes are present in the neer named Harry Beck drew up the streamlined (although
genome? What functions do they have? What posi- distorted) map that has remained to this day an icon of
tions do they occupy on the chromosomes? Their pursuit London. The street and underground maps of London
of the third question is broadly called mapping. Mapping are compared in Figure 4-1. Note that the positions of the
is the main focus of this chapter, but all three questions are underground stations and the exact distances between them
interrelated, as we will see. are of no interest in themselves, except as a way of getting
We all have an everyday feeling for the importance of to a destination of interest such as Westminster Abbey. We
maps in general, and, indeed, we have all used them at will see three parallels with the London maps when chro-
some time in our lives to find our way around. Relevant mosome maps are used to zero in on individual “destina-
to the focus of this chapter is that, in some situations, sev- tions,” or specific genes. First, several different types of
eral maps need to be used simultaneously. A good example chromosome maps are often necessary and must be used
in everyday life is in navigating the dense array of streets in conjunction; second, maps that contain distortions are
and buildings of a city such as London, England. A street still useful; and third, many sites on a chromosome map are
map that shows the general layout is one necessity. How- charted only because they are useful in trying to zero in on
ever, the street map is used by tourists and Londoners alike other sites that are the ones of real interest.
in conjunction with another map, that of the underground Obtaining a map of gene positions on the chromosomes
railway system. The underground system is so complex is an endeavor that has occupied thousands of geneticists
FIGURE 4-1 These London maps illustrate the principle that, often, several maps are needed to
get to a destination of interest. The map of the underground railway (“the Tube”) is used to get to a
destination of interest such as a street address, shown on the street map. In genetics, two different
kinds of genome maps are often useful in locating a gene, leading to an understanding of its structure
and function. [(right) Transport for London.]
114
for the past 100 years or so. Why is it so important? There are linked, they mean that the loci of those genes are on the
are several reasons: same chromosome and, hence, the alleles on any one homo-
log are physically joined (linked) by the DNA between
1. Gene position is crucial information needed to build com-
them. The way in which early geneticists deduced linkage
plex genotypes required for experimental purposes or for
is a useful means of introducing most of the key ideas and
commercial applications. For example, in Chapter 5, we
procedures in the analysis.
will see cases in which special allelic combinations must
be put together to explore gene interaction.
2. Knowing the position occupied by a gene provides a Using recombinant frequency
way of isolating it and discovering its structure and to recognize linkage
function. A gene’s position can be used to define it at In the early 1900s, William Bateson and R. C. Punnett
the DNA level. In turn, the DNA sequence of a wild- (for whom the Punnett square was named) were study-
type gene or its mutant allele is a necessary part of ing the inheritance of two genes in sweet peas. In a stan-
deducing its underlying function. dard self of a dihybrid F1, the F2 did not show the 9:3:3:1
3. The genes present and their arrangement on chromo- ratio predicted by the principle of independent assort-
somes are often slightly different in related species. For ment. In fact, Bateson and Punnett noted that certain
example, the rather long human chromosome number 2 combinations of alleles showed up more often than
is split into two shorter chromosomes in the great apes. expected, almost as though they were physically attached
By comparing such differences, geneticists can deduce in some way. However, they had no explanation for this
the evolutionary genetic mechanisms through which discovery.
these genomes diverged. Hence, chromosome maps are Later, Thomas Hunt Morgan found a similar devia-
useful in interpreting mechanisms of evolution. tion from Mendel’s second law while studying two auto-
somal genes in Drosophila. Morgan proposed linkage as a
The arrangement of genes on chromosomes is represented
hypothesis to explain the phenomenon of apparent allele
diagrammatically as a unidimensional chromosome map,
association.
showing gene positions known as loci (sing., locus), and the
Let’s look at some of Morgan’s data. One of the genes
distances between the loci based on some kind of scale. Two
affected eye color (pr, purple, and pr + , red), and the other
basic types of chromosome maps are currently used in genet-
gene affected wing length (vg, vestigial, and vg + , normal).
ics; they are assembled by quite different procedures yet are
(Vestigial wings are very small compared to wild type.) The
used in a complementary way. R ecombination-based maps,
wild-type alleles of both genes are dominant. Morgan per-
which are the topic of this chapter, map the loci of genes
formed a cross to obtain dihybrids and then followed with
that have been identified by mutant phenotypes showing
a testcross:
single-gene inheritance. Physical maps (see Chapter 14) show
the genes as segments arranged along the long DNA mole- P pr /pr ⋅ vg /vg × pr + /pr + ⋅ vg + /vg +
cule that constitutes a chromosome. These maps show differ-
↓
ent views of the genome, but, just like the maps of London,
they can be used together to arrive at an understanding of Gametes pr ⋅ vg pr + ⋅ vg +
what a gene’s function is at the molecular level and how that
↓
function influences phenotype.
F1 dihybrid pr /pr ⋅ vg + /vg
+
Morgan’s testcross results were as follows (listed as the Linked alleles tend to be inherited together
gametic classes from the dihybrid):
pr vg pr + vg +
pr + ⋅ vg + 1339
P ×
pr ⋅ vg 1195 pr pr +
vg vg +
+
pr ⋅ vg 151
pr ⋅ vg + 154
pr vg pr + vg +
2839 Gametes
How crossovers produce recombinants are on different homologs, Chiasmata are the
for linked genes in what is called a trans sites of crossing over
conformation (trans means
The linkage hypothesis explains why allele combinations “opposite”). The two con-
from the parental generations remain together: the genes formations are written as
are physically attached by the segment of chromosome follows:
between them. But exactly how are any recombinants
produced when genes are linked? Morgan suggested that, Cis AB /ab or + + /ab
when homologous chromosomes pair at meiosis, the chro- Trans Ab /aB or + b /a +
mosomes occasionally break and exchange parts in a pro-
cess called crossing over. Figure 4-3 illustrates this physical Note the following con-
exchange of chromosome segments. The two new combina- ventions that pertain to
tions are called crossover products. linkage symbolism:
Is there any microscopically observable process that 1. Alleles on the same
could account for crossing over? At meiosis, when dupli- homolog have no punc-
cated homologous chromosomes pair with each other, a tuation between them.
cross-shaped structure called a chiasma (pl., chiasmata)
2. A slash symbolically sep-
often forms between two nonsister chromatids. Chiasmata FIGURE 4-4 Several chiasmata
arates the two homologs.
are shown in Figure 4-4. To Morgan, the appearance of the appear in this photograph taken
chiasmata visually corroborated the concept of crossing 3. Alleles are always writ- in the course of meiosis in a
ten in the same order grasshopper testis. [Republished
over. (Note that the chiasmata seem to indicate that chro-
with permission of Elsevier,
matids, not unduplicated chromosomes, participate in a on each homolog.
Gareth H. Jones, F. Chris H.
crossover. We will return to this point later.) 4. As in earlier chapters, Franklin, “Meiotic Crossingover:
genes known to be on Obligation and Interference,” Cell,
KEY CONCEPT For linked genes, recombinants are different chromosomes 2006, July; 126 (2): 246–248, Figure
produced by crossovers between nonsister chromatids 1. Permission conveyed through
(unlinked genes) are
during meiosis. Chiasmata are the visible manifestations of Copyright Clearance Center, Inc.]
shown separated by a
crossovers.
semicolon—for example, A /a ; C /c.
5. In this book, genes of unknown linkage are shown sepa-
rated by a dot, A /a ⋅ D /d .
Linkage symbolism and terminology
The work of Morgan showed that linked genes in a dihybrid
may be present in one of two basic conformations. In one, Evidence that crossing over is a
the two dominant, or wild-type, alleles are present on the breakage-and-rejoining process
same homolog (as in Figure 4-3); this arrangement is called a The idea that recombinants are produced by some kind of
cis conformation (cis means “adjacent”). In the other, they exchange of material between homologous chromosomes
was a compelling one. But experimentation was necessary to
test this hypothesis. A first step was to find a case in which
Crossing over produces new allelic combinations the exchange of parts between chromosomes would be visible
under the microscope. Several investigators approached this
pr vg +
vg pr vg pr problem in the same way, and one of their analyses follows.
In 1931, Harriet Creighton and Barbara McClintock
were studying two genes in corn that they knew were both
located on chromosome 9. One affected seed color (C, col-
ored; c, colorless), and the other affected endosperm com-
position (Wx, waxy; wx, starchy). The plant was a dihybrid
vg + pr + vg +
pr +
pr + vg in cis conformation. However, in one plant, the chromo-
Crossover between some 9 carrying the alleles C and Wx was unusual in that
nonsister chromatids it also carried a large, densely staining element (called a
Parental Meiosis Crossover knob) on the C end and a longer piece of chromosome on
chromosomes chromosomes the Wx end; thus, the heterozygote was
In the progeny of a testcross of this plant, they exam- Crossing over is between
ined the chromosomes of the recombinants and parental chromatids, not chromosomes
genotypes. They found that all the recombinants inher-
Two-chromosome stage
ited one or the other of the two following chromosomes, A b
depending on their recombinant makeup: A B A b A b
wx C a B
a b a B a B
Four-chromatid stage
Wx c A B
A B
A B A b
Multiple crossovers can include two or more chromatids 4.2 MAPPING BY RECOMBINANT
(a) Two chromatids FREQUENCY
Position of crossovers Tetrad genotypes
LO 4.2 Show the impact of one or more crossovers in
A B C A B C producing recombinants between linked genes.
A B C A b C
(b) Three chromatids LO 4.4 Extend the same type of analysis to a three-gene
(“three-point”) testcross to assess whether or not
Position of crossovers Tetrad genotypes three genes are linked, and if so, their relative order
A B C A B c and map distances.
ANIMATED A B A B
Meioses Parental
ART
with no A B A B
Meiotic recombination crossover Parental
between linked genes by between a b a b
Parental
crossing over the genes
a b a b
Parental
A B A B
Meioses Parental
with a A B A b
Recombinant
crossover
between a b a B
Recombinant
the genes
a b a b
Parental
A C
Map based on A–C recombination
3 m.u.
C A B
3 m.u. 5 m.u.
Possible combined maps 8 m.u.
A C B
3 m.u. 2 m.u.
5 m.u.
arranged in some linear order, making map distances addi- know that 100 − 11 = 89 percent will be nonrecombinant in
tive. (There are some minor but not insignificant exceptions, two equal classes, 44.5 percent pr + vg + and 44.5 percent
as we will see later.) Since we now know from molecular pr vg. (Note that the tester contribution pr vg was ignored
analysis that a chromosome is a single DNA molecule with in writing out these genotypes.)
the genes arranged along it, it is no surprise for us today There is a strong implication that the “distance” on a
to learn that recombination-based maps are linear because linkage map is a physical distance along a chromosome,
they reflect a linear array of genes. and Morgan and Sturtevant certainly intended to imply
How is a map represented? As an example, in Dro- just that. But we should realize that the linkage map is
sophila, the locus of the eye-color gene and the locus of the a hypothetical entity constructed from a purely genetic
wing-length gene are approximately 11 m.u. apart, as men- analysis. The linkage map could have been derived with-
tioned earlier. The relation is usually diagrammed in the out even knowing that chromosomes existed. Further-
following way: more, at this point in our discussion, we cannot say
whether the “genetic distances” calculated by means of
pr 11.0 m.u. vg recombinant frequencies in any way represent actual
physical distances on chromosomes. However, physical
mapping has shown that genetic distances are, in fact,
Generally, we refer to the locus of this eye-color gene roughly proportional to recombination-based distances.
in shorthand as the “pr locus,” after the first discovered There are exceptions caused by recombination hotspots,
mutant allele, but we mean the place on the chromosome places in the genome where crossing over takes place
where any allele of this gene will be found, mutant or wild more frequently than usual. The presence of hotspots
type. causes proportional expansion of some regions of the
As stated in Chapters 2 and 3, genetic analysis can be map. Recombination blocks, which have the opposite
applied in two opposite directions. This principle is appli- effect, also are known.
cable to recombinant frequencies. In one direction, recom- A summary of the way in which recombinants from
binant frequencies can be used to make maps. In the other crossing over are used in mapping is shown in Figure 4-10.
direction, when given an established map with genetic dis- Crossovers occur more or less randomly along the chromo-
tance in map units, we can predict the frequencies of prog- some pair. In general, in longer regions, the average num-
eny in different classes. For example, the genetic distance ber of crossovers is higher and, accordingly, recombinants
between the pr and vg loci in Drosophila is approximately are more frequently obtained, translating into a longer map
11 m.u. So, knowing this value, we know that there will be distance.
11 percent recombinants in the progeny from a testcross
of a female dihybrid heterozygote in cis conformation KEY CONCEPT Recombination between linked genes can be
(pr vg /pr + vg + ). These recombinants will consist of two used to map their distance apart on a chromosome. The unit
of mapping (1m.u.) is defined as a recombinant frequency of
reciprocal recombinants of equal frequency: thus, 5.5 per-
1 percent. Map distances are roughly additive.
cent will be pr vg + and 5.5 percent will be pr + vg . We also
a b c
A B C
Meiocyte 2
a b c
A B C
Meiocyte 3
a b c
A B C
Meiocyte 4
…
etc.
a b c
Few Numerous
recombinants recombinants
Chromosome
map
A B C
locus locus locus
flies. The columns alongside show which genotypes are Double recombinants arising from two crossovers
recombinant (R) for the loci taken two at a time. We must
be careful in our classification of parental and recombinant v ct + cv +
types. Note that the parental input genotypes for the triple v ct + cv +
heterozygotes are v+ ⋅ cv ⋅ ct and v ⋅ cv+ ⋅ ct +; any combina-
tion other than these two constitutes a recombinant. v+ ct cv
v+ ct cv
Recombinant for loci
Gametes v and cv v and ct cv and ct
FIGURE 4-11 Example of a double crossover between two
v ⋅ cv+ ⋅ ct + 580 chromatids. Notice that a double crossover produces double-
recombinant chromatids that have the parental allele combinations at
v+ ⋅ cv ⋅ ct 592 the outer loci. The position of the centromere cannot be determined
from the data. It has been added for completeness.
v ⋅ cv ⋅ ct + 45 R R
+ +
v ⋅ cv ⋅ ct 40 R R
left and cv to the right, but the map could equally well be
v ⋅ cv ⋅ ct 89 R R
drawn with the placement of these loci inverted.
v+ ⋅ cv+ ⋅ ct + 94 R R Third, note that linkage maps merely map the loci in rela-
v ⋅ cv+ ⋅ ct 3 R R tion to one another, with the use of standard map units. We
do not know where the loci are on a chromosome—or even
v+ ⋅ cv ⋅ ct + 5 R R which specific chromosome they are on. In subsequent anal-
1448 268 191 93 yses, as more loci are mapped in relation to these three, the
complete chromosome map would become “fleshed out.”
Let’s analyze the loci two at a time, starting with the v and
cv loci. In other words, we look at just the first two columns KEY CONCEPT Three-point (and higher) testcrosses enable
under “Gametes” and cover up the third one. Because the geneticists to evaluate linkage between three (or more) genes
parentals for this pair of loci are v+ ⋅ cv and v ⋅ cv+, we know and to determine gene order, all in one cross.
that the recombinants are by definition v ⋅ cv and v+ ⋅ cv+.
There are 45 + 40 + 89 + 94 = 268 of these recombinants. Of A final point to note is that the two smaller map dis-
a total of 1448 flies, this number gives an RF of 18.5 percent. tances, 13.2 m.u. and 6.4 m.u., add up to 19.6 m.u., which is
For the v and ct loci, the recombinants are v ⋅ ct and greater than 18.5 m.u., the distance calculated for v and cv.
v+ ⋅ ct +. There are 89 + 94 + 3 + 5 = 191 of these recombi- Why? The answer to this question lies in the way in which
nants among 1448 flies, and so the RF = 13.2 percent. we have treated the two rarest classes of progeny (totaling
For ct and cv, the recombinants are cv ⋅ ct + and cv+ ⋅ ct . 8) with respect to the recombination of v and cv. Now that
There are 45 + 40 + 3 + 5 = 93 of these recombinants we have the map, we can see that these two rare classes are
among the 1448, and so the RF = 6.4 percent . in fact double recombinants, arising from two crossovers
Clearly, all the loci are linked, because the RF values are (Figure 4-11). However, when we calculated the RF value for
all considerably less than 50 percent. Because the v and cv v and cv, we did not count the v ct cv+ and v+ ct + cv geno-
loci have the largest RF value, they must be farthest apart; types; after all, with regard to v and cv, they are parental
therefore, the ct locus must lie between them. A map can be combinations (v cv+ and v+ cv). In light of our map, how-
drawn as follows: ever, we see that this oversight led us to underestimate the
v ct cv distance between the v and the cv loci. Not only should
we have counted the two rarest classes, we should have
13.2 6.4 counted each of them twice because each represents dou-
m.u. m.u.
ble recombinants. Hence, we can correct the value by add-
The testcross can be rewritten as follows, now that we ing the numbers 45 + 40 + 89 + 94 + 3 + 3 + 5 + 5 = 284 .
know the linkage arrangement: Of the total of 1448, this number is exactly 19.6 percent,
v+ ct cv /v ct + cv+ × v ct cv /v ct cv which is identical with the sum of the two component val-
ues. (In practice, we do not need to do this calculation,
Note several important points here. First, we have because the sum of the two shorter distances gives us the
deduced a gene order that differs from that used in our list best estimate of the overall distance.)
of the progeny genotypes. Because the point of the exer-
cise was to determine the linkage relation of these genes,
the original listing was of necessity arbitrary; the order was Deducing gene order by inspection
simply not known before the data were analyzed. Hence- Now that we have had some experience with the three-
forth, the genes must be written in correct order. point testcross, we can look back at the progeny listing
Second, we have definitely established that ct is between and see that, for trihybrids of linked genes, gene order can
v and cv. In the diagram, we have arbitrarily placed v to the usually be deduced by inspection, without a recombinant
frequency analysis. Typically, for linked genes, we have the frequency of double recombinants: that frequency would
eight genotypes at the following frequencies: equal the product of the recombinant frequencies in the
adjacent regions. In the v-ct-cv recombination data, the v-ct
two at high frequency
RF value is 0.132 and the ct-cv value is 0.064; so, if there is
two at intermediate frequency no interference, double recombinants might be expected at
two at a different intermediate frequency the frequency 0.132 × 0.064 = 0.0084 (0.84 percent). In the
two rare sample of 1448 flies, 0.0084 × 1448 = 12 double recombi-
nants are expected. But the data show that only 8 double
Only three gene orders are possible, each with a dif- recombinants were actually observed. If this deficiency of
ferent gene in the middle position. It is generally true that double recombinants were consistently observed, it would
the double-recombinant classes are the smallest ones. show us that the two regions are not independent and sug-
Only one order is compatible with the smallest classes gest that the distribution of crossovers favors singles at the
(v ct cv+ and v+ ct + cv) having been formed by dou- expense of doubles. In other words, there is some kind of
ble crossovers, as shown in Figure 4-12. A simple rule of interference: a crossover does reduce the probability of a
thumb for deducing the gene in the middle is to compare crossover in an adjacent region.
the gametes in the parental classes with the gametes in the Interference is quantified by first calculating a term
double-recombinant classes, and to identify the allele pair called the coefficient of coincidence (c.o.c.), which is the
that has moved in relation to the other two allele pairs ratio of observed to expected double recombinants. Inter-
that remain in the original parental arrangement. In other ference (I) is defined as 1 – c.o.c. Hence,
words, the gene in the middle is the allele pair that has
“flipped” position in the double-recombinant classes. observed frequency or
number of double recombinants
I = 1−
Interference expected frequency or
number of double recombinants
Knowing the existence of double crossovers permits us to ask
questions about their possible interdependence. We can ask, are In our example,
the crossovers in adjacent chromosome regions independent 8 4
I = 1 − 12 = 12
= 13 , or 33 percent
events, or does a crossover in one region affect the likelihood
of there being a crossover in an adjacent region? The answer In some regions, there are never any observed double
is that, generally, crossovers inhibit each other somewhat in an recombinants. In these cases, c.o.c. = 0, and so I = 1 and inter-
interaction called interference. Double-recombinant classes can ference is complete. Interference values anywhere between 0
be used to deduce the extent of this interference. and 1 are found in different regions and in different organisms.
Interference can be measured in the following way. A statistical approach to correcting map distances for
If the crossovers in the two regions are independent, we unseen multiple crossovers is shown in Box 4-1.
can use the product rule (see Chapter 3) to predict the You may have wondered why we always use het-
erozygous females for testcrosses in Drosophila.
Different gene orders give different double recombinants The explanation lies in an unusual feature of Dro-
sophila males. When, for example, pr vg /pr + vg +
Possible gene orders Double-recombinant chromatids males are crossed with pr vg /pr vg females, only
pr vg /pr + vg + and pr vg /pr vg progeny are recov-
v ct + cv + v ct cv + ered. This result shows that there is no crossing over
in Drosophila males. However, this absence of cross-
ing over in one sex is limited to certain species; it is
v+ ct cv v+ ct + cv not the case for males of all species (or for the het-
erogametic sex). In other organisms, there is cross-
ct + v cv + ct + v + cv + ing over in XY males and in WZ females. The reason
for the absence of crossing over in Drosophila males
is that they have an unusual prophase I, with no
ct v + cv ct v cv synaptonemal complexes (the molecular assemblages
generally visible between paired chromosomes).
Incidentally, there is a recombination difference between
ct + cv + v ct + cv v
human sexes as well. Women show higher recombinant
frequencies for the same autosomal loci than do men.
With the use of a reiteration of the preceding
ct cv v+ ct cv + v+
recombination-based techniques, maps have been
produced of thousands of genes for which variant
FIGURE 4-12 The three possible gene orders shown on the left yield the six
(mutant) phenotypes have been identified. A sim-
products of a double crossover shown on the right. Only the first possibility is
compatible with the data in the text. Note that only the nonsister chromatids ple illustrative example from the tomato is shown
taking part in the double crossover are shown. in Figure 4-13. The tomato chromosomes are shown
Any number of crossovers gives 50 percent (showing two nonparental genotypes). From a cross
recombinants AB × ab, they are:
5
9 LII
3
12
11 9SI
2
7 8SI 10
4
1 8I
5 7
(c) 1 2
23
Normal (F ) 23 Fasciated (f )
Normal (M ) 12 Mottled (m ) Red (R) Yellow (r ) Green-base Uniform fruit
(U) (u )
15
Purple (A) Green (a )
4 21
Tall (D) Dwarf (d) 18 Smooth (H) Hairy (h)
Yellow (Wf ) White (wf )
Hairy (H I ) Hairless (h l )
3 20
17 Non-tangerine 30 Tangerine (t )
Smooth (P) Peach (p)
(T)
2
Normal (L f ) 16 Leafy (lf )
30
Normal (O ) Oblate (o )
Normal (B r ) Brachytic (b r) Xanthophyllous Green (x a)
16 (X a / x a)
Jointed (J ) Jointless (j)
Woolly Normal (wo )
(Wo / wo ) 30 8
35
Yellow skin (Y) Clear skin (y ) Susceptibility Resistance to
to leaf mold leaf mold from
5 Resistance Susceptibility to (c f p2 ) Potentate #2
to leaf mold leaf mold from (C f p2 )
Normal (N e ) 4 Necrotic (n e) (C f s c ) Stirling Castle
(c f s c)
Normal (N t) Nipple-tip (n t )
14
FIGURE 4-13 (a) Photomicrograph of a meiotic prophase I (pachytene) from anthers, showing the
12
12 pairs of chromosomes. (b) Illustration of the 12 chromosomes shown in part a. The chromosomes
are identified by the currently used chromosome-numbering system. The centromeres are shown in
orange, and the flanking, densely staining regions (heterochromatin) in green. (c) 1952 linkage map.
Each locus is flanked by drawings of the normal and variant phenotypes. Interlocus map distances Normal Macrocalyx
(M c ) (m c)
are shown in map units. [From C. M. Rick, “The Tomato,” Scientific American, 1978.]
A a a A a A A and a segregate into separate nuclei at the second meiotic division when there is a
crossover between the centromere and the A locus.
A a a A a A
a A A a a A The other patterns are produced similarly; the differ-
a A A a a A ence is that the chromatids move in different directions at
the second division, as shown on the next page.
a A a A A a
You can see that the frequency of octads with an
a A a A A a MII pattern should be proportional to the size of the
126 132 9 11 10 12 centromere— A /a region and could be used as a measure
Total = 300 of the size of that region. In our example, the MII frequency
is 42/300 = 14 percent. Does this percentage mean that
The first two columns on the left are from meioses the A /a locus is 14 m.u. from the centromere? The answer
with no crossover in the region between the A locus and is no, but this value can be used to calculate the number of
the centromere. The letter M is used to stand for a type of map units. The 14 percent value is a percentage of meio-
segregation at meiosis. The patterns for the first two col- ses, which is not the way that map units are defined. Map
umns are called MI patterns, or first-division segregation units are defined as the percentage of recombinant chro-
patterns, because the two different alleles segregate into matids issuing from meiosis. Because a crossover in any
the two daughter nuclei at the first division of meiosis. The meiosis results in only 50 percent recombinant chromatids
A a A a
A a A a
a a A A a a A A
A A a a a a A A
a A A a
a A A a
1 2 3 4
In the second meiotic division, the centromeres attach to the spindle at random, producing the four arrangements
shown. The four arrangements are equally frequent.
(four out of eight), we must divide the 14 percent by 2 to region must be 7 m.u. in length, and this measurement can
convert the MII frequency (a frequency of meioses) into map be introduced into the map of that chromosome.
units (a frequency of recombinant chromatids). Hence, this
in Figure 4-13a, their numbering in Figure 4-13b, and Using ratios as diagnostics
recombination-based gene maps in Figure 4-13c. The chro-
The analysis of ratios is one of the pillars of genetics. In
mosomes are shown as they appear under the microscope,
the text so far, we have encountered many different ratios
together with chromosome maps based on linkage analysis
whose derivations are spread out over several chapters.
of various allelic pairs shown with their phenotypes.
Because recognizing ratios and using them in diagnosis of
In some organisms, recombination-based techniques
the genetic system under study are part of everyday genet-
can be used to map the positions of centromeres. Cen-
ics, let’s review the main ratios that we have covered so
tromeres are not genes, but they are regions of DNA on
far. They are shown in Figure 4-14. You can read the ratios
which the orderly reproduction of living organisms abso-
from the relative widths of the colored boxes in a row.
lutely depends and are therefore of great interest in genetics
Figure 4-14 deals with selfs and testcrosses of monohybrids,
(Box 4-2).
Dihybrid testcrossed
(independent assortment) 1:1:1:1
Dihybrid selfed
(independent assortment) 9:3:3:1
Dihybrid testcrossed
(linked) [Example only (P:R:R:P)]
Trihybrid testcrossed
(independent assortment)
1:1:1:1:1:1:1:1
Trihybrid testcrossed
(all linked) [Example only
(P:P:SCO:SCO:SCO:SCO:DCO:DCO)]
dihybrids (with independent assortment and linkage), and is a locus with a simple DNA sequence difference, and
D
trihybrids (also with independent assortment and link- R is a locus showing variable numbers of tandem repeats,
age of all genes). One situation not represented is a trihy-
then we might analyze a cross of the following type:
brid in which only two of the three genes are linked; as
an exercise, you might like to deduce the general pattern D1 R1/D2 R2 × D3 R3/D3 R3
that would have to be included in such a diagram from
from which recombinants would be
this situation. Note that, in regard to linkage, the sizes of
the classes depend on map distances. A geneticist deduces D1 R2/D3 R3
unknown genetic states in something like the following
and
way: “a 9:3:3:1 ratio tells me that this ratio was very likely
produced by a selfed dihybrid in which the genes are on D2 R1/D3 R3
different chromosomes.”
The frequency of recombinants would then be used to cal-
KEY CONCEPT Single gene inheritance and two-gene inher-
culate the map distance between the two loci.
itance (linked and unlinked) can be inferred from diagnostic Thus, molecular markers can be mapped in relation to
phenotypic ratios in both selfing and testcrossing. each other in this way, but a molecular marker can also be
mapped in relation to a phenotypic marker. The location
of the gene for the human disease cystic fibrosis was orig-
4.3 MAPPING WITH MOLECULAR inally discovered in this way through its linkage to molec-
ular markers known to be located on chromosome 7. This
MARKERS discovery led to the isolation and sequencing of the gene,
resulting in the further discovery that it encodes the pro-
So far in this chapter we have mapped loci of genes with
tein now called cystic fibrosis transmembrane conductance
alleles that result in visible phenotypic differences. One
regulator (CFTR). The gene for Huntington disease was
might call these loci phenotypic markers as they mark cer-
also located in this way, leading to the discovery that it
tain points on the chromosome that can produce visibly
encodes a muscle protein now called huntingtin. The gen-
different phenotypes in the outward appearance of prog-
eral experimental procedure might be as follows. Let A and
eny. However, molecular studies such as DNA sequencing
a be the disease-gene alleles and M1 and M2 be alleles of
have revealed many DNA differences between chromo-
a specific molecular-marker locus. Assume that the cross
somes that are neutral; that is, they do not seem to pro-
is A /a ⋅ M1/M2 × a /a ⋅ M1/M1, a kind of testcross. Prog-
duce visible phenotypic effects. These are generally called
eny would be first scored for the A and a phenotypes, and
molecular markers. Because they are so numerous, they are
then DNA would be extracted from each individual and
useful in “fleshing out” the chromosome map. They can be
sequenced or otherwise assessed to determine the molecular
compared to milestones that mark the distances along a
alleles. Assume that we obtain the following results:
well-traveled road; in themselves they are not particularly
interesting, but they can help in the process of locating A /a ⋅ M1/M1 49 percent A /a ⋅ M2/M1 1 percent
important towns—in this case, genes of interest that have
a /a ⋅ M2/M1 49 percent /
a a ⋅ M1/M1 1 percent
been shown to be relevant in a specific piece of genetic
research. Locating a gene of interest is an important step These results tell us that the testcross must have been in
toward isolating the gene and charting the general topogra- the following conformation:
phy of its chromosomal neighborhood.
A M1/a M2 × a M1/a M1
Here, we will focus on two general types of molecular
markers, and will revisit the subject in more detail in later chap- and the two progeny genotypes on the right in the list must
ters. (Hence, for the time being, we are treating them merely be recombinants, giving a map distance of 2 m.u. between
as milestones and ignoring their origin and detection.) First, the A /a locus and the molecular locus M1/M2. Hence, we
there are molecular markers that are merely loci showing neu- now know the general location of the gene in the genome
tral simple DNA sequence differences, perhaps a G-C base pair and can narrow its location down with more finely scaled
replaced by a T-A base pair. Second, there are markers that are approaches. Additional molecular markers can be mapped
loci showing variable numbers of tandem (adjacent) repeats of to each other, creating a map that can act like a series of
short, simple DNA sequences; for example, one chromosome stepping stones on the way to some gene with an interesting
might have five repeats at that locus, and the homolog might phenotype.
have eight. Both simple sequence differences and repeated In our examples above, we have been dealing with what
DNA differences are highly polymorphic; that is, there are are effectively testcrosses. However, because molecular
often many “alleles” of each marker in the population. markers do not show dominance or recessiveness, and can
Molecular markers can be mapped by recombinant fre- be scored directly in molecular tests, often crosses that are
quency in exactly the same way as phenotypic markers. If not testcrosses can be assessed for recombinants. Such an
we use D and R to represent two linked molecular marker analysis is shown in Figure 4-15. The figure shows experi-
loci, where mental details of how the molecular marker “alleles” are
1 2 3 4 5 6
M´´´´
M´´
M´
M´´´
PCR products
first law, so skewed allele ratios cannot be responsible for nonidentical sister-spore pair. Normally, postmeiotic repli-
the quite large deviations from the expected numbers of cation gives identical sister-spore pairs as follows: the A A
50 :50: 50: 50. a a tetrad becomes
We must apply the χ2 analysis to test a hypothesis of no
A-A A-A a-a a-a
linkage. If that hypothesis is rejected, we can infer linkage.
(We cannot test a hypothesis of linkage directly because (the hyphens show sister spores). In contrast, an aberrant
we have no way of predicting what recombinant frequency 5A : 3a octad must be
to test.) The calculation for testing lack of linkage is as
A-A A-A A-a a-a
follows:
In other words, there is one nonidentical sister-spore pair
Observed (O) Expected (E) O – E (O – E)2 (O – E)2 /E (in red).
60 50 10 100 2.00 The observation of a nonidentical sister-spore pair sug-
gests that the DNA of one of the final four meiotic homo-
37 50 −13 169 3.38
logs contains heteroduplex DNA. Heteroduplex DNA is
41 50 −9 81 1.62 DNA in which there is a mismatched nucleotide pair in the
62 50 12 144 2.88 gene under study. The logic is as follows. If in a cross of
2 2 A × a, one allele (A) is G :C and the other allele (a) is A : T,
χ = ∑ (O – E) /E for all classes = 9.88
the two alleles would usually replicate faithfully. However,
Since there are four genotypic classes, we must use a heteroduplex, which forms only rarely, would be a mis-
4 − 1 = 3 degrees of freedom. Consulting the chi-square matched nucleotide pair such as G : T or A :C (effectively,
table in Chapter 3 (Table 3-1), we see our values of 9.88 a DNA molecule bearing both A and a information). Note
and 3 df give a p value of ∼ 0.025, or 2.5 percent. This is that a heteroduplex involves only one nucleotide position:
less than the standard cut-off value of 5 percent, so we can the surrounding DNA segment might be as follows, where
reject the hypothesis of no linkage. Hence, we are left with the heteroduplex site is shown in red:
the conclusion that the genes are very likely linked, approx-
imately 39 m.u. apart. GCTAATGTTATTAG
Notice, in retrospect, that it was important to make sure CGATTATAATAATC
alleles were segregating 1:1 to avoid a compound hypothe-
sis of 1:1 allele ratios and no linkage. If we rejected such a At replication to form an octad, a G : T heteroduplex
compound hypothesis, we would not know which part of it would pull apart and replicate faithfully, with G bonding to
was responsible for the rejection. C and A bonding to T. The result would be a nonidentical
spore pair of G :C (allele A) and A : T (allele a).
KEY CONCEPT The chi-square test is useful in testing the Nonidentical sister spores (and aberrant octads gener-
significance of deviations from a 1: 1: 1: 1 ratio in deducing ally) were found to be statistically correlated with cross-
linkage between two genes. ing over in the region of the gene concerned, providing an
important clue that crossing over might be based on the
formation of heteroduplex DNA.
4.5 THE MOLECULAR MECHANISM In the currently accepted model (follow it in
OF CROSSING OVER Figure 4-16), the heteroduplex DNA and a crossover are
both produced by a double-stranded break in the DNA of
LO 4.6 Diagram how double-strand DNA breaks and one of the chromatids participating in the crossover. Let’s
heteroduplex formation can lead to a crossover at see how that works. Molecular studies show that broken
the molecular level. ends of DNA will promote recombination between dif-
ferent chromatids. In step 1, both strands of a chromatid
In this chapter we have analyzed the genetic consequences of break in the same location. From the break, DNA is eroded
the cytologically visible process of crossing over without wor- at the 5′ end of each broken strand, leaving both 3′ ends
rying about the mechanism of crossing over. However, crossing single stranded (step 2). One of the single strands “invades”
over is remarkable in itself as a molecular process: how can two the DNA of the other participating chromatid; that is, it
large coiled molecules of DNA exchange segments with a preci- enters the center of the helix and base-pairs with its homol-
sion so exact that no nucleotides are lost or gained? ogous sequence (step 3), displacing the other strand. Then
Studies on fungal octads gave a clue. Although most the tip of the invading strand uses the adjacent sequence
octads show the expected 4 : 4 segregation of alleles as a template for new polymerization, which proceeds by
such as 4 A : 4 a , some rare octads show aberrant ratios. forcing the two resident strands of the helix apart (step 4).
There are several types, but as an example we will use The displaced single-stranded loop hydrogen bonds with
5:3 octads (either 5A :3a or 5a :3A). Two things are pecu- the other single strand (the blue one in Figure 4-16). If the
liar about this ratio. First, there is one too many spores of invasion and strand displacement spans a site of heterozy-
one allele and one too few of the other. Second, there is a gosity (such as A /a), then a region of heteroduplex DNA
G
5' 3'
3' 5'
C
T
3' 5'
5' 3'
A
Erosion
Invasion and
displacement
Polymerization
Heteroduplex region
Resolution to
crossover by nicks ( )
G
T
T
FIGURE 4-16 A molecular model of crossing over. Only the two chromatids A N I M ATED A RT
(blue and red) participating in the crossover are shown. The 3′-to-5′ strand
A molecular model of crossing over
is placed on the inside of both for clarity. The chromatids differ at one site,
GC, in one allele (perhaps allele A) and AT in the other (perhaps a). Only the
outcome with mispaired heteroduplex DNA and a crossover are shown. The
final crossover products are shaded in yellow and blue.
is formed. Replication also takes place from the other result is a strange structure with two single-stranded junc-
single-stranded end to fill the gap left by the invading tions called Holliday junctions after their original proposer,
strand (also shown on the upper blue strand in step 4 of Robin Holliday. These junctions are potential sites of sin-
Figure 4-16). The replicated ends are sealed, and the net gle-strand breakage and reunion; two such events, shown
by the darts in the figure, then lead to a complete dou- here. A physical map is simply a map of the actual genomic
ble-stranded crossover (step 5). DNA, a very long DNA nucleotide sequence, showing
Note that when the invading strand uses the invaded where genes are, their sequence, how big they are, what is
DNA as a replication template, this automatically results in between them, and other landmarks of interest. The units
an extra copy of the invaded sequence at the expense of the of distance on a physical map are numbers of DNA bases;
invading sequence, thus explaining the departure from the for convenience, the kilobase is the preferred unit. The com-
expected 4: 4 ratio. plete sequence of a DNA molecule is obtained by sequenc-
This same sort of recombination takes place at many ing large numbers of small genomic fragments and then
different chromosomal sites where the invasion and strand assembling them into one whole sequence. The sequence is
displacement do not span a heterozygous mutant site. Here, then scanned by a computer, programmed to look for gene-
DNA would be formed that is heteroduplex in the sense like segments recognized by characteristic base sequences
that it is composed of strands of each participating chro- including known signal sequences for the initiation and
matid, but there would not be a mismatched nucleotide termination of transcription. When the computer’s pro-
pair, and the resulting octad would contain only identical gram finds a gene, it compares its sequence with the pub-
spore pairs. Those rare occasions in which the invasion lic database of other sequenced genes for which functions
and polymerization do span a heterozygous site are simply have been discovered in other organisms. In many cases,
lucky cases that provided the clue for the mechanism of there is a “hit”; in other words, the sequence closely resem-
crossing over. bles that of a gene of known function in another species.
In such cases, the functions of the two genes also may be
KEY CONCEPT A crossover is initiated by a double-stranded similar. The sequence similarity (often close to 100 percent)
break in the DNA of a chromatid at meiosis. A series of molec- is explained by the inheritance of the gene from some com-
ular events ensues that eventually produces crossover DNA mon ancestor and the general conservation of functional
molecules. In addition, if the site of the crossover happens sequences through evolutionary time. Other genes discov-
to be near a site of DNA heterozygosity in meiosis, aberrant
ered by the computer show no sequence similarity to any
non-Mendelian allele ratios for the heterozygous site may be
gene of known function. Hence, they can be considered
produced.
“genes in search of a function.” In reality, of course, it is the
researcher, not the gene, who searches and who must find
the function. Sequencing different individual members of a
population also can yield sites of molecular heterozygosity,
4.6 USING RECOMBINATION- which, just as they do in recombination maps, act as orien-
tation markers on the physical map.
BASED MAPS IN Because physical maps are now available for most of
CONJUNCTION WITH the main genetic model organisms, is there really any need
PHYSICAL MAPS for recombination maps? Could they be considered out-
moded? The answer is that both maps are used in con-
Recombination maps have been the main topic of this chap- junction with each other to “triangulate” in determining
ter. They show the loci of genes for which mutant alleles gene function, a principle illustrated earlier by the London
(and their mutant phenotypes) have been found. The posi- maps. The general approach is illustrated in Figure 4-17,
tions of these loci on a map is determined on the basis of which shows a physical map and a recombination map of
the frequency of recombinants at meiosis. The frequency of the same region of a genome. Both maps contain genes and
recombinants is assumed to be proportional to the distance molecular markers. In the lower part of Figure 4-17, we see
apart of two loci on the chromosome; hence, recombinant a section of a recombination-based map, with positions of
frequency becomes the mapping unit. Such recombination- genes for which mutant phenotypes have been found and
based mapping of genes with known mutant phenotypes mapped. Not all the genes in that segment are included.
has been done for nearly a century. We have seen how sites For some of these genes, a function may have been discov-
of molecular heterozygosity (unassociated with mutant ered on the basis of biochemical or other studies of mutant
phenotypes) also can be incorporated into such recombi- strains; genes for proteins A and B are examples. The gene
nation maps. Like any heterozygous site, these molecular in the middle is a “gene of interest” that a researcher has
markers are mapped by recombination and then used to found to affect the aspect of development being studied.
navigate toward a gene of biological interest. We make the To determine its function, the physical map can be use-
perfectly reasonable assumption that a recombination map ful. The genes in the physical map that are in the general
represents the arrangement of genes on chromosomes, but, region of the gene of interest on the recombination map
as stated earlier, these maps are really hypothetical con- become candidate genes, any one of which could be the
structs. In contrast, physical maps are as close to the real gene of interest. Further studies are needed to narrow the
genome map as science can get. choice to one. If that single case is a gene whose function is
The topic of physical maps will be examined more known for other organisms, then a function for the gene of
closely in Chapters 10 and 14, but we can foreshadow it interest is suggested. In this way, the phenotype mapped on
Recombination
map
FIGURE 4-17 Comparison of relative positions on physical and recombination maps can connect
phenotype with an unknown gene function.
the recombination map can be tied to a function deduced melded to understand the gene’s contribution to the devel-
from the physical map. Molecular markers on both maps opment of the organism.
(not shown in Figure 4-17) can be aligned to help in the There are several other genetic-mapping techniques,
zeroing-in process. Hence, we see that both maps contain some of which we will encounter in Chapters 6, 18, and 19.
elements of function: the physical map shows a gene’s pos-
sible action at the cellular level, whereas the recombination KEY CONCEPT The union of recombination and physical
map contains information related to the effect of the gene maps can ascribe biochemical function to a gene identified by
at the phenotypic level. At some stage, the two have to be its mutant phenotype.
SUMMARY
In a dihybrid testcross in Drosophila, Thomas Hunt wondered if these values corresponded to the actual dis-
Morgan found a deviation from Mendel’s law of indepen- tances between genes on a chromosome. Alfred Sturtevant,
dent assortment. He postulated that the two genes were a student of Morgan’s, developed a method of determining
located on the same pair of homologous chromosomes. This the distance between genes on a linkage map, based on the
relation is called linkage. RF. The easiest way to measure RF is with a testcross of a
Linkage explains why the parental gene combinations dihybrid or trihybrid. RF values calculated as percentages
stay together, but not how the recombinant (nonparental) can be used as map units to construct a chromosomal map
combinations arise. Morgan postulated that, in meiosis, showing the loci of the genes analyzed. In ascomycete fungi,
there may be a physical exchange of chromosome parts by centromeres also can be located on the map by measuring
a process now called crossing over. A result of the physical second-division segregation frequencies.
breakage and reunion of chromosome parts, crossing over Single nucleotide differences in sequences, and dif-
takes place at the four-chromatid stage of meiosis. Thus, ferences in the number of repeating units, can be used as
there are two types of meiotic recombination. Recombi- molecular markers for mapping genes.
nation by Mendelian independent assortment results in a Although the basic test for linkage is deviation from
recombinant frequency of 50 percent. Crossing over results independent assortment, such a deviation may not be obvi-
in a recombinant frequency (RF) of generally less than ous in a testcross, and a statistical test is needed. The χ2
50 percent. test, which tells how often observations deviate from expec-
As Morgan studied more linked genes, he discovered tations purely by chance, is particularly useful in determin-
many different values for recombinant frequency and ing whether loci are linked.
The mechanism of crossing over is thought to start with In genetics generally, the recombination-based map of
a double-stranded break in one participating chromatid. Ero- loci conferring mutant phenotypes is used in conjunction
sion leaves the ends single stranded. One single strand invades with a physical map such as the complete DNA sequence,
the double helix of the other participating chromatid, leading which shows all the gene-like sequences. Knowledge of gene
to the formation of heteroduplex DNA. Gaps are filled by position in both maps enables the melding of cellular func-
polymerization. The molecular resolution of this structure tion with a gene’s effect on phenotype.
becomes a full double-stranded crossover at the DNA level.
KEY TERMS
centimorgan (cM) (p. 120) genetic map unit (m.u.) (p. 120) Poisson distribution (p. 125)
chromosome map (p. 115) heteroduplex DNA (p. 132) recombinant frequency (RF)
cis conformation (p. 117) interference (p. 124) (p. 120)
coefficient of coincidence (c.o.c.) linkage map (p. 120) recombination map (p. 115)
(p. 124) linked (p. 115) second-division segregation pattern
crossing over (p. 117) locus (p. 115) (M II pattern) (p. 128)
crossover product (p. 117) mapping function (p. 125) three-point testcross (three factor-
double-stranded break (p. 132) molecular marker (p. 130) testcross) (p. 122)
first-division segregation pattern octad (p. 128) trans conformation (p. 117)
(M I pattern) (p. 128) physical map (p. 134)
SOLVED PROBLEMS
1 2
I
i/i I B/i
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
II
i/i I B/i I B/i i/i I B/i I B/i I A/i i/i i/i i/i I B/i i/i i/i i/i I B/i I B/i
1 2 3 4 5
III
I B/i I B/i I B/i I A/i I A/i
}ni
B nail–patella
NI 41.5% ← b ⋅ wx ⋅ cn 6
83.0% blood type B
41.5% + + + 5
b ⋅ wx ⋅ cn
17.0% } nNIi B
8.5%
8.5%
1000
Now, we can calculate the recombinant frequen- b. The parental chromosomes in the triple heterozygote are
cies. For b–wx, b cn wx
69 + 67 + 48 + 44
RF = = 22.8%
1000 b cn wx
for b–cn, c. The expected number of double recombinants is
0.103 × 0.147 × 1000 = 15.141. The observed number is
48 + 44 + 6 + 5 6 + 5 = 11, and so interference can be calculated as
RF = = 10.3%
1000
I = 1 − (11/15.141) = 1 − 0.726 = 0.274 = 27.4%
and for wx–cn,
SOLVED PROBLEM 3
69 + 67 + 6 + 5
RF = = 14.7% A cross is made between a haploid strain of Neurospora of
1000
genotype nic+ ad and another haploid strain of genotype
The map is therefore nic ad + . From this cross, a total of 1000 linear asci are iso-
b cn w lated and categorized as in the table below. Map the ad and
nic loci in relation to centromeres and to each other. (Note:
10.3 m.u. 14.7 Information from Box 4-2 is required for this problem.)
1 2 3 4 5 6 7
+ + + + + + + + +
nic ⋅ ad nic ⋅ ad nic ⋅ ad nic ⋅ ad nic ⋅ ad nic ⋅ ad nic ⋅ ad +
+
SOLUTION ad
What principles can we draw on to solve this problem? It is
a good idea to begin by doing something straightforward, 9.30 m.u.
which is to calculate the two locus-to-centromere distances.
We do not know if the ad and the nic loci are linked, but we Now we have to put the two together and decide
do not need to know. The frequencies of the M II patterns for between the following alternatives, all of which are compat-
each locus give the distance from locus to centromere. (We ible with the preceding locus-to-centromere distances:
can worry about whether it is the same centromere later.) a. nic ad
Remember that an M II pattern is any pattern that is not
two blocks of four. Let’s start with the distance between the
nic locus and the centromere. All we have to do is add the 5.05 m.u. 9.30 m.u.
ascus types 4, 5, 6, and 7, because all of them are M II pat- b. nic ad
terns for the nic locus. The total is 5 + 90 + 1 + 5 = 101 of
1000, or 10.1 percent. In this chapter, we have seen that to 5.05 m.u. 9.30 m.u.
convert this percentage into map units, we must divide by 2,
c. nic ad
which gives 5.05 m.u.
nic 5.05 m.u.
We do the same for the ad locus. Here, the total of Here, a combination of common sense and simple anal-
the M II patterns is given by types 3, 5, 6, and 7 and is ysis tells us which alternative is correct. First, an inspection
90 + 90 + 1 + 5 = 186 of 1000, or 18.6 percent, which is of the asci reveals that the most common single type is the
9.3 m.u. one labeled 1, which contains more than 80 percent of all
the asci. This type contains only nic+ ⋅ ad and nic ⋅ ad + gen- Is the map distance from nic to ad simply 9.30 − 5.05 =
otypes, and they are parental genotypes. So we know that 4.25 m.u.? Close, but not quite. The best way of calculat-
recombination is quite low and the loci are certainly linked. ing map distances between loci is always by measuring the
This rules out alternative a. recombinant frequency. We could go through the asci and
Now consider alternative c. If this alternative were cor- count all the recombinant ascospores, but using the for-
rect, a crossover between the centromere and the nic locus mula RF = T + NPD is simpler. The T asci are classes 3,
would generate not only an M II pattern for that locus, but 4, and 7, and the NPD asci are classes 2 and 6. Hence,
also an M II pattern for the ad locus, because it is farther RF = [(100) + 2]/1000 = 5.2 percent, or 5.2 m.u., and a bet-
from the centromere than nic is. The ascus pattern produced ter map is
by a crossover between nic and the centromere in alterna-
nic ad
tive c should be
nic ad 5.05 m.u. 5.2 m.u.
nic ad
10.25 m.u.
nic ad nic ad
The reason for the underestimation of the ad-to-cen-
nic ad nic ad
tromere distance calculated from the M II frequency is the
nic ad nic ad occurrence of double crossovers, which can produce an M I
nic ad pattern for ad, as in ascus type 4:
nic ad
nic ad
nic ad
nic ad
nic ad
nic ad nic ad
Remember that the nic locus shows M II patterns in asci
types 4, 5, 6, and 7 (a total of 101 asci); of them, type 5 is nic ad nic ad
the very one that we are talking about and contains 90 asci. nic ad nic ad
Therefore, alternative c appears to be correct because ascus
nic ad
type 5 comprises about 90 percent of the M II asci for the nic ad
nic locus. This relation would not hold if alternative b were
nic ad
correct because crossovers on either side of the centromere
would generate the M II patterns for the nic and the ad loci nic ad
independently.
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a Problem Solving Video. Problems with the icon have an Unpacking the Problem exercise.
www
www
WORKING WITH THE FIGURES 5. a. Redraw Figure 4-5 with the heterozygous alleles in
(The first 20 questions require inspection of text figures.) trans configuration.
1. (i) In Figure 4-1, under which street do you think the b. Is the outcome from the upper case possible from
red underground line lies? a meiosis in which crossing over takes place at the
four strand stage?
(ii) Using the underground map, determine where
Oxford Circus is on the street map. 6. Redraw Figure 4-6 part b with one crossover occurring
between chromatids 2 and 3, and the other between 2
2. Redraw Figure 4-2 with the heterozygous alleles in the
and 4, and list the meiotic products.
trans configuration.
7. Redraw Figure 4-7 with the heterozygous alleles
3. In Figure 4-3, would there be any meiotic products that
in trans configuration, and state genotypes of the
did not undergo a crossover in the meiosis illustrated?
recombinants.
If so, what colors would they be in the color convention
used? 8. a. In Figure 4-8, some meiotic products are labeled
parental. Which parent is being referred to in this
4. In Figure 4-4, can you tell if the three crossovers on the
terminology?
long chromosome all involve the same two chromatids?
b. Redraw the figure with the heterozygous alleles in 18. For the second figure in Box 4-2, draw the outcome if
trans configuration and then compare the recombi- each pair of arrow is “flipped.” Does this change sig-
nant genotypes with those from this figure. nificantly affect the outcomes?
9. a. In Figure 4-9, why is only locus A shown in a con- 19. a. In Figure 4-16, let GC = A and AT = a, then draw
stant position? the fungal octad that would result from the final
b. In an actual experiment 50 progeny are analyzed structure (5).
and 2 are found to be recombinant. What is the b. (Challenging) Insert some closely linked flank-
likely gene order? ing markers into the diagram, say P /p to the left
10. In Figure 4-10 , what is the mean frequency of and Q /q to the right (assume either cis or trans
crossovers per meiosis in the region A–B? The region arrangements). Assume neither of these loci show
B –C ? non-Mendelian segregation. Then draw the final
octad based on the structure in part 5.
11. a. In Figure 4-11, is it true to say that from such a
cross the product v cv+ can have two different 20. In Figure 4-17, suggest a way in which you might be
origins? able to decide between which of the five or six candi-
date genes that might actually be the central locus on
b. Redraw the figure showing a double recombinant the recombination map.
arising from a three-chromatid double crossover.
12. Redraw Figure 4-12 using parental chromosomes A B BASIC PROBLEMS
C and a b c.
21. A plant of genotype
13. Point to the approximate location of the wooly locus
A B
in Figure 4-13b.
14. a. In Figure 4-14, in the bottom row, four colors are a b
labeled SCO. Why are they not all the same size is testcrossed with
(frequency)? a b
b. Consider the three cases in which there are four a b
progeny types and in words state the different
If the two loci are 10 m.u. apart, what proportion of
diagnostics they represent.
progeny will be AB /ab?
c. Consider the two cases in which there are eight
22. The A locus and the D locus are so tightly linked that
progeny types, and in words state the different
no recombination is ever observed between them.
diagnostics they represent.
If Ad /Ad is crossed with aD /aD and the F1 is inter-
15. a. Using the conventions of Figure 4-15, draw parents crossed, what phenotypes will be seen in the F2, and in
and progeny classes from a cross what proportions?
P M ′′′/p M ′ × p M ′/p M ′′′′ 23. The R and S loci are 35 m.u. apart. If a plant of
genotype
b. Why are there no M′M″ individuals in the progeny?
R S
c. Are there any exceptions to the rule that P is asso-
ciated with M ′′ ? r s
d. In the gel, why are there only two DNA bands is selfed, what progeny phenotypes will be seen, and in
shown for each progeny individual? what proportions?
16. a. For the figure in Box 4-1, what would be the RF 24. The cross E /E ⋅ F /F × e /e ⋅ f /f is made, and the F1 is then
between A /a and B /b in a cross in which purely by backcrossed with the recessive parent. The progeny gen-
chance all meioses had four-strand double cross- otypes are inferred from the phenotypes. The progeny
overs in that region? genotypes, written as the gametic contributions of the
b. Redraw the two-crossover section, making the con- heterozygous parent, are in the following proportions:
stant crossover between chromatids 1 and 3. Does E⋅F 2
6
this affect the main conclusion of that part of the
E⋅ f 1
figure? 6
e⋅F 1
17. For the first figure in Box 4-2, draw the arrangements 6
of alleles in an octad from a similar meiosis in which e⋅f 2
6
the upper product of the first division is segregated in
an upside-down manner at the second division. Explain these results.
25. A strain of Neurospora with the genotype H ⋅ I is 29. A fruit fly of genotype B R /b r is testcrossed with
crossed with a strain with the genotype h ⋅ i. Half the b r /b r. In 84 percent of the meioses, there are no chi-
progeny are H ⋅ I , and the other half are h ⋅ i. Explain asmata between the linked genes; in 16 percent of the
how this outcome is possible. meioses, there is one chiasma between the genes. What
26. A female animal with genotype A /a ⋅ B /b is crossed proportion of the progeny will be B r /b r?
with a double-recessive male (a /a ⋅ b /b) . Their progeny 30. A three-point testcross was made in corn. The results and
include 442 A /a ⋅ B /b, 458 a /a ⋅ b /b , 46 A /a ⋅ b /b , and a recombination analysis are shown in the display below,
54 a /a ⋅ B /b . Explain these results. which is typical of three-point testcrosses (p = purple
27. If A /A ⋅ B /B is crossed with a /a ⋅ b /b and the F1 is test- leaves, + = green; v = virus-resistant seedlings, + = sen-
crossed, what percentage of the testcross progeny sitive; b = brown midriff to seed, + = plain). Study the
will be a /a ⋅ b /b if the two genes are (a) unlinked; display, and answer parts a through c.
(b) completely linked (no crossing over at all); (c) 10 m.u. P + /+ ⋅ +/+ ⋅ +/+× p /p ⋅ v /v ⋅ b /b
apart; (d) 24 m.u. apart?
Gametes +⋅+⋅+ p⋅v⋅b
28. In a haploid organism, the C and D loci are 8 m.u.
apart. From a cross C d × c D, give the proportion of a. Determine which genes are linked.
each of the following progeny classes: (a) C D; (b) c d; b. Draw a map that shows distances in map units.
(c) C d; (d) all recombinants combined. c. Calculate interference, if appropriate.
Recombinant for
Class Progeny phenotypes F1 gametes Numbers p-b p-v v-b
1 gre sen pla +⋅+⋅+ 3210
2 pur res bro p⋅v⋅b 3222
3 gre res pla +⋅v ⋅+ 1024 R R
4 pur sen bro p⋅+⋅b 1044 R R
5 pur res pla p⋅v ⋅+ 690 R R
6 gre sen bro +⋅+⋅b 678 R R
7 gre res bro +⋅v⋅b 72 R R
8 pur sen pla p⋅+⋅+ 60 R R
Total 10,000 1,500 2,200 3,436
www
UNPACKING PROBLEM 30 7. What does the line marked “Gametes” represent,
www
and how is it different from the column marked “F1
Before attempting a solution to this problem, try answer-
gametes”? In what way is comparison of these two
ing the following questions:
types of gametes relevant to recombination?
1. Sketch cartoon drawings of the P, F1, and tester corn 8. Which meiosis is the main focus of study? Label it
plants, and use arrows to show exactly how you on your drawing.
would perform this experiment. Show where seeds
are obtained. 9. Why are the gametes from the tester not shown?
2. Why do all the +’s look the same, even for different 10. Why are there only eight phenotypic classes? Are
genes? Why does this not cause confusion? there any classes missing?
3. How can a phenotype be purple and brown, for 11. What classes (and in what proportions) would be
example, at the same time? expected if all the genes are on separate chromo-
4. Is it significant that the genes are written in the somes?
order p-v-b in the problem? 12. To what do the four pairs of class sizes (very big,
5. What is a tester, and why is it used in this analysis? two intermediates, very small) correspond?
6. What does the column marked “Progeny pheno- 13. What can you tell about gene order simply
types” represent? In class 1, for example, state by inspecting the phenotypic classes and their
exactly what “gre sen pla” means. frequencies?
b. What is the coefficient of coincidence? 35. In Drosophila, the allele dp+ determines long wings and
dp determines short (“dumpy”) wings. At a separate
locus, e+ determines gray body and e determines ebony 38. In the two pedigrees diagrammed here, a vertical bar
body. Both loci are autosomal. The following crosses in a symbol stands for steroid sulfatase deficiency, and
were made, starting with pure-breeding parents: a horizontal bar stands for ornithine transcarbamylase
P long, ebony♀ × short, gray♂ deficiency.
F2 long, ebony 54
long, gray 47
short, gray 52 II II
1 2 1 2 3
short, ebony 47
200
Use the χ2 test to determine if these loci are linked. In
III III
doing so, indicate (a) the hypothesis, (b) calculation of 1 2 3 4 5 6 1 2 3
χ2, (c) p value, (d) what the p value means, (e) your
conclusion, and (f) the inferred chromosomal constitu-
tions of parents, F1, tester, and progeny.
IV
36. The mother of a family with 10 children has blood 1 2
type Rh+ . She also has a very rare condition (ellipto-
cytosis, phenotype E) that causes red blood cells to be a. Is there any evidence in these pedigrees that the
oval rather than round in shape but that produces no genes determining the deficiencies are linked?
adverse clinical effects. The father is Rh− (lacks the b. If the genes are linked, is there any evidence in the
Rh+ antigen) and has normal red blood cells (phe- pedigree of crossing over between them?
notype e). The children are 1 Rh+ e, 4 Rh+ E, and
c. Assign genotypes of these individuals as far as
5 Rh− e. Information is available on the mother’s par-
possible.
ents, who are Rh+ E and Rh− e. One of the 10 children
(who is Rh+ E) marries someone who is Rh+ e, and 39. In the accompanying pedigree, the vertical lines stand
they have an Rh+ E child. for protan color blindness, and the horizontal lines
stand for deutan color blindness. These are separate
a. Draw the pedigree of this whole family.
conditions causing different misperceptions of colors;
b. Is the pedigree in agreement with the hypoth- each is determined by a separate gene.
esis that the Rh+ allele is dominant and Rh− is
I
recessive? 1 2
c. What is the mechanism of transmission of ellipto-
cytosis?
d. Could the genes governing the E and Rh pheno- II
1 2 3 4 5
types be on the same chromosome? If so, estimate
the map distance between them, and comment on
your result.
III
37. From several crosses of the general type A /A ⋅ B /B × 1 2 3 4 5
a /a ⋅ b /b , the F1 individuals of type A /a ⋅ B /b were test-
crossed with a /a ⋅ b /b . The results are as follows: a. Does the pedigree show any evidence that the genes
are linked?
Testcross progeny
b. If there is linkage, does the pedigree show any evi-
Testcross of A /a ⋅ a /a ⋅ A /a ⋅ a /a ⋅ dence of crossing over?
F1 from cross B /b b /b b /b B /b
Explain your answers to parts a and b with the aid of
1 310 315 287 288 the diagram.
2 36 38 23 23 c. Can you calculate a value for the recombination
3 360 380 230 230 between these genes? Is this recombination by
4 74 72 50 44 independent assortment or by crossing over?
40. In corn, a triple heterozygote was obtained carrying the
For each set of progeny, use the χ2 test to decide if mutant alleles s (shrunken), w (white aleurone), and y
there is evidence of linkage. (waxy endosperm), all paired with their normal wild-type
alleles. This triple heterozygote was testcrossed, and the 42. In the tiny model plant Arabidopsis, the recessive allele
progeny contained 116 shrunken, white; 4 fully wild type; hyg confers seed resistance to the drug hygromycin,
2538 shrunken; 601 shrunken, waxy; 626 white; 2708 and her, a recessive allele of a different gene, confers
white, waxy; 2 shrunken, white, waxy; and 113 waxy. seed resistance to herbicide. A plant that was homo-
a. Determine if any of these three loci are linked and, zygous hyg /hyg ⋅ her /her was crossed with wild type,
if so, show map distances. and the F1 was selfed. Seeds resulting from the F1 self
were placed on petri dishes containing hygromycin
b. Show the allele arrangement on the chromosomes and herbicide.
of the triple heterozygote used in the testcross.
a. If the two genes are unlinked, what percentage of
c. Calculate interference, if appropriate. seeds are expected to grow?
41. a. A mouse cross A /a ⋅ B /b × a /a ⋅ b /b is made, and in b. In fact, 13 percent of the seeds grew. Does this
the progeny there are percentage support the hypothesis of no linkage?
25% A /a ⋅ B /b, 25% a /a ⋅ b /b, Explain. If not, calculate the number of map units
between the loci.
25% A /a ⋅ b /b, 25% a /a ⋅ B /b
c. Under your hypothesis, if the F1 is testcrossed, what
Explain these proportions with the aid of simplified
proportion of seeds will grow on the medium con-
meiosis diagrams.
taining hygromycin and herbicide?
b. A mouse cross C /c ⋅ D /d × c /c ⋅ d /d is made, and in
43. In a diploid organism of genotype A /a ; B /b ; D /d , the
the progeny there are
allele pairs are all on different chromosome pairs. The
45% C /c ⋅ d /d , 45% c /c ⋅ D /d , diagrams below purport to show anaphases (“pulling
5% c /c ⋅ d /d , 5% C /c ⋅ D /d apart” stages) in individual cells. State whether each
drawing represents mitosis, meiosis I, or meiosis II or
Explain these proportions with the aid of simplified is impossible for this particular genotype.
meiosis diagrams.
a. A B D
f.
A b
a b d d
A B D
d
a b a B
d
b. g.
A b d a b
A b d d
A b d
d
A b d a b
c. h.
A B d
A A B B D D
A b d A A B B D D
a B D a a b b d d
a a b b d d
a b D
d. i.
A b D A a B b D d
A b D
a B d
a B d A a B b D d
e. j.
A B A a B b D d
D
D
A B a A b B d D
44. The Neurospora cross al -2+ × al -2 is made. A linear predicted to have (a) no crossovers? (b) one crossover?
tetrad analysis reveals that the second-division segre- (c) two crossovers? (Note: Information from Box 4-1
gation frequency is 8 percent. is required for this problem.)
a. Draw two examples of second-division segregation 47. A Neurospora cross was made between a strain that
patterns in this cross. carried the mating-type allele A and the mutant allele
b. What can be calculated by using the 8 percent arg-1 and another strain that carried the mating-type
value? allele a and the wild-type allele for arg -1(+). Four hun-
dred linear octads were isolated, and they fell into the
( Note: Information from Box 4-2 is required for this seven classes given in the table below. For simplicity,
problem.) they are shown as tetrads. (Note: Information from
45. From the fungal cross arg -6 ⋅ al -2 × arg -6+ ⋅ al -2+ , Box 4-2 is required for this problem.)
what will the spore genotypes be in unordered tetrads a. Deduce the linkage arrangement of the mating-type
that are (a) parental ditypes? (b) tetratypes? (c) non- locus and the arg-1 locus. Include the centromere
parental ditypes? (Note: Information from Box 4-2 is or centromeres on any map that you draw. Label
required for this problem.) all intervals in map units.
46. For a certain chromosomal region, the mean number b. Diagram the meiotic divisions that led to class 6.
of crossovers at meiosis is calculated to be two per Label clearly.
meiosis. In that region, what proportion of meioses are
1 2 3 4 5 6 7
A ⋅ arg A⋅+ A ⋅ arg A ⋅ arg A ⋅ arg A⋅+ A⋅+
A ⋅ arg A⋅+ A⋅+ a ⋅ arg a⋅+ a ⋅ arg a ⋅ arg
a⋅+ a ⋅ arg a ⋅ arg A⋅+ A ⋅ arg A⋅+ A ⋅ arg
a⋅+ a ⋅ arg a⋅+ a⋅+ a⋅+ a ⋅ arg a⋅+
www
UNPACKING PROBLEM 47 12. How is a cross made in a fungus such as Neuros-
www
pora? Explain how to isolate asci and individual
Before attempting a solution to this problem, try answer-
ascospores. How does the term tetrad relate to the
ing the following questions:
terms ascus and octad?
1. Are fungi generally haploid or diploid? 13. Where does meiosis take place in the Neurospora
2. How many ascospores are in the ascus of Neuros- life cycle? (Show it on a diagram of the life cycle.)
pora? Does your answer match the number pre- 14. What does Problem 47 have to do with meiosis?
sented in this problem? Explain any discrepancy.
15. Can you write out the genotypes of the two paren-
3. What is mating type in fungi? How do you think it tal strains?
is determined experimentally?
16. Why are only four genotypes shown in each class?
4. Do the symbols A and a have anything to do with
dominance and recessiveness? 17. Why are there only seven classes? How many ways
have you learned for classifying tetrads generally?
5. What does the symbol arg-1 mean? How would Which of these classifications can be applied to
you test for this genotype? both linear and unordered tetrads? Can you apply
6. How does the arg-1 symbol relate to the symbol +? these classifications to the tetrads in this problem?
7. What does the expression wild type mean? (Classify each class in as many ways as possible.)
Can you think of more possibilities in this cross? If
8. What does the word mutant mean? so, why are they not shown?
9. Does the biological function of the alleles shown 18. Do you think there are several different spore
have anything to do with the solution of this prob- orders within each class? Why would these different
lem? spore orders not change the class?
10. What does the expression linear octad analysis 19. Why is the following class not listed?
mean?
a⋅+ A ⋅ arg
11. In general, what more can be learned from linear
tetrad analysis that cannot be learned from unor- a⋅+ A ⋅ arg
dered tetrad analysis?
20. What does the expression linkage arrangement 49. Three different crosses in Neurospora are analyzed on
mean? the basis of unordered tetrads. Each cross combines a
21. What is a genetic interval? different pair of linked genes. The results are shown in
the following table:
22. Why does the problem state “centromere or cen-
tromeres” and not just “centromere”? What is the Non-
general method for mapping centromeres in tetrad Parental Tetra- parental
analysis? ditypes types ditypes
23. What is the total frequency of A ⋅ + ascospores? Cross Parents (%) (%) (%) (%)
(Did you calculate this frequency by using a for- 1 a ⋅ b+ × a+ ⋅ b 51 45 4
mula or by inspection? Is this a recombinant geno- + +
2 c ⋅ d ×c ⋅ d 64 34 2
type? If so, is it the only recombinant genotype?)
+ +
3 e ⋅ f ×e ⋅ f 45 50 5
24. The first two classes are the most common and
are approximately equal in frequency. What does
For each cross, calculate
this information tell you? What is their content of
parental and recombinant genotypes? a. the frequency of recombinants (RF).
Now try to solve the problem. If you are unable to do so, b. the uncorrected map distance, based on RF.
try to identify the obstacle and write a sentence or two c. the corrected map distance, based on tetrad
describing your difficulty. Then go back to the expansion frequencies.
questions and see if any of them relate to your difficulty. If
this approach does not work, inspect the Learning Objec- d. the corrected map distance, based on the mapping
tives and Key Concepts of this chapter and ask yourself function.
which might be relevant to your difficulty. (Note: Information from both Box 4-1 and Box 4-2 is
required for this problem.)
48. A geneticist studies 11 different pairs of Neurospora
loci by making crosses of the type a ⋅ b × a+ ⋅ b+ and 50. On Neurospora chromosome 4, the leu3 gene is just
then analyzing 100 linear asci from each cross. For to the left of the centromere and always segregates at
the convenience of making a table, the geneticist orga- the first division, whereas the cys2 gene is to the right
nizes the data as if all 11 pairs of genes had the same of the centromere and shows a second-division segre-
designation—a and b—as shown below. For each gation frequency of 16 percent. In a cross between a
cross, map the loci in relation to each other and to cen- leu3 strain and a cys2 strain, calculate the predicted
tromeres. (Note: Information from Box 4-2 is required frequencies of the seven classes of linear tetrads shown
for this problem.) on the next page, where l = leu3 and c = cys2. Ignore
double and other multiple crossovers. (Note: Informa-
tion from Box 4-2 is required for this problem.)
(i) l c (ii) l + (iii) l c (iv) l c (v) l c (vi) l + (vii) l + a. Determine the linkage relations of these two genes
to their centromere(s) and to each other. Specify
l c l + l + +c ++ +c +c distances in map units.
++ +c ++ ++ ++ +c ++ b. Draw a diagram to show the origin of the ascus
++ +c +c l + l c l + l c type with only one single representative (second
51. A rice breeder obtained a triple heterozygote carrying from right).
the three recessive alleles for albino flowers (al), brown (Note: Information from Box 4-2 is required for this
awns (b), and fuzzy leaves (fu), all paired with their problem.)
normal wild-type alleles. This triple heterozygote was 54. A corn geneticist wants to obtain a corn plant that has
testcrossed. The progeny phenotypes were the three dominant phenotypes: anthocyanin (A), long
tassels (L), and dwarf plant (D). In her collection of
170 wild type 710 albino
pure lines, the only lines that bear these alleles are AA
150 albino, brown, fuzzy 698 brown, fuzzy LL dd and aa ll DD. She also has the fully recessive
5 brown 42 fuzzy line aa ll dd. She decides to intercross the first two and
3 albino, fuzzy 38 albino, brown testcross the resulting hybrid to obtain in the progeny
a plant of the desired phenotype (which would have
a. Are any of the genes linked? If so, draw a map to be Aa Ll Dd in this case). She knows that the three
labeled with map distances. (Do not bother with a genes are linked in the order written, that the dis-
correction for multiple crossovers.) tance between the A /a and the L /l loci is 16 m.u., and
that the distance between the L /l and the D /d loci is
b. The triple heterozygote was originally made by
24 m.u.
crossing two pure lines. What were their genotypes?
a. Draw a diagram of the chromosomes of the par-
52. In a fungus, a proline mutant (pro) was crossed with a
ents, the hybrid, and the tester.
histidine mutant (his). A nonlinear tetrad analysis gave
the following results: b. Draw a diagram of the crossover(s) necessary to
produce the desired genotype.
+ + + + + his
c. What percentage of the testcross progeny will be of
+ + + his + his the phenotype that she needs?
pro his pro + pro + d. What assumptions did you make (if any)?
pro his pro his pro + 55. In the model plant Arabidopsis thaliana, the following
6 82 112 alleles were used in a cross:
56. In corn, the cross WW ee FF × ww EE ff is made. The a. What were the genotypes of the two parental mice
three loci are linked as follows: in the cross?
W/w E/e F/f b. Draw the chromosomes of the parents.
c. If you deduced linkage, state the map unit value or
8 m.u. 24 m.u.
values and show how they were obtained.
Assume no interference. 59. Consider the Neurospora cross +; +× f ; p
a. If the F1 is testcrossed, what proportion of progeny It is known that the + /f locus is very close to the cen-
will be ww ee ff? tromere on chromosome 7 — in fact, so close that there
b. If the F1 is selfed, what proportion of progeny will are never any second-division segregations. It is also
be ww ee ff? known that the + /p locus is on chromosome 5, at such
57. The fungal cross + ⋅ +×c ⋅ m was made, and nonlinear a distance that there is usually an average of 12 per-
(unordered) tetrads were collected. The results were cent second-division segregations. With this informa-
tion, what will be the proportion of octads that are
++ ++ +m a. parental ditypes showing M I patterns for both
++ +m +m loci?
cm c+ c+ b. nonparental ditypes showing M I patterns for both
cm cm c+ loci?
Total 112 82 6 c. tetratypes showing an M I pattern for + /f and an
M II pattern for + /p?
a. From these results, calculate a simple recombinant d. tetratypes showing an M II pattern for + /f and an
frequency. M I pattern for + /p?
b. Compare the Haldane mapping function and the (Note: Information from Box 4-2 is required for this
Perkins formula in their conversions of the RF problem.)
value into a “corrected” map distance.
60. In a haploid fungus, the genes al-2 and arg-6 are
c. In the derivation of the Perkins formula, only the 30 m.u. apart on chromosome 1, and the genes lys-
possibility of meioses with zero, one, and two 5 and met-1 are 20 m.u. apart on chromosome 6. In
crossovers was considered. Could this limit explain a cross
any discrepancy in your calculated values? Explain
briefly (no calculation needed). al -2 +; + met -1×+ arg-6; lys-5 +
( Note: Information from Box 4-1 is required for this
what proportion of progeny would be prototrophic
problem.)
+ + ; + +?
58. In mice, the following alleles were used in a cross:
61. The recessive alleles k (kidney-shaped eyes instead of
W = waltzing gait w = nonwaltzing gait wild-type round), c (cardinal-colored eyes instead of
wild-type red), and e (ebony body instead of wild-
G = normal gray color g = albino
type gray) identify three genes on chromosome 3 of
B = bent tail b = straight tail Drosophila. Females with kidney-shaped, cardinal-
A waltzing gray bent-tailed mouse is crossed with colored eyes were mated with ebony males. The F1 was
a nonwaltzing albino straight-tailed mouse and, wild type. When F1 females were testcrossed with kk
over several years, the following progeny totals are cc ee males, the following progeny phenotypes were
obtained: obtained:
a. Determine the order of the genes and the map dis- nail–patella syndrome, blood type A 66%
tances between them. normal nails and patellae, blood type O 16%
b. Draw the chromosomes of the parents and the F1. normal nails and patellae, blood type A 9%
c. Calculate interference and say what you think of nail–patella syndrome, blood type O 9%
its significance.
62. From parents of genotypes A /A ⋅ B /B and a /a ⋅ b /b , Fully analyze these data, explaining the relative fre-
a dihybrid was produced. In a testcross of the dihy- quencies of the four phenotypes. (See pages 156–157
brid, the following seven progenies were obtained: for the genetic basis of these blood types.)
66. Assume that three pairs of alleles are found in Dro-
A /a ⋅ B /b, a /a ⋅ b /b, A /a ⋅ B /b, A /a ⋅ b /b,
sophila: x+ and x, y+ and y, and z+ and z. As shown by
a /a ⋅ b /b, A /a ⋅ B /b, and a /a ⋅ B /b the symbols, each non-wild-type allele is recessive to its
Do these results provide convincing evidence of linkage? wild-type allele. A cross between females heterozygous
at these three loci and wild-type males yields prog-
CHALLENGING PROBLEMS eny having the following genotypes: 1010 x+⋅ y+ ⋅ z+
females, 430 x ⋅ y+ ⋅ z males, 441 x+ ⋅ y ⋅ z + males,
63. Use the Haldane map function to calculate the cor- 39 x ⋅ y ⋅ z males, 32 x+ ⋅ y+ ⋅ z males, 30 x+ ⋅ y+ ⋅ z+
rected map distance in cases where the measured m a l e s , 27 x ⋅ y ⋅ z + m a l e s , 1 x+ ⋅ y ⋅ z m a l e , a n d
RF = 5%, 10%, 20%, 30%, and 40%. Sketch a graph 0 x ⋅ y+ ⋅ z+ males.
of RF against corrected map distance, and use it to
answer the question, When should one use a map func- a. On what chromosome of Drosophila are the genes
tion? (Note: Information from Box 4-1 is required for carried?
this problem.) b. Draw the relevant chromosomes in the heterozy-
64. An individual heterozygous for four genes, A /a ⋅ B /b ⋅ gous female parent, showing the arrangement of
C /c ⋅ D /d , is testcrossed with a /a ⋅ b /b ⋅ c /c ⋅ d /d , and the alleles.
1000 progeny are classified by the gametic contribu- c. Calculate the map distances between the genes and
tion of the heterozygous parent as follows: www the coefficient of coincidence.
a ⋅ B⋅C ⋅ D 42 www 67. The five sets of data given in the following table rep-
A⋅b⋅c⋅d 43 resent the results of testcrosses using parents with the
same alleles but in different combinations. Determine
A ⋅ B⋅C ⋅ d 140 the order of genes by inspection—that is, without cal-
a⋅b⋅c⋅D 145 culating recombination values. Recessive phenotypes
a⋅B⋅c⋅D 6 are symbolized by lowercase letters and dominant phe-
notypes by pluses.
A ⋅ b ⋅C ⋅ d 9
A⋅B⋅c⋅d 305 Phenotypes Data sets
observed in
a ⋅ b ⋅C ⋅ D 310 3-point testcross 1 2 3 4 5
a. Which genes are linked? +++ 317 1 30 40 305
b. If two pure-breeding lines had been crossed ++ c 58 4 6 232 0
to produce the heterozygous individual, what +b+ 10 31 339 84 28
would their genotypes have been?
+b c 2 77 137 201 107
c. Draw a linkage map of the linked genes, show-
a ++ 0 77 142 194 124
ing the order and the distances in map units.
a + c 21 31 291 77 30
d. Calculate an interference value, if appropriate.
a b + 72 4 3 235 1
65. An autosomal allele N in humans causes abnormalities
in nails and patellae (kneecaps) called the nail–patella a b c 203 1 34 46 265
syndrome. Consider matings in which one partner has 68. From the phenotype data given in the following table
the nail–patella syndrome and blood type A and the for two 3-point testcrosses for (1) a, b, and c and (2)
other partner has normal nails and patellae and blood b, c, and d, determine the sequence of the four genes a,
type O. These matings produce some children who b, c, and d and the three map distances between them.
have both the nail–patella syndrome and blood type A. Recessive phenotypes are symbolized by lowercase let-
Assume that unrelated children from this phenotypic ters and dominant phenotypes by pluses.
group mature, mate, and have children. Four pheno-
types are observed in the following percentages in this
second generation:
35
+b+ 2215 b c + 141
The geneticist crosses the F1 with a recessive tester and If the following cross is made
classifies the progeny by the gametic contribution of
+ + + /+ + +× w s e /w s e
the F1:
A ⋅ B⋅C ⋅ D⋅ E 316 and the F1 is testcrossed with w s e /w s e , and it is
assumed that there is no interference on this region of
a ⋅ b ⋅C ⋅ d ⋅ E 314 the chromosome, what proportion of progeny will be
A ⋅ B⋅C ⋅ d ⋅ E 31 of the following genotypes?
a ⋅ b ⋅C ⋅ D⋅ E 39 a. + + + e. + + e
A ⋅ b ⋅C ⋅ d ⋅ E 130 b. w s e f. w s +
a ⋅ B⋅C ⋅ D⋅ E 140 c. + s e g. w + e
A ⋅ b ⋅C ⋅ D⋅ E 17 d. w + + h. + s +
a ⋅ B⋅C ⋅ d ⋅ E 13 76. Every Friday night, genetics student Jean Allele,
exhausted by her studies, goes to the student union’s
1000
bowling lane to relax. But, even there, she is haunted
The second cross of pure lines is A /A ⋅ B /B ⋅ C /C ⋅ D /D ⋅ by her genetic studies. The rather modest bowling lane
E /E × a /a ⋅ B /B ⋅ c /c ⋅ D /D ⋅ e /e. has only four bowling balls: two red and two blue.
The geneticist crosses the F1 from this cross with a They are bowled at the pins and are then collected and
recessive tester and obtains returned down the chute in random order, coming to
A ⋅ B⋅C ⋅ D⋅ E 243 rest at the end stop. As the evening passes, Jean notices
familiar patterns of the four balls as they come to rest
a⋅B⋅c⋅D⋅e 237
at the stop. Compulsively, she counts the different pat-
A⋅B⋅c⋅D⋅e 62 terns. What patterns did she see, what were their fre-
a ⋅ B⋅C ⋅ D⋅ E 58 quencies, and what is the relevance of this matter to
genetics?
A ⋅ B⋅C ⋅ D⋅ e 155
a⋅B⋅c⋅D⋅E 165 77. In a tetrad analysis, the linkage arrangement of the p
and q loci is as follows:
a ⋅ B⋅C ⋅ D⋅ e 46
(i) (ii)
A⋅B⋅c⋅D⋅E 34
1000 p q
The geneticist also knows that genes D and E assort
Assume that
independently.
• in region i, there is no crossover in 88 percent of
a. Draw a map of these genes, showing distances in
meioses, and there is a single crossover in 12 per-
map units wherever possible.
cent of meioses;
b. Is there any evidence of interference?
• in region ii, there is no crossover in 80 percent of
74. In the plant Arabidopsis, the loci for pod length (L, meioses, and there is a single crossover in 20 per-
long; l, short) and fruit hairs (H, hairy; h, smooth) are cent of meioses; and
linked 16 m.u. apart on the same chromosome. The
• there is no interference (in other words, the situa-
following crosses were made:
tion in one region does not affect what is going on
(i) L H /L H × l h /l h → F1 in the other region).
(ii) L h /L h × l H /l H → F1 What proportions of tetrads will be of the following
types? (a) M I M I , PD; (b) M I M I , NPD; (c) M I M II ,
If the F1’s from cross i and cross ii are crossed, T; (d) M II M I , T; (e) M II M II , PD; (f) M II M II , NPD;
a. what proportion of the progeny are expected to be (g) M II M II , T. (Note: Here the M pattern written first
l h /l h ? is the one that pertains to the p locus.) Hint: The eas-
b. what proportion of the progeny are expected to be iest way to do this problem is to start by calculating
L h /l h? the frequencies of asci with crossovers in both regions,
region i, region ii, and neither region. Then determine
75. In corn (Zea mays), the genetic map of part of chro- what M I and M II patterns result. (Note: Information
mosome 4 is as follows, where w, s, and e represent from Box 4-2 is required for this problem.)
recessive mutant alleles affecting the color and shape
of the pollen: 78. For an experiment with haploid yeast, you have two
different cultures. Each will grow on minimal medium
w s e
to which arginine has been added, but neither will
grow on minimal medium alone. (Minimal medium is
8 m.u. 14 m.u.
inorganic salts plus sugar.) Using appropriate methods,
you induce the two cultures to mate. The diploid cells a long tandem repeat (l) and the latter for a short tan-
then divide meiotically and form unordered tetrads. dem repeat (s). The two were crossed to form an F1,
Some of the ascospores will grow on minimal medium. which was then backcrossed to the second pure line. A
You classify a large number of these tetrads for the thousand progeny were scored as follows:
phenotypes ARG− (arginine requiring) and ARG+
(arginine independent) and record the following data: Aa Bb ss 9 Aa bb ss 43
Aa Bb ls 362 Aa bb ls 93
Segregation of aa bb ls 11 aa Bb ls 37
ARG− : ARG+ Frequency (%)
aa bb ss 358 aa Bb ss 87
4:0 40
3:1 20 a. What do these results tell us about linkage?
2:2 40 b. Draw a map if appropriate.
c. Incorporate the variable tandem repeat locus into
a. Using symbols of your own choosing, assign geno-
your map.
types to the two parental cultures. For each of the
three kinds of segregation, assign genotypes to the
GENETICS AND SOCIETY
segregants.
b. If there is more than one locus governing arginine Mapping the human chromosomes by the analysis of
requirement, are these loci linked? recombinant frequencies between phenotypic markers was
of only limited success. What might be some possible rea-
79. A molecular analysis of two pure lines A /A ⋅ B /B and
sons for this?
a /a ⋅ b /b showed that the former was homozygous for
Gene Interaction
153
154
mRNA
Chromosome
+ + m
+ m m
Chromosome
mRNA
Protein
FIGURE 5-1 In the heterozygote, even though the mutated copy of the gene produces
nonfunctional protein, the wild-type copy generates enough functional protein to produce
the wild-type phenotype.
A /a . As we saw in Chapter 2, the alleles that result in PAH to break down the phenylalanine entering the body.
henylketonuria (PKU) and many other single-gene human
p One “dose” of functional PAH protein, produced by one
diseases are fully recessive, whereas their wild-type alleles P allele, results in the wild-type phenotype. Thus, the PAH
are dominant. Other single-gene diseases such as pseudo- wild-type allele is said to be haplosufficient. Haplo means
achondroplasia result from alleles that are fully dominant, a haploid dose (one), and sufficient refers to the ability of
whereas, in those cases, the wild-type allele is recessive. that single dose to produce the wild-type phenotype. Hence,
How can these dominance relations be interpreted at the both P /P (two doses) and P /p (one dose) have enough PAH
cellular level? activity to result in the normal cellular chemistry. People
The disease PKU is a good general model for recessive with p /p have zero doses of PAH activity. Figure 5-1 illus-
mutations. Recall from Chapter 2 that PKU is caused by a trates this general notion.
defective allele of the gene encoding the enzyme phenylal- How can we explain fully dominant mutations? There
anine hydroxylase (PAH). In the absence of normal PAH, are several molecular mechanisms for dominance. A regu-
the phenylalanine entering the body in food is not broken larly encountered mechanism is that the wild-type allele of
down and hence accumulates. Under such conditions, phe- a gene is haploinsufficient. In haploinsufficiency, one wild-
nylalanine is converted into phenylpyruvic acid, which is type dose is not enough to achieve normal levels of func-
transported to the brain through the bloodstream and there tion. Assume that 16 units of a gene’s product are needed
impedes normal development, leading to intellectual dis- for normal chemistry and that each wild-type allele can
abilities. The reason that the defective allele is recessive is make 10 units. Two wild-type alleles will produce 20 units
that just one copy of the wild-type allele P produces enough of product, well over the minimum. But consider what
alleles are i, I A , and I B , but a person can have only two of the
three alleles or two copies of one of them. The pairwise com- FIGURE 5-4 The sickle-shaped cell is caused by a single mutation in
the gene for hemoglobin. [Eye of Science/Science Source.]
binations result in six different genotypes: the three homozy-
gotes and three different types of heterozygotes, as follows: Figure 5-4 shows an electron micrograph of blood cells
Genotype Blood type including some sickled cells. In regard to the presence or
A A A absence of anemia, the Hb A allele is dominant. In the het-
I / I , I /i A
erozygote, a single Hb A allele produces enough functioning
B B B
I /I , I / i B hemoglobin to prevent anemia. In regard to blood-cell shape,
A
I /I B
AB however, there is incomplete dominance, as shown by the fact
i/ i O
that, in the heterozygote, many of the cells have a slight sickle
shape. Finally, in regard to hemoglobin itself, there is codom-
In this allelic series, the alleles determine the presence inance. The alleles Hb A and HbS encode two different forms
and form of a complex sugar molecule present on the sur- of hemoglobin that differ by a single amino acid, and both
face of red blood cells. This sugar molecule is an antigen, a forms are synthesized in the heterozygote. The A and S forms
cell-surface molecule that can be recognized by the immune of hemoglobin can be separated by electrophoresis because it
system. The alleles I A and I B determine two different forms happens that they have different charges (Figure 5-5). We see
of this cell-surface molecule. However, the allele i results in that homozygous Hb A /Hb A people have one type of hemo-
no cell-surface molecule of this type (it is a null allele). In globin (A), and anemics have another (type S), which moves
the genotypes I A /i and I B /i , the alleles I A and I B are fully more slowly in the electric field. The heterozygotes have both
dominant over i. However, in the genotype I A /I B, each of types, A and S. In other words, there is codominance at the
the alleles produces its own form of the cell-surface mole- molecular level. The fascinating population genetics of the
cule, and so the A and B alleles are codominant. Hb A and HbS alleles will be considered in Chapter 20.
The human disease sickle-cell anemia illustrates the Sickle-cell anemia illustrates the arbitrariness of the
somewhat arbitrary ways in which we classify dominance. terms dominance, incomplete dominance, and codomi-
The gene concerned encodes the molecule hemoglobin, nance. The type of dominance inferred depends on the
which is responsible for transporting oxygen in blood ves- phenotypic level at which the assay is made—organismal,
sels and is the major constituent of red blood cells. There cellular, or molecular. Indeed, caution should be applied to
are two main alleles, Hb A and HbS, and the three possible many of the categories that scientists use to classify struc-
genotypes have different phenotypes, as follows: tures and processes; these categories are devised by humans
for the convenience of analysis.
Hb A /Hb A : normal; red blood cells never sickle
HbS /HbS : severe, often fatal anemia; abnormal KEY CONCEPT In general, three main types of dominance
hemoglobin causes red blood cells to can be distinguished: full dominance, incomplete dominance,
have sickle shape and codominance. The type of dominance is determined by
the molecular functions of the alleles of a gene and by the
Hb A /HbS : no anemia; red blood cells sickle only
investigative level of analysis.
under low oxygen concentrations
Heterozygotes can express the protein product of both alleles Seven alleles and their interactions
in leaf patterning of clover
Sickle-cell Sickle-cell
Phenotype Unaffected anemia trait
Genotype HbA / HbA HbS / HbS HbS / HbA
Origin vv
V baV ba V lV ba V hV ba V fV ba
The leaves of clover plants show several variations on
the dominance theme. Clover is the common name for
plants of the genus Trifolium. There are many species.
Some are native to North America, whereas others grow
there as introduced weeds. Much genetic research has been V bV b V lV b V hV b V fV b V ba V b
done with white clover, which shows considerable varia-
tion among individual plants in the curious V, or chevron,
pattern on the leaves. The different chevron forms (and the
absence of chevrons) are determined by a series of seven V byV by V lV by V hV by V fV by V b a V by V b V by
alleles, as seen in Figure 5-6, which shows the many different
types of interactions possible for even one allele. In most FIGURE 5-6 Multiple alleles determine the chevron pattern on the
leaves of white clover. The genotype of each plant is shown below
practical cases, many alleles of a gene can be found together
it. There is a variety of dominance interactions. [Research by W. Ellis
in a population, constituting an allelic series. The pheno- Davies.]
types shown by the allelic combinations are many and var-
ied, reflecting the relative nature of dominance: an allele can
show dominance with one partner but not with another. function) is essential to the organism’s operation. Essential
Hence, the complexity illustrated by the ABO blood type genes are those without which an organism dies. (An exam-
system is small compared with that in a case such as clover ple of an essential gene might be a ribosomal gene without
chevrons. which no protein would be made.) Indeed, with the use of
modern DNA technology, a null mutant allele of a gene of
KEY CONCEPT A gene can take on many forms, called interest can now be made intentionally and made homo-
alleles, each caused by various mutations of the DNA zygous to see if it is lethal and under which environmental
sequence. Some mutant alleles have phenotypic impact; conditions. Lethal alleles are also useful in determining the
others do not. developmental stage at which the gene normally acts. In
this case, geneticists look for whether death from a lethal
mutant allele occurs early or late in the development of a
zygote. The phenotype associated with death can also be
Recessive lethal alleles informative in regard to gene function; for example, if a
An allele that is capable of causing the death of an organ- certain organ appears to be abnormal, the gene is likely to
ism is called a lethal allele. In the characterization of a set be expressed in that organ.
of newly discovered mutant alleles, a recessive mutation What is the diagnostic test for lethality? The test is well
(a mutation in the homozygous state) is sometimes found illustrated by one of the prototypic examples of a lethal
to be lethal. This information is potentially useful in that allele, a coat-color allele in mice (see the Model Organism
it shows that the newly discovered gene (of yet unknown box on page 159). Normal wild-type mice have coats with a
rather dark overall pigmentation. A mutation called yellow The expected monohybrid ratio of 1: 2:1 would be
(a lighter coat color) shows a curious inheritance pattern. If found among the zygotes, but it is altered to a 2:1 ratio in
any yellow mouse is mated with a homozygous wild-type the progeny actually seen at birth because zygotes with a
mouse, a 1:1 ratio of yellow to wild-type mice is always lethal AY /AY genotype do not survive to be counted. This
observed in the progeny. This result suggests that a yellow hypothesis is supported by the removal of uteri from preg-
mouse is always heterozygous for the yellow allele and that nant females of the yellow × yellow cross; one-fourth of the
the yellow allele is dominant over wild type. However, if embryos are found to be dead.
any two yellow mice are crossed with each other, the result
is always as follows: A recessive lethal allele, yellow coat
yellow × yellow → 2 yellow, 1 wild type
3 3
159
Tailless, a recessive lethal allele in cats Geneticists commonly encounter situations in which
expected phenotypic ratios are consistently skewed in one
direction because a mutant allele reduces viability. For
example, in the cross A /a × a /a , we predict a progeny ratio
of 50 percent A /a and 50 percent a/a, but we might con-
sistently observe a ratio such as 55 percent:45 percent or
60 percent:40 percent. In such a case, the recessive allele
is said to be sublethal because the lethality is expressed in
only some but not all of the homozygous individuals. Thus,
lethality may range from 0 to 100 percent, depending on
the gene itself, the rest of the genome, and the environment.
We have seen that lethal alleles are useful in diagnosing
the time at which a gene acts and the nature of the phenotypic
defect that kills. However, maintaining stocks bearing lethal
alleles for laboratory use is a challenge. In diploids, recessive
lethal alleles can be maintained as heterozygotes. In haploids,
heat-sensitive lethal alleles are useful. They are members of a
general class of temperature-sensitive (ts) mutations. Their
phenotype is wild type at the permissive temperature (often
FIGURE 5-8 A Manx cat. A dominant allele causing taillessness is room temperature) but mutant at some higher restrictive
lethal in the homozygous state. The phenotype of two eye colors is
unrelated to taillessness. [Gerard Lacz/NHPA/Photoshot.]
temperature. Temperature-sensitive alleles are thought to be
caused by mutations that make the protein prone to twist or
bend its shape to an inactive conformation at the restrictive
The AY allele produces effects on two characters: coat temperature. Research stocks can be maintained easily under
color and survival. In general, the term pleiotropic is used permissive conditions, and the mutant phenotype can be
for any allele that affects several properties of an organism. assayed in a subset of individuals by a switch to the restrictive
The tailless Manx phenotype in cats (Figure 5-8) also is conditions. In diploids, temperature-sensitive dominant lethal
produced by an allele that is lethal in the homozygous state. mutations also are useful. This type of mutation is lethal even
A single copy of the Manx allele, M L , severely interferes when present in a single dose, but only when the experimenter
with normal spinal development, resulting in the absence switches the organism to the restrictive temperature.
of a tail in the M L /M heterozygote. But in the M L /M L Null alleles for genes identified through genomic
homozygote, two copies of the Manx allele produces such sequencing can be made by using a variety of “reverse
an extreme abnormality in spinal development that the genetic” procedures that specifically knock out the function
embryo does not survive. of that gene. These will be described in Chapter 14.
KEY CONCEPT Some mutant alleles are lethal; that is, they KEY CONCEPT To see if a gene is essential, a null allele is
can result in the death of the organism. Lethality is most often tested for lethality.
recessive.
The yellow and M L alleles have their own phenotypes Penetrance and expressivity
in a heterozygote, but most recessive lethals are silent in the
In the analysis of single-gene inheritance, there is a natural
heterozygote. In such a situation, recessive lethality is diag-
tendency to choose mutants that produce clear Mendelian
nosed by observing the death of 25 percent of the progeny
ratios. In such cases, we can use the phenotype to distin-
at some stage of development.
guish mutant and wild-type genotypes with almost 100 per-
Whether an allele is lethal or not often depends on the
cent certainty. In these cases, we say that the mutation is
environment in which the organism develops. Whereas
100 percent penetrant into the phenotype. However, many
certain alleles are lethal in virtually any environment,
mutations show incomplete penetrance; that is, not every
others are viable in one environment but lethal in another.
individual with the genotype expresses the corresponding
Human hereditary diseases provide some examples. Cystic
phenotype. Thus, penetrance is defined as the percentage of
fibrosis and sickle-cell anemia are diseases that would be
individuals with a given allele who exhibit the phenotype
lethal without treatment. Furthermore, many of the alleles
associated with that allele.
favored and selected by animal and plant breeders would
Why would an organism have a particular genotype
almost certainly be eliminated in nature as a result of
and yet not express the corresponding phenotype? There
competition with the members of the natural population.
are several possible reasons:
The dwarf mutant varieties of grain, which are very high
yielding, provide good examples; only careful nurturing by 1. The influence of the environment. Individuals with
farmers has maintained such alleles for our benefit. the same genotype may show a range of phenotypes,
depending on the environment. The range of phenotypes Penetrance and expressivity contrasted
for mutant and wild-type individuals may overlap: the
phenotype of a mutant individual raised in one set of Phenotypic expression
(each oval represents an individual)
circumstances may match the phenotype of a wild-type
individual raised in a different set of circumstances.
Should this matching happen, the mutant cannot be dis-
tinguished from the wild type. Variable penetrance
2. The influence of other interacting genes. Uncharacter-
ized modifiers, epistatic genes, or suppressors in the rest
of the genome (all discussed shortly) may act to prevent
Variable expressivity
the expression of the typical phenotype.
3. The subtlety of the mutant phenotype. The subtle effects
brought about by the absence of a gene function may be
difficult to measure in a laboratory situation. Variable penetrance and expressivity
A typical encounter with incomplete penetrance is FIGURE 5-10 Assume that all the individuals shown have the same
shown in Figure 5-9. In this human pedigree, we see a nor- pigment allele (P) and possess the same potential to produce pigment.
mally dominantly Effects from the rest of the genome and the environment may suppress
or modify pigment production in any one individual. The color indicates
inherited phenotype
Inferring incomplete penetrance the level of expression.
disappearing in the sec-
ond generation only to (genotype b /b) from different stocks might show very dif-
1 2 reappear in
3 the next. ferent4intensities of brown pigment
5 from light to dark. As
Another measure for penetrance, variable expressivity may be due to varia-
Q for describing the tion in the allelic constitution of the rest of the genome or to
range of phenotypic environmental factors. Figure 5-10 illustrates the distinction
expression is called between penetrance and expressivity. An example of vari-
expressivity. Expressiv- able expressivity in dogs is found in Figure 5-11.
R
ity measures the degree The phenomena of incomplete penetrance and variable
FIGURE 5-9 In this human pedigree to which a given allele expressivity can make any kind of genetic analysis substan-
6 of a dominant allele that is7not fully is expressed
8 at the phe- tially9 more difficult, including
10human pedigree analysis and
penetrant, person Q does not display notypic level; that is, predictions in genetic counseling. For example, it is often
the phenotype but passed the dominant expressivity measures the case that a disease-causing allele is not fully penetrant.
allele to at least two progeny. Because
the allele is not fully penetrant, the other
the intensity of the Thus, someone could have the allele but not show any signs
progeny (for example, R) may or may not phenotype. For exam- of the disease. If that is the case, it is difficult to give a clean
have inherited the dominant allele. ple, “brown” animals genetic bill of health to any person in a disease pedigree
1 2 3 4
5 6 7 8
9 10
(for example, person R in Figure 5-9). On the other hand, maleylacetoacetic acid; so he proposed that, in AKU, there
pedigree analysis can sometimes identify persons who do is a defect in this conversion. Consequently, homogentisic
not express but almost certainly do have a disease geno- acid builds up and is excreted. Garrod’s observations raised
type (for example, individual Q in Figure 5-9). Similarly, the possibility that the cell’s chemical pathways were under
variable expressivity can complicate counseling because the control of a large set of interacting genes. However, the
persons with low expressivity might be misdiagnosed. direct demonstration of this control was provided by the
Even though penetrance and expressivity can be quanti- later work of Beadle and Tatum on the fungus Neurospora.
fied, they nevertheless represent “fuzzy” situations because
rarely is it possible to identify the specific factors causing Biosynthetic pathways in Neurospora
variation without substantial extra research.
The landmark study by George Beadle and Edward
Tatum in the 1940s not only clarified the role of genes but
KEY CONCEPT The terms penetrance and expressivity
uantify the modification of a gene’s effect by varying environ-
q also demonstrated the interaction of genes in biochemi-
ment and genetic background; they measure, respectively, the cal pathways. They later received a Nobel Prize for their
percentage of cases in which the phenotype is observed and study, which marks the beginning of all molecular biology.
its severity. Beadle and Tatum did their work on the haploid fungus
Neurospora, which we have met in earlier chapters. Their
plan was to investigate the genetic control of cellular chem-
We now turn to the approaches that can be used to istry. In what has become the standard forward genetic
detect the interaction between two or more loci. approach, they first irradiated Neurospora cells to produce
mutations and then tested cultures grown from ascospores
for interesting mutant phenotypes relevant to biochemical
5.2 INTERACTION OF GENES function. They found numerous mutants that had defec-
IN PATHWAYS tive nutrition. Specifically, these mutants were auxotrophic
mutants, meaning that the mutants would not grow unless
their medium contained one or more specific cellular build-
LO 5.4 Describe the hypotheses proposed to explain various
types of gene interaction at the molecular level. ing blocks. Whereas wild-type Neurospora can use its cellular
biochemistry to synthesize virtually all its cellular compo-
Genes act by controlling cellular chemistry. Early in the nents from the inorganic nutrients and a carbon source in the
twentieth century, Archibald Garrod, an English physician medium, auxotrophic mutants cannot. In order to grow, such
( Figure 5-12), made the first observation supporting this mutants require a nutrient to be supplied (a nutrient that a
insight. Garrod noted that several recessive human diseases wild-type fungus is able to synthesize for itself), suggesting
show defects in what is called metabolism, the general set that the mutant is defective for some normal synthetic step.
of chemical reactions taking place in an organism. This As their first step, Beadle and Tatum confirmed that each
observation led to the notion that such genetic diseases are mutation that generated a nutrient requirement was inherited
“inborn errors of metabolism.” Garrod worked on a dis- as a single-gene mutation because each gave a 1:1 ratio when
ease called alkaptonuria (AKU), or black urine disease. He crossed with a wild type (remember, Neurospora is a haploid
discovered that the substance responsible for black urine organism). Letting aux represent an auxotrophic mutation,
was homogentisic acid, which is present in high amounts + × aux
and secreted into the urine in AKU patients. He knew that,
↓
in unaffected people, homogentisic acid is converted into
progeny: 1 + and 1 aux
2 2
Discoverer of inborn errors Their second step was to classify the specific nutri-
of metabolism tional requirement of each auxotroph. Some would grow
only if proline was supplied, others methionine, others
pyridoxine, others arginine, and so on. Beadle and Tatum
decided to focus on arginine auxotrophs. They found that
the genes that mutated to give arginine auxotrophs mapped
to three different loci on three separate chromosomes. (To
determine whether their collection of arginine auxotrophs
resulted from mutation in the same gene or multiple differ-
ent genes, Beadle and Tatum methodically analyzed pairs
of mutants using the complementation test, discussed in
Section 5.3.) Let’s call the genes at the three loci the arg-1,
arg-2, and arg-3 genes. A key breakthrough was Beadle and
Tatum’s discovery that the auxotrophs for each of the three
FIGURE 5-12 British physician Archibald loci differed in their response to the structurally related
Garrod (1857–1936). [SPL/Science Source.] compounds ornithine and citrulline (Figure 5-13). The arg-1
Arginine and its chemical relatives This brilliant model, which was initially known as
the one-gene–one-enzyme hypothesis, was the source of
NH2 NH2 the first exciting insight into the functions of genes: genes
C"O C " NH somehow were responsible for the function of enzymes,
NH2 NH NH and each gene apparently controlled one specific enzyme
in a series of interconnected steps in a biochemical path-
(CH2)3 (CH2)3 (CH2)3
way. Other researchers obtained similar results for other
CHNH2 CHNH2 CHNH2 biosynthetic pathways, and the hypothesis soon achieved
COOH COOH COOH general acceptance. All proteins, whether or not they are
Ornithine Citrulline Arginine enzymes, also were found to be encoded by genes, and
so the phrase was refined to become the one-gene–one-
FIGURE 5-13 The chemical structures polypeptide hypothesis. (Recall that a polypeptide is the
of arginine and the structurally related
simplest type of protein, a single chain of amino acids.)
compounds citrulline and ornithine.
It soon became clear that a gene encodes the physical
mutants grew when supplied with any one of the chemi- structure of a protein, which in turn dictates its function.
cals ornithine, citrulline, or arginine. The arg-2 mutants Beadle and Tatum’s hypothesis became one of the great
grew when given arginine or citrulline but not ornithine. unifying concepts in biology because it provided a bridge
The arg-3 mutants grew only when arginine was supplied. that brought together the two major research areas of
These results are summarized in Table 5-1. genetics and biochemistry.
Cellular enzymes were already known to interconvert We must add parenthetically that, although the great
such related compounds. On the basis of the properties of the majority of genes encode proteins, some are known to
arg mutants, Beadle and Tatum and their colleagues proposed encode RNAs that have special functions. All genes are
a biochemical pathway for such conversions in Neurospora: transcribed to make RNA. Protein-encoding genes are tran-
scribed to messenger RNA (mRNA), which is then trans-
enzyme X enzyme Y
precursor → ornithine → lated into protein. However, the RNA encoded by a
enzyme Z
citrulline → arginine minority of genes is never translated into protein because
the RNA itself has a unique function. These are called
This pathway nicely explains the three classes of mutants f unctional RNAs. Some examples are transfer RNAs, ribo-
shown in Table 5-1. Under the model, the arg-1 mutants somal RNAs, and small cytoplasmic RNAs; more about
have a defective enzyme X, and so they are unable to con- them will be covered in later chapters.
vert the precursor into ornithine as the first step in produc-
ing arginine. However, they have normal enzymes Y and Z, KEY CONCEPT Chemical synthesis in cells is by pathways of
and so the arg-1 mutants are able to produce arginine if sup- sequential steps catalyzed by enzymes. The genes encoding
plied with either ornithine or citrulline. Similarly, the arg-2 the enzymes of a specific pathway constitute a functionally
mutants lack enzyme Y, and the arg-3 mutants lack enzyme interacting subset of the genome.
Z. Thus, a mutation at a particular gene is assumed to inter-
fere with the production of a single enzyme. The defective
enzyme creates a block in some biosynthetic pathway. The Gene interaction in other types
block can be circumvented by supplying to the cells any com- of pathways
pound that normally comes after the block in the pathway. The notion that genes interact through pathways is a
We can now diagram a more complete biochemical model: powerful one that finds application in all organisms. The
arg-1+ arg-2+ Neurospora arginine pathway is an example of a synthetic
↓ ↓ pathway, a chain of enzymatic conversions that synthe-
precursor enzyme X
→ ornithine
→ enzyme Y sizes essential molecules. We can extend the idea again to
a human case already introduced, the disease phenylke-
arg-3+
tonuria (PKU), which is caused by an autosomal recessive
↓
allele. This disease results from an inability to convert phe-
enzyme Z
citrulline → arginine nylalanine into tyrosine. As a result of the block, phenylal-
anine accumulates and is spontaneously converted into a
TABLE 5-1 Growth of arg Mutants in Response to toxic compound, phenylpyruvic acid. The PKU gene is part
Supplements of a metabolic pathway like the Neurospora arginine path-
Supplements way, a section of which is shown in Figure 5-14. The illustra-
Mutant Ornithine Citrulline Arginine tion includes several other diseases caused by blockages in
steps in this pathway (including alkaptonuria, the disease
arg-1 + + +
investigated by Garrod).
arg-2 − + +
Another type of pathway is a signal-transduction
arg-3 − − + pathway. This type of pathway is a chain of complex sig-
Note: A plus sign means growth; a minus sign means no growth. nals, from the environment to the internal components
A synthetic pathway and associated diseases analysis of mutants with aberrant mating response, and
the steps were pieced together by using the approaches in
Dietary protein the next section. A mutation at any one of these steps may
disrupt the mating process.
Developmental pathways comprise the steps by which
a zygote becomes an adult organism. This process involves
Phenylalanine If [Phe] high
(Phe)
many genetically controlled steps, including establishment of
the anterior-posterior and dorsal-ventral axes, laying down
Phenylpyruvic
acid the basic body plan of organs, and tissue differentiation and
Phe hydroxylase PKU
movement. These steps can require gene regulation and sig-
nal transduction. Developmental pathways will be taken up
Albinism Cretinism in detail in Chapter 13, but the interaction of genes in these
Tyrosine pathways is analyzed in the same way, as we will see next.
(Tyr)
Melanin Thyroxine
KEY CONCEPT Gene interaction occurs in any cellular
athway, such as biosynthetic, signal transduction, and
p
Tyr aminotransferase developmental.
Hydroxyphenylpyruvic acid
(HPA)
5.3 INFERRING GENE
HPA oxidase Tyrosinosis INTERACTIONS
LO 5.2 Recognize phenotypic ratios diagnostic of the
Homogentisic acid presence of a lethal allele.
(HA)
HA oxidase Alkaptonuria LO 5.5 Determine whether two mutations are in the same
gene or in different genes, using progeny ratios or
using complementation tests.
Maleylacetoacetic
acid
LO 5.6 Infer how two genes may be interacting, based on
modified Mendelian ratios.
CO2 + H2O
LO 5.7 For known cases of gene interaction, predict
progeny ratios in crosses.
FIGURE 5-14 A section of the phenylalanine metabolic pathway in
humans, including diseases associated with enzyme blockages. The
The genetic approach that reveals the interacting genes for
disease PKU is produced when the enzyme phenylalanine hydroxylase
malfunctions. Accumulation of phenylalanine results in an increase a particular biological property is briefly as follows:
in phenylpyruvic acid, which interferes with the development of the
Step 1. Obtain many single-gene mutants and test for
nervous system.
dominance.
Step 2. Test the mutants for allelism—are they at one or
of the cell, that result in activation of cellular responses. several loci?
These pathways are crucial to the proper function of an
Step 3. Combine the mutants in pairs to form double
organism. One of the best-understood signal-transduction
mutants to see if the genes interact.
pathways was worked out from a genetic analysis of the
mating response in baker’s yeast. Two mating types, deter- Gene interaction is inferred from the phenotype of the dou-
mined by the alleles MATa and MATα, are necessary for ble mutant: if the genes interact, then the phenotype differs
yeast mating to occur. When a cell is in the presence of from the simple combination of both single-gene mutant
another cell of opposite mating type, it undergoes a series phenotypes. If mutant alleles from different genes interact,
of changes in shape and behavior to prepare for mating. then we infer that the wild-type genes interact normally as
This mating response is triggered when a mating pher- well. In cases in which the two mutants interact, a modified
omone (hormone) is released by a cell of the opposite 9:3:3:1 Mendelian ratio will often result.
mating type and binds to a membrane receptor on the A procedure that must be carried out before testing
receiving cell. This signal promotes the sequential action of interactions is to determine whether each mutation is of
a set of genes, which ultimately activates the transcription a different locus (step 2 above). The mutant screen could
of mating-specific genes that enable the cell to mate. This have unintentionally favored certain genes. Thus, the set of
set of genes was discovered through a standard interaction gene loci needs to be defined, as shown in the next section.
KEY CONCEPT The complementation test is a standard way Let’s illustrate the complementation test with an example
of determining whether or not two recessive mutations are in from harebell plants (genus Campanula). The wild-type flower
the same gene. The mutations are united in one cell, and if the color of this plant is blue. Let’s assume that, from a mutant
cell shows the wild-type phenotype, the mutations have com- hunt, we have obtained three white-petaled mutants and that
plemented and must be in different genes. they are available as homozygous pure-breeding strains. They
all look the same, and so we do not know a priori whether
they are genetically identical. We will call the mutant strains
Sorting mutants using the $, £, and ¥ to avoid any symbolism using letters, which might
complementation test imply dominance. When crossed with wild type, each mutant
gives the same results in the F1 and F2 as follows:
How is it possible to decide whether two mutations belong to
3 1
the same gene? There are several ways. First, each mutant allele white $ × blue → F1 , all blue → F2 , 4
blue, 4
white
could be mapped. Then, if two mutations map to two different 3 1
white £ × blue → F1 , all blue → F2 , 4
blue, 4
white
chromosomal loci, they are likely of different genes. However,
3 1
this approach is time consuming on a large set of mutations. A white ¥ × blue → F1 , all blue → F2 , 4
blue, 4
white
quicker approach often used is the complementation test.
In each case, the results show that the mutant condition
In a diploid, the complementation test is performed by
is determined by the recessive allele of a single gene. How-
intercrossing two individuals that are homozygous for differ-
ever, are they three alleles of one gene, of two genes, or of
ent recessive mutations. The next step is to observe whether
three genes? Because the mutants are recessive, the question
the progeny have the wild-type phenotype. If the progeny
can be answered by the complementation test, which asks if
are wild type, the two recessive mutations must be in dif-
the mutants complement one another.
ferent genes because the respective wild-type alleles pro-
Let us intercross the mutants to test for complemen-
vide wild-type function. In this case, the two mutations are
tation. Assume that the results of intercrossing mutants
said to have complemented. Consider two genes a1 and a2,
$, £, and ¥ are as follows:
named after their mutant alleles. We can represent the het-
erozygotes as follows, depending on whether the genes are white $ × white £ → F1 , all white
on the same chromosome or are on different chromosomes: white $ × white ¥ → F1 , all blue
Same chromosome: white £ × white ¥ → F1 , all blue
a1 1 From this set of results, we can conclude that mutants $ and
£ must be caused by alleles of one gene (say, w1) because they
1 a2 do not complement, but ¥ must be caused by a mutant allele of
another gene (w2) because ¥ complements both $ and £.
Different chromosomes:
a1 1 KEY CONCEPT When two independently derived recessive
mutant alleles producing similar recessive phenotypes fail to
complement, they must be alleles of the same gene.
1 a2
You can see that each locus has one wild-type allele to pro- How does complementation work at the molecular
vide wild-type function, resulting in wild-type progeny. level? The normal blue color of the harebell flower is caused
However, if the progeny are not wild type, then the by a blue pigment called anthocyanin. Pigments are chem-
recessive mutations must be alleles of the same gene. icals that absorb certain colors of light; in regard to the
Because both alleles of the gene are mutants, there is no
wild-type allele to provide wild-type function. These alleles Harebell plant
could have different mutant sites within the same gene, but
they would both be nonfunctional. Consider two recessive
mutations, a′ and a′′, of a gene whose wild type allele is a+.
The heterozygote a′ /a′′ would be
a
a
= mutation
+
Since there is no a allele to provide wild-type function, the
progeny will not be wild type.
At the operational level, complementation is defined as
the production of a wild-type phenotype when two haploid
genomes bearing different recessive mutations are united in Flowers of the harebell plant (Campanula species).
the same cell. [Gregory G. Dimijian, M.D./Science Source.]
harebell, the anthocyanin absorbs all wavelengths except The nuclei in a heterokaryon do not generally fuse. In one
blue, which is reflected into the eye of the observer. How- sense, this condition is a “mimic” diploid.
ever, this anthocyanin is made from chemical precursors Assume that, in different strains, there are mutations
that are not pigments; that is, they do not absorb light of in two different genes conferring the same mutant pheno-
any specific wavelength and simply reflect back the white type—for example, an arginine requirement. We will call
light of the sun to the observer, giving a white appearance. these genes arg-1 and arg-2. The genotypes of the two strains
The blue pigment is the end product of a series of biochem- can be represented as arg -1 ⋅ arg -2+ and arg -1+ ⋅ arg -2 .
ical conversions of nonpigments. Each step is catalyzed by a These two strains can be fused to form a heterokaryon with
specific enzyme encoded by a specific gene. We can explain the two nuclei in a shared cytoplasm:
the results with a pathway as follows:
Nucleus 1 is arg -1 ⋅ arg -2+
gene w1 gene w2 Nucleus 2 is arg -1+ ⋅ arg -2
Because gene products are made in a common cytoplasm,
the two wild-type alleles can exert their dominant effect and
enzyme 1 enzyme 2
cooperate to produce a heterokaryon of wild-type pheno-
type. In other words, the two mutations complement, just as
precursor 1 precursor 2 blue anthocyanin they would in a diploid. If the mutations had been alleles of
the same gene, there would have been no complementation.
A homozygous mutation in either of the genes will lead
to the accumulation of a precursor that will simply make
Analyzing double mutants of random
the plant white. Now the mutant designations could be
written as follows:
mutations
Recall that, to learn whether two genes interact, we need
$ w1$ /w1$ ⋅ w2+ /w2+
to assess the phenotype of the double mutant to see if it is
£ w1£ /w1£ ⋅ w2+ /w2+ different from the combination of both single mutations.
¥ w1+ /w1+ ⋅ w2¥ /w2¥ The double mutant is obtained by intercrossing. The F1 is
obtained as part of the complementation test; so with the
However, in practice, the subscript symbols would be
assumption that complementation has been observed, sug-
dropped and the genotypes would be written as follows:
gesting different genes, the F1 is selfed or intercrossed to
$ w1/w1 ⋅ w2+ /w2+ obtain an F2 homozygous for both mutations. This double
£ w1/w1 ⋅ w2+ /w2+ mutant may then be identified by looking for Mendelian
ratios. For example, if a standard 9:3:3:1 Mendelian ratio
¥ w1+ /w1+ ⋅ w2 /w2
is obtained, the phenotype present in only 1/16 of the prog-
Hence, an F1 from $ × £ will be eny represents the double mutant (the “1” in 9:3:3:1). In
cases of gene interaction, however, the phenotype of the
w1/w1 ⋅ w2+ /w2+
double mutant may not be distinct but will match that of
These F1 plants will have two defective alleles for w1 and one of the single mutants. In this case, a modified Mende-
will therefore be blocked at step 1. Even though enzyme 2 is lian ratio will result, such as 9:3: 4 or 9: 7.
fully functional, it has no substrate on which to act; so no blue The standard 9:3:3:1 Mendelian ratio is the simplest
pigment will be produced, and the phenotype will be white. case, expected if there is no gene interaction and if the two
The F1 plants from the other crosses, however, will have the mutations under test are on different chromosomes. This
wild-type alleles for both of the enzymes needed to take the 9:3:3:1 ratio is the null hypothesis: any modified Mende-
intermediates to the final blue product. Their genotypes will be lian ratio representing a departure from this null hypothesis
would be informative, as the following examples will show.
w1+ /w1 ⋅ w2+ /w2
KEY CONCEPT A range of modified 9: 3: 3:1 F1 ratios can
Hence, we see that complementation is actually a result
reveal specific types of gene interaction.
of the cooperative interaction of the wild-type alleles of the
two genes. Figure 5-15 summarizes the interaction of the
complementing and noncomplementing white mutants at The 9 : 3 : 3 :1 ratio: no gene interaction As a base-
the genetic and cellular levels. line, let’s start with the case in which two mutated genes
In a haploid organism, the complementation test can- do not interact, a situation where we expect the 9:3:3:1
not be performed by intercrossing. In fungi, an alternative ratio. Let’s look at the inheritance of skin coloration
method brings mutant alleles together to test complemen- in corn snakes. The snake’s natural color is a repeat-
tation: fusion resulting in a heterokaryon ( Figure 5-16). ing black-and-orange camouflage pattern, as shown in
Fungal cells fuse readily. When two different strains fuse, Figure 5-17a. The phenotype is produced by two separate
the haploid nuclei from the different strains occupy one pigments, both of which are under genetic control. One
cell, which is the heterokaryon (Greek; different kernels). gene determines the orange pigment, and the alleles that
F1
No
complementation Complementation
$ + + ¥
£ + £ +
Enzyme 1 Enzyme 2
No substrate Enzyme 2
Colorless No Colorless Colorless
White Blue
precursor 1 precursor 2 precursor 1 precursor 2
we will consider are o+ (presence of orange pigment) and black pigment (Figure 5-17c). The double homozygous
o (absence of orange pigment). Another gene determines recessive o /o ; b /b is albino (Figure 5-17d). Notice, how-
the black pigment, and its alleles are b+ (presence of black ever, that the faint pink color of the albino is from yet
pigment) and b (absence of black pigment). These two another pigment, the hemoglobin of the blood that is visi-
genes are unlinked. The natural pattern is produced by ble through this snake’s skin when the other pigments are
the genotype o+ /− ; b+ /− . (The dash represents the pres- absent. The albino snake also clearly shows that there is
ence of either allele.) A snake that is o /o ; b+ /− is black another element to the skin-pigmentation pattern in addi-
because it lacks the orange pigment (Figure 5-17b), and tion to pigment: the repeating motif in and around which
a snake that is o+ /− ; b /b is orange because it lacks the pigment is deposited.
Fusion
Heterokaryon grows
without arginine
FIGURE 5-16 A heterokaryon of Neurospora and similar fungi mimics a diploid state. When
vegetative cells fuse, haploid nuclei share the same cytoplasm in a heterokaryon. In this example,
haploid nuclei with mutations in different genes in the arginine synthetic pathway complement to
produce a Neurospora culture that no longer requires arginine. Functional enzyme produced by
arg-1+ is shown in purple, and functional enzyme produced by arg-2+ is shown in red.
If a homozygous orange and a homozygous black snake Here, however, an F2 shows a standard 9:3:3:1 ratio:
are crossed, the F1 is wild type (camouflaged), demonstrat-
♀ o+ /o ; b+ /b × ♂ o+ /o ; b+ /b
ing complementation:
(camouflaged) (camouflaged)
♀ o+ /o+ ; b /b × ♂ o /o ; b+ /b+ ↓
(orange) (black) F2 9 o+ /– ; b+ / – (camouflaged)
↓
3 o+ /– ; b /b (orange)
F1 o+ /o ; b+ /b
(camouflaged) 3 o /o ; b+ / – (black)
1 o /o ; b /b (albino)
The 9:3:3:1 ratio is produced because the two pigment
Independently synthesized and inherited pigments genes act independently at the cellular level.
}
(a) (b) b +
precursor → black pigment
camouflaged
+
o
precursor → orange pigment
If the presence of one mutant makes one pathway fail,
the other pathway is still active, producing the other pig-
ment color. Only when both mutants are present do both
pathways fail, and no pigment of any color is produced.
The 9:7 ratio: genes in the same pathway The F2 ratio
from the harebell dihybrid cross shows both blue and white
(c) (d) plants in a ratio of 9: 7. How can such results be explained?
Introduction to Genetic Analysis, 11e The 9: 7 ratio is clearly a modification of the dihybrid 9:3:3:1
Figure 06.13 #619
04/29/14 ratio with the 3:3:1 combined to make 7; hence, some kind
Dragonfly Media Group of interaction is inferred. The cross of the two white lines and
subsequent generations can be represented as follows:
w1/w1 ; w2+ /w2+ (white) × w1+ /w1+ ; w2 /w2 (white)
↓
F1 w1 /w1 ; w2+ /w2 (blue)
+
↓
FIGURE 5-17 In corn snakes, combinations of orange and black
F2 9 w1+ /– ; w2+ /– (blue) 9
pigments determine the four phenotypes shown. (a) A wild-type black-
}
and-orange camouflaged snake synthesizes both black and orange 3 w1+ / – ; w2 /w2 (white)
pigments. (b) A black snake does not synthesize orange pigment. +
3 w1/w1 ; w2 / – (white) 7
(c) An orange snake does not synthesize black pigment. (d) An albino
snake synthesizes neither black nor orange pigment. [Anthony Griffiths.] 1 w1/w1 ; w2 /w2 (white)
Clearly, in this case, the only way in which a 9: 7 ratio the F1 dihybrid will also result in a 9: 7 phenotypic ratio in
is possible is if the double mutant has the same phenotype the F2:
as the two single mutants. Hence, the modified ratio consti-
tutes a way of identifying the double mutant’s phenotype. Functional a +
Furthermore, the identical phenotypes of the single and Proportion Genotype protein Ratio
double mutants suggest that each mutant allele controls a 9 + +
r /− ; a /− Yes 9
16
different step in the same pathway. The results show that
}
a plant will have white petals if it is homozygous for the 3
16
r + /− ; a /a No
recessive mutant allele of either gene or both genes. To have 3 r /r ; a+ /− No 7
the blue phenotype, a plant must have at least one copy of 16
(b) r a+
Mutation in
the gene that No protein A
encodes the produced
regulatory Nonfunctional
protein regulatory protein
(c) r+ a
Mutation in
the gene that Mutant protein A
encodes the produced
structural
protein
r a
(d)
Mutation in No protein A
both genes produced
A model for recessive epistasis In the F2 , the 9:3: 4 phenotypic ratio is diagnostic of
recessive epistasis. As in the preceding case, we see, again,
that the ratio tells us what the phenotype of the double
must be, because the 16 4 component of the ratio must be a
Dihybrid w +/ w ; m + /m
one of the two mutant phenotypes; so, by definition, white
must be epistatic to magenta. (To find the double mutant
within the group, white F2 plants would have to be individ-
ually testcrossed.) This interaction is called recessive epis-
Selfed tasis because a recessive phenotype (white) overrides the
other phenotype. Dominant epistasis will be considered in
the next section.
9
16
w + / – ; m + / – Both enzymes active At the cellular level, we can account for the recessive
w+ m+ epistasis in Collinsia by the following type of pathway (see
Enzyme 1 Enzyme 2 9
also Figure 5-19).
+ +
gene w gene m
colorless → magenta → blue
Notice that the epistatic mutation occurs in a step in the
3
w +/ – ; m/m Blocked at second enzyme pathway leading to blue pigment; this step is upstream of
16
w+ the step that is blocked by the masked mutation.
Another informative case of recessive epistasis is the
Enzyme 1 3 yellow coat color of some Labrador retriever dogs. Two
alleles, B and b, stand for black and brown coats, respec-
tively. The two alleles produce black and brown melanin.
3 The allele e of another gene is epistatic on these alleles,
16
w/w ; m + / – Blocked at first enzyme
giving a yellow coat ( Figure 5-20 ). Therefore, the geno-
m+
types B /− ; e /e and b /b ; e /e both produce a yellow phe-
Enzyme 2 notype, whereas B /− ; E / − and b /b ; E /− are black and
No substrate brown, respectively. This case of epistasis is not caused by
an upstream block in a pathway leading to dark pigment.
4
1 Yellow dogs can make black or brown pigment, as can be
w/w ; m/m Blocked at first enzyme
16 seen in their noses and lips. The action of the allele e is
to prevent the deposition of the pigment in hairs. In this
case, the epistatic gene is developmentally downstream; it
represents a kind of developmental target that must be of
FIGURE 5-19 Wild-type alleles of two genes (w + and m+ ) encode E genotype before pigment can be deposited.
enzymes catalyzing successive steps in the synthesis of a blue petal
pigment. Homozygous m /m plants produce magenta flowers, and
KEY CONCEPT Epistasis is inferred when a mutant allele of
homozygous w /w plants produce white flowers. The double mutant
w /w ; m /m also produces white flowers, indicating that white is epistatic to one gene masks the expression of a mutant allele of another
magenta. gene and expresses its own phenotype instead.
FIGURE 5-20 Three different coat colors in Labrador retrievers. Two alleles B and b of a pigment
gene determine (a) black and (b) brown, respectively. At a separate gene, E allows color deposition in
the coat, and e /e prevents deposition, resulting in (c) the gold phenotype. Part c illustrates recessive
epistasis. [Anthony Griffiths.]
phenotype is the b phenotype, then a is being overridden. If Dominant epistasis due to a white mutation
both phenotypes are present, then there is no epistasis.
mutant to mutation-causing agents such as high-energy radi- A molecular mechanism for suppression
ation, and screen the descendants for wild types. In haploids
such as fungi, screening is accomplished by simply plating
mutagenized cells and looking for colonies with wild-type
phenotypes. Most wild types arising in this way are merely m+ s+
reversals of the original mutational event and are called
revertants. However, some will be “pseudorevertants,” dou- Wild
type Active protein
ble mutants in which one of the mutations is a suppressor. complex
Revertant and suppressed states can be distinguished by
appropriate crossing. For example, in yeast, the two results
m s+
would be distinguished as follows:
First
true revertant a+ × standard wild-type a+ mutation
Inactive
↓
Progeny all a+
m s
suppressed mutant a ⋅ s × standard wild-type a+ ⋅ s+ Second
↓ mutation
acting as
Progeny a+ ⋅ s+ wild type suppressor
Active protein
complex
a+ ⋅ s wild type
a ⋅ s+ original mutant m+ s
a⋅s wild type (suppressed) Suppressor
mutation
The appearance of the original mutant phenotype iden- alone Inactive
tifies the parent as a suppressed mutant.
In diploids, suppressors produce various modified F2
ratios, which are useful in confirming suppression. Let’s
look at a real-life example from Drosophila. The recessive FIGURE 5-22 A first mutation alters the binding site of one protein so
allele pd results in purple eye color when unsuppressed. that it can no longer bind to a partner. A suppressor mutation in the partner
A recessive allele su has no detectable phenotype itself but alters the binding site so that both proteins are able to bind once again.
suppresses the unlinked recessive allele pd. Hence, pd /pd ;
suppression is based on the physical binding of gene prod-
su /su is wild type in appearance and has red eyes. The
ucts in the cell—for example, protein–protein binding.
following analysis illustrates the inheritance pattern.
Assume that two proteins normally fit together to provide
A homozygous purple-eyed fly is crossed with a homozy-
some type of cellular function. When a mutation causes a
gous red-eyed stock carrying the suppressor.
shape change in one protein, it no longer fits together with
pd /pd ; su+ /su+ (purple) × pd + /pd + ; su /su (red) the other; hence, the function is lost (Figure 5-22). How-
↓ ever, a suppressor mutation that causes a compensatory
shape change in the second protein can restore fit and hence
F1 all pd + /pd ; su+ /su (red)
normal function. In this figure, if the genotypes were dip-
Self pd + /pd ; su+ /su (red) × pd + /pd ; su+ /su (red) loids representing an F2 from a dihybrid, then a 14: 2 ratio
↓ would result because the only mutant genotypes would be
}
F2 + + m/m ⋅ s+ /s+ (1 16) and m+ /m+ ⋅ s/s (1 16), totaling (2 16). If this
9 pd / – ; su /– red
were a haploid dihybrid cross (such as m+ s+ × m s), a 1:1
+
3 pd / – ; su /su red 13 ratio would result. From suppressor ratios generally, interact-
1 pd /pd ; su /su red ing proteins often can be deduced.
3 pd /pd ; su / –+
purple 3 Alternatively, in situations in which a mutation
causes a block in a metabolic pathway, the suppressor
13
The overall ratio in the F2 is 13 red:3 purple. The 16 com- finds some way of bypassing the block—for example, by
ponent must include the double mutant, which is clearly wild rerouting into the blocked pathway intermediates similar
type in phenotype. This ratio is expected from a recessive to those beyond the block. In the following example, the
suppressor that itself has no detectable phenotype. suppressor provides an intermediate B to circumvent the
Suppression is sometimes confused with epistasis. How- block.
ever, the key difference is that a suppressor cancels the
expression of a mutant allele and restores the correspond- No suppressor
ing wild-type phenotype. Furthermore, often only two phe- A B product
notypes segregate (as in the preceding examples) rather
With suppressor
than three, as in epistasis.
How do suppressors work at the molecular level? There A B product
are many possible mechanisms. A particularly useful type of B
In several organisms, nonsense suppressors have been A model for synthetic lethality
found—mutations in tRNA genes resulting in an anticodon
that will bind to a premature stop codon within a mutant
coding sequence. Hence, the suppressor allows translation A+ B+ Wild type
full binding;
to proceed past the former block and make a complete
fully functional
protein rather than a truncated one. Such suppressor muta-
tions often have little effect on the phenotype other than in DNA
suppression.
A– B+ Mutant A
KEY CONCEPT Mutant alleles called suppressors cancel the partial binding;
effect of a mutant allele of another gene, resulting in wild-type functional
phenotype.
SUMMARY
A gene does not act alone; rather, it acts in concert with The interaction of different genes can be detected by
many other genes in the genome. In forward genetic analy- testing double mutants because allele interaction implies
sis, deducing these complex interactions is an important stage interaction of gene products at the functional level. Some
of the research. Individual mutations are first tested for their key types of interaction are epistasis, suppression, and
dominance relations, a type of allelic interaction. Recessive synthetic lethality. Epistasis is the replacement of a mutant
mutations are often a result of haplosufficiency of the wild- phenotype produced by one mutation with a mutant pheno-
type allele, whereas dominant mutations are often the result type produced by mutation of another gene. The observation
either of haploinsufficiency of the wild type or of the mutant of epistasis suggests a common developmental or chemical
acting as a dominant negative (a rogue polypeptide). Some pathway. A suppressor is a mutation of one gene that can
mutations cause severe effects or even death (lethal muta- restore wild-type phenotype to a mutation at another gene.
tions). Lethality of a homozygous recessive mutation is a way Suppressors often reveal physically interacting proteins or
to assess if a gene is essential in the genome. nucleic acids. Some combinations of viable mutants are
The interaction of different genes is a result of their par- lethal, a result known as synthetic lethality. Synthetic lethals
ticipation in the same or connecting pathways of various can reveal a variety of interactions, depending on the nature
kinds—synthetic, signal transduction, or developmental. of the mutations.
Genetic dissection of gene interactions begins by the exper- The different types of gene interactions produce F2 dihy-
imenter amassing mutants affecting a character of interest. brid ratios that are modifications of the standard 9:3:3:1.
The complementation test determines whether two distinct For example, recessive epistasis results in a 9:3: 4 ratio.
recessive mutations are of one gene or of two different In more general terms, gene interaction and gene-
genes. The mutant genotypes are brought together in an F1 environment interaction are revealed by incomplete
individual, and if the phenotype is mutant, then no comple- penetrance (the ability of a genotype to express itself in
mentation has occurred and the two alleles must be of the the phenotype) and variable expressivity (the quantitative
same gene. If the phenotype is wild type, then complementa- degree of phenotypic manifestation of a genotype).
tion has occurred, and the alleles must be of different genes.
KEY TERMS
allelic series (multiple alleles) (p. 154) functional RNA (p. 163) permissive temperature (p. 160)
codominance (p. 156) heterokaryon (p. 166) pleiotropic allele (p. 160)
complementation (p. 165) incomplete dominance (p. 156) restrictive temperature (p. 160)
complementation test (p. 165) lethal allele (p. 158) revertant (p. 172)
dominant negative mutation (p. 156) modifier (p. 173) suppressor (p. 171)
double mutants (p. 164) multiple alleles (p. 154) synthetic lethal (p. 173)
epistasis (p. 169) null mutation (p. 156) temperature-sensitive (ts) mutations
essential gene (p. 158) one-gene–one-polypeptide hypothesis (p. 160)
expressivity (p. 161) (p. 163)
full (complete) dominance (p. 154) penetrance (p. 160)
SOLVED PROBLEMS
I
1 2
II
1 2 3 4 5 6 7 8 9 10 11
III 4
5 6 7 8 9 10 11 12 13 14 15 16 17
IV 4
5 6 7 8 9
SOLUTION 8 w t /w t × w t /w t all w t /w t
96: 24:8 are expected, but the actual numbers fit these Cross 2 can be partly written out without further
expectations remarkably well. thought by using our arbitrary gene symbols:
One of the principles of this chapter is that modi-
A/A ; –/– a/a ; B/B
fied Mendelian ratios reveal gene interactions. Cross 3 3
A/– ; –/–
gives F2 numbers appropriate for a modified dihybrid 4
A/a ; B/–
Mendelian ratio, and so it looks as if we are dealing 1
with a two-gene interaction. It seems the most promis- 4 a/a ; B/–
ing place to start; we can return to crosses 1 and 2 and We know that the F2 of cross 2 shows single-gene seg-
try to fit them in later. regation, and it seems certain now that the A /a allelic
Any dihybrid ratio is based on the phenotypic pro- pair has a role. But the B allele is needed to produce
portions 9:3:3:1. Our observed modification groups the spiny-tip phenotype, and so all plants must be
them as follows: homozygous B /B:
9 A /– ; B /–
3 A/– ; b /b
} 12 piping A/A ; B/B a/a ; B/B
3
4 A/– ; B/B
3 a /a ; B /– 3 spiny tip A/a ; B/B
1
1 a /a ; b /b 1 spiny 4 a/a ; B/B
So, without worrying about the name of the type of Notice that the two single-gene segregations in
gene interaction (we are not asked to supply this any- crosses 1 and 2 do not show that the genes are not
way), we can already define our three pineapple-leaf interacting. What is shown is that the two-gene inter-
phenotypes in relation to the proposed allelic pairs A /a action is not revealed by these crosses—only by cross
and B /b: 3, in which the F1 is heterozygous for both genes.
piping = A /– (B /b irrelevant) b. Now it is simply a matter of using Mendel’s laws to
predict cross outcomes:
spiny tip = a /a ; B /–
spiny = a /a ; b /b
}
(1) A /a ; B /b × a /a ; b /b
→ 1 A /a ; B /b
4
piping
What about the parents of cross 3? The spiny parent (independent 1
4
A /a ; b /b
must be a /a ; b /b, and, because the B gene is needed assortment in
to produce F2 spiny-tip leaves, the piping parent must
1 a /a ; B /b spiny tip
a standard 4
be A /A ; B /B. (Note that we are told that all parents testcross) 1 a /a ; b /b spiny
4
are pure, or homozygous.) The F1 must therefore be
A /a ; B /b. (2) A /a ; B /b × a /a ; B /b
→
Without further thought, we can write out cross 1
}
3 3
as follows: 4
B /− → 8
1 A /a 1
2 2
piping
1 b /b → 1
a/a ; B/B a/a ; b/b 4 8
3
a/a ; B/– 3 3
4
B /− → spiny tip
a/a ; B/b 1 a /a
4 8
1 2 1 1
4 a/a ; b/b 4
b /b → 8
spiny
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a Problem Solving Video.
WORKING WITH THE FIGURES c. In the system defined in this figure, if we assume
(The first 19 questions require inspection of text figures.) the gene codes for an enzyme catalyzing the
synthesis of a black pigment, what would be the
1. a. In Figure 5-1, what do the yellow stars represent?
phenotype of the heterozygote?
b. Explain in your own words why the heterozygote
is functionally wild type.
d. At the structural level, what might be the difference e. Do you think it would be possible to treat albinism
between the proteins represented by the colors by ingesting melanin? (Research this possibility
orange and yellow? yourself.)
2. a. In Figure 5-2, explain how the mutant polypeptide 12. a. In Figure 5-15, what do the dollar, pound, and yen
acts as a spoiler and what its net effect on pheno- symbols represent?
type is. b. Why can’t the left-hand F1 heterozygote synthesize
b. What might cause a bend in the mutant protein? blue pigment?
c. In Model 1, what can you say about the pos- c. Draw out the results of crossing the $ and ¥ lines.
sibility of up-regulation of the protein in the d. Write out all the genotypes and phenotypes of
heterozygote? progeny from a self of the blue F1.
3. In Figure 5-4, does the photo show the blood of one e. Write out all the genotypes and phenotypes pro-
individual (if so, which one), or a mixture of bloods (if duced by crossing the blue F1 to the white F1.
so, which ones)?
13. a. In Figure 5-16, explain at the protein level why this
4. a. In Figure 5-5, what is the object represented by the heterokaryon can grow on minimal medium.
color blue?
b. A heterokaryon produces spores by pinching off
b. Is it true to say that the sickle-cell hemoglobin cells that contain a few nuclei. Will any of these
migrates faster than normal hemoglobin? spores be arginine-dependent? Explain.
c. What might cause the different migration rates? 14. a. In Figure 5-17, write possible genotypes for each of
5. a. In Figure 5-6, assess the allele V f with respect to the four snakes illustrated.
the V by allele. Is it dominant? Recessive? Codomi- b. Explain in short sentences the meaning of the
nant? Incompletely dominant? header for this figure.
b. In this figure, is there any case in which the hetero- 15. a. In Figure 5-18, which panel represents the double
zygote has a truly new phenotype? mutant?
c. Predict the phenotype of the heterozygotes of v b. State the function of the regulatory gene.
combined with the other alleles.
c. In the situation in panel b, would protein from the
6. In Figure 5-7, if you assume that all the progeny are active protein gene be made?
visible, is the observed color ratio the one expected?
d. What is the function of the pale green region?
7. In Figure 5-9, propose a specific genetic explanation
e. What is the element represented in yellow?
for individual Q (give a possible genotype, defining the
alleles). f. Panels b and d have the same outcome: state the
two different mechanisms that produce this.
8. In Figure 5-10, point to the individuals that show full
expressivity. 16. a. In Figure 5-19, if you selfed 10 different F2 pink
plants, would you expect to find any white-
9. Speculate logically on the minimum number of mod-
flowered plants among the offspring? Any blue-
ifying genes and alleles that could produce the varia-
flowered plants?
tion shown by the allele S P in Figure 5-11.
b. Some white F2 plants have a functional enzyme 2:
10. From a knowledge of the structures shown in Figure
Since enzyme 2 produces the blue pigment, why
5-13, do you think Beadle and Tatum might have had
are these plants not blue?
a clue about the sequential steps in the synthetic path-
way before doing their genetic tests? 17. In Figure 5-21, write down possible genotypes for each
of the three petals.
11. a. In Figure 5-14, in view of the position of HPA oxi-
dase earlier in the pathway compared to that of 18. a. In Figure 5-22, what do the square/triangular pegs
HA oxidase, would you expect people with tyrosi- and holes represent?
nosis to show symptoms of alkaptonuria? b. Is the suppressor mutation alone wild type in
b. If a double mutant could be found, would you phenotype?
expect tyrosinosis to be epistatic to alkaptonuria? c. Would it be reasonable to call the s gene a suppres-
c. Do you think it might be possible to cure the symp- sor gene in a wild-type cell?
toms of PKU by ingesting tyrosine? (Research this 19. a. In Figure 5-23, explain why the interacting alleles
possibility yourself.) are called synthetic lethals.
d. How might you treat cretinism? (Research this b. For the model to work, is it essential that the red
possibility yourself.) and blue proteins bind to each other?
28. Radishes may be long, round, or oval, and they may be 32. Consider two blood polymorphisms that humans
red, white, or purple. You cross a long, white variety have in addition to the ABO system. Two alleles LM
with a round, red one and obtain an oval, purple F1. and LN determine the M, N, and MN blood groups.
The F2 shows nine phenotypic classes as follows: 9 The dominant allele R of a different gene causes a
long, red; 15 long, purple; 19 oval, red; 32 oval, pur- person to have the Rh+ (rhesus positive) phenotype,
ple; 8 long, white; 16 round, purple; 8 round, white; whereas the homozygote for r is Rh− (rhesus nega-
16 oval, white; and 9 round, red. tive). Two men took a paternity dispute to court,
a. Provide a genetic explanation of these results. Be each claiming three children to be his own. The blood
sure to define the genotypes and show the constitu- groups of the men, the children, and their mother
tion of the parents, the F1, and the F2. were as follows:
c. What additional experimental evidence might you a. Were the mutations in the two auxotrophs in the
collect to test your hypothesis in part b? same gene in the pathway for synthesizing leu-
d. In Berlin, about 1 human in 1000 shows a Pelger cine, or in two different genes in that pathway?
anomaly of white blood cells very similar to that Explain.
described for rabbits. The anomaly is inherited as a b. Write the genotype of the two strains according to
simple dominant, but the homozygous type has not your model.
been observed in humans. Based on the condition c. What progeny, and in what proportions, would
in rabbits, why do you suppose the human homo- you predict from crossing the two auxotrophic
zygous has not been observed? mutants? (Assume independent assortment.)
e. Again by analogy with rabbits, what phenotypes 39. A yeast geneticist irradiates haploid cells of a strain
and genotypes would you expect among the that is an adenine-requiring auxotrophic mutant,
children of a man and woman who both show the caused by mutation of the gene ade1. Millions of the
Pelger anomaly? irradiated cells are plated on minimal medium, and a
(Data from A. M. Srb, R. D. Owen, and R. S. Edgar, small number of cells divide and produce prototro-
General Genetics, 2nd ed. W. H. Freeman and Com- phic colonies. These colonies are crossed individu-
pany, 1965.) ally with a wild-type strain. Two types of results are
35. Two normal-looking fruit flies were crossed, and, in obtained:
the progeny, there were 202 females and 98 males. 1. prototroph × wild type: progeny all prototrophic
a. What is unusual about this result? 2. prototroph × wild type: progeny 75% prototrophic,
b. Provide a genetic explanation for this anomaly. 25% adenine-requiring auxotrophs
a. Explain the difference between these two types
c. Provide a test of your hypothesis.
of results.
36. You have been given a virgin Drosophila female. You
b. Write the genotypes of the prototrophs in each
notice that the bristles on her thorax are much shorter
case.
than normal. You mate her with a normal male (with
long bristles) and obtain the following F1 progeny: c. What progeny phenotypes and ratios do you
1 short-bristled females, 1 long-bristled females, and predict from crossing a prototroph of type 2 by
3 3
1 long-bristled males. A cross of the F long-bristled the original ade1 auxotroph?
3 1
females with their brothers gives only long-bristled F2 . 40. In roses, the synthesis of red pigment is by two steps in
A cross of short-bristled females with their brothers a pathway, as follows:
gives 13 short-bristled females, 13 long-bristled females, gene P
and 13 long-bristled males. Provide a genetic hypothesis colorless intermediate →
to account for all these results, showing genotypes in gene Q
magenta intermediate → red pigment
every cross.
a. What would the phenotype be of a plant homozy-
37. A dominant allele H reduces the number of body bris- gous for a null mutation of gene P?
tles that Drosophila flies have, giving rise to a “hair-
less” phenotype. In the homozygous condition, H is b. What would the phenotype be of a plant homozy-
lethal. An independently assorting dominant allele S gous for a null mutation of gene Q?
has no effect on bristle number except in the presence c. What would the phenotype be of a plant homozy-
of H, in which case a single dose of S suppresses the gous for null mutations of genes P and Q?
hairless phenotype, thus restoring the hairy pheno- d. Write the genotypes of the three strains in parts a,
type. However, S also is lethal in the homozygous (S /S) b, and c.
condition.
e. What F2 ratio is expected from crossing plants from
a. What ratio of hairy to hairless flies would you find parts a and b? (Assume independent assortment.)
in the live progeny of a cross between two hairy
flies both carrying H in the suppressed condition? 41. Because snapdragons (Antirrhinum) possess the pig-
ment anthocyanin, they have reddish purple petals.
b. When the hairless progeny are backcrossed with Two pure anthocyaninless lines of Antirrhinum were
a parental hairy fly, what phenotypic ratio would developed, one in California and one in Holland.
you expect to find among their live progeny? They looked identical in having no red pigment
38. After irradiating wild-type cells of Neurospora (a hap- at all, manifested as white (albino) flowers. How-
loid fungus), a geneticist finds two leucine-requiring ever, when petals from the two lines were ground
auxotrophic mutants. He combines the two mutants in up together in buffer in the same test tube, the solu-
a heterokaryon and discovers that the heterokaryon is tion, which appeared colorless at first, gradually
prototrophic. turned red.
blue × white blue 101 blue, 33 white 22. Draw chromosomes representing the meioses in the
parents in the cross blue × white and representing
blue × pink blue 192 blue, 63 pink meiosis in the F1.
pink × white blue 272 blue, 121 white, 89 pink 23. Repeat step 22 for the cross blue × pink.
a. Explain these results genetically. Define the allele Now try to solve the problem. If you are unable to do so,
symbols that you use, and show the genetic con- try to identify the obstacle and write a sentence or two
stitution of the parents, the F1, and the F2 in each describing your difficulty. Then go back to the expansion
cross. questions and see if any of them relate to your difficulty. If
this approach does not work, inspect the Learning Objec-
b. A cross between a certain blue F2 plant and a certain
tives and Key Concepts of this chapter and ask yourself
white F2 plant gave progeny of which 83 were blue, 81
which might be relevant to your difficulty.
were pink, and 12 were white. What must the geno-
types of these two F2 plants have been? 44. A woman who owned a purebred albino poodle (an
autosomal recessive phenotype) wanted white pup-
www
UNPACKING PROBLEM 43 pies. She took the dog to a breeder, who said he would
www
mate the female with an albino stud male, also from a
Before attempting a solution to this problem, try answer- pure stock. When six puppies were born, all of them
ing the following questions: were black; so the woman sued the breeder, claiming
1. What is the character being studied? that he replaced the stud male with a black dog, giv-
ing her six unwanted puppies. You are called in as an
2. What is the wild-type phenotype?
expert witness, and the defense asks you if it is possible
a. From these results, deduce and explain the inheri- b. Another inherited color deviation in mice substi-
tance of these colors. tutes brown for the black color in the wild-type
hair. Such brown-agouti mice are called cinna-
b. Write the genotypes of each of the parents, the F1, mons. When wild-type mice are crossed with cin-
and the F2 in all crosses. namons, all of the F1 are wild type and the F2 has
47. Two albinos marry and have four normal children. a 3:1 ratio of wild type to cinnamon. Diagram this
How is this possible? cross as in part a, letting B stand for the wild-type
48. Consider the production of flower color in the Japa- black allele and b stand for the cinnamon brown
nese morning glory (Pharbitis nil). Dominant alleles allele.
of either of two separate genes (A /− ⋅ b /b or a /a ⋅ B /−) c. When mice of a true-breeding cinnamon line are
produce purple petals. A /− ⋅ B /− produces blue petals, crossed with mice of a true-breeding nonagouti
and a /a ⋅ b /b produces scarlet petals. Deduce the geno- (black) line, all of the F1 are wild type. Use a genetic
types of parents and progeny in the following crosses: diagram to explain this result.
d. why your scheme is the fastest and the most conve- Females 1
2
one band Males 1
2
one band
nient. 1 1
2
three bands 2
two bands
58. The following pedigree is for a dominant phenotype
governed by an autosomal allele. What does this ped- a. Provide a clear explanation of the inheritance of
igree suggest about the phenotype, and what can you these phenotypes.
deduce about the genotype of individual A? b. In accord with your model, what would be the
outcome of a cross between a three-banded daugh-
ter and a one-banded son?
62. In minks, wild types have an almost black coat. Breeders
A have developed many pure lines of color variants for
the mink-coat industry. Two such pure lines are plat-
inum (blue gray) and aleutian (steel gray). These lines
59. Petal coloration in foxgloves is determined by were used in crosses, with the following results:
three genes. M encodes an enzyme that synthesizes
anthocyanin, the purple pigment seen in these petals; Cross Parents F1 F2
m /m produces no pigment, resulting in the pheno- 1 wild × platinum wild 18 wild, 5 platinum
type albino with yellowish spots. D is an enhancer
2 wild × aleutian wild 27 wild, 10 aleutian
of anthocyanin, resulting in a darker pigment; d /d
does not enhance. At the third locus, w /w allows pig- 3 platinum × aleutian wild 133 wild
ment deposition in petals, but W prevents pigment 41 platinum
deposition except in the spots and so results in the 46 aleutian
white, spotted phenotype. Consider the following
two crosses: 17 sapphire (new)
a. Devise a genetic explanation of these three crosses. Do these results support the hypothesis that the
Show genotypes for the parents, the F1, and the F2 original plant was P /p ; Q /q?
in the three crosses, and make sure that you show 67. A plant of phenotype 1 was selfed, and, in the progeny,
the alleles of each gene that you hypothesize for there were 100 plants of phenotype 1 and 60 plants of
every mink. an alternative phenotype 2. Are these numbers com-
b. Predict the F1 and F2 phenotypic ratios from cross- patible with expected ratios of 9: 7, 13:3, and 3:1?
ing sapphire with platinum and with aleutian pure Formulate a genetic hypothesis on the basis of your
lines. calculations.
63. In Drosophila, an autosomal gene determines the 68. Four homozygous recessive mutant lines of Drosoph-
shape of the hair, with B giving straight and b giving ila melanogaster (labeled 1 through 4) showed abnor-
bent hairs. On another autosome, there is a gene of mal leg coordination, which made their walking highly
which a dominant allele I inhibits hair formation so erratic. These lines were intercrossed; the phenotypes
that the fly is hairless ( i has no known phenotypic of the F1 flies are shown in the following grid, in which
effect). “+ ” represents wild-type walking and “−” represents
a. If a straight-haired fly from a pure line is crossed abnormal walking:
with a fly from a pure-breeding hairless line 1 2 3 4
known to be an inhibited bent genotype, what
1 − + + +
will the genotypes and phenotypes of the F1 and
the F2 be? 2 + − − +
b. What cross would give the ratio 4 hairless: 3 + − − +
3 straight:1 bent? 4 + + + −
64. The following pedigree concerns eye phenotypes in
a. What type of test does this analysis represent?
Tribolium beetles. The solid symbols represent black
eyes, the open symbols represent brown eyes, and the b. How many different genes were mutated in creat-
cross symbols (X) represent the “eyeless” phenotype, ing these four lines?
in which eyes are totally absent. c. Invent wild-type and mutant symbols, and write
out full genotypes for all four lines and for the F1
I
1 2 3 flies.
d. Do these data tell us which genes are linked? If not,
II
1 2 3 4 5 how could linkage be tested?
e. Do these data tell us the total number of genes
III taking part in leg coordination in this animal?
1 2 3 4
IV 69. Three independently isolated tryptophan-requiring
1 mutants of haploid yeast are called trpB, trpD, and
trpE. Cell suspensions of each are streaked on a plate
a. From these data, deduce the mode of inheritance of
of nutritional medium supplemented with just enough
these three phenotypes.
tryptophan to permit weak growth for a trp strain.
b. Using defined gene symbols, show the genotype of The streaks are arranged in a triangular pattern so that
beetle II-3. they do not touch one another. Luxuriant growth is
65. A plant believed to be heterozygous for a pair of alleles noted at both ends of the trpE streak and at one end of
B /b (where B encodes yellow and b encodes bronze) the trpD streak (see the figure below).
was selfed, and, in the progeny, there were 280 yellow
and 120 bronze plants. Do these results support the
hypothesis that the plant is B /b?
66. A plant thought to be heterozygous for two inde-
pendently assorting genes (P /p ; Q /q) was selfed, and
the progeny were
88 P /– ; Q /– 25 p /p ; Q /–
32 P /– ; q /q 14 p /p ; q /q
70. A pure-breeding strain of squash that produced disk- long 32 sphere 178 disk 270
shaped fruits (see the accompanying illustration) was
crossed with a pure-breeding strain having long fruits. Propose an explanation for these results, and show the
The F1 had disk fruits, but the F2 showed a new pheno- genotypes of the P, F1, and F2 generations.
type, sphere, and was composed of the following pro- 71. Marfan’s syndrome is a disorder of the fibrous connec-
portions, shown in the next column: tive tissue, characterized by many symptoms, including
long, thin digits; eye defects; heart disease; and long
limbs. (Flo Hyman, the American volleyball star, suf-
fered from Marfan’s syndrome. She died from a rup-
tured aorta.)
II
III
Symptoms
Unknown, presumed normal Eye lens displacement Long fingers and toes
Examined, normal Congenital heart disease Very long, thin fingers and toes
Questionably affected
a. Use the pedigree above to propose a mode of inher- dominant allele, as specified in Problem 72. The domi-
itance for Marfan’s syndrome. nant allele Pr of a fourth independently assorting gene
b. What genetic phenomenon is shown by this is required to convert the biochemical precursor into
pedigree? a purple pigment, and its recessive allele pr makes the
pigment red. Plants that do not produce pigment have
c. Speculate on a reason for such a phenomenon. yellow seeds. Consider a cross of a strain of genotype
(Data from J. V. Neel and W. J. Schull, Human Heredity. A /A ; C /C ; R /R ; pr /pr with a strain of genotype a /a ;
University of Chicago Press, 1954.) c /c ; r /r ; Pr /Pr.
72. In corn, three dominant alleles, called A, C, and R, a. What are the phenotypes of the parents?
must be present to produce colored seeds. Genotype b. What will be the phenotype of the F1?
A /− ; C/− ; R /− is colored; all others are colorless. A col-
ored plant is crossed with three tester plants of known c. What phenotypes, and in what proportions, will
genotype. With tester a /a ; c /c ; R /R, the colored plant appear in the progeny of a selfed F1?
produces 50 percent colored seeds; with a /a ; C /C ; r /r, d. What progeny proportions do you predict from the
it produces 25 percent colored; and with A /A ; c /c ; r /r, testcross of an F1?
it produces 50 percent colored. What is the genotype 74. The allele B gives mice a black coat, and b gives a
of the colored plant? brown one. The genotype e /e of another, independently
73. The production of pigment in the outer layer of seeds of assorting gene prevents the expression of B and b,
corn requires each of the three independently assorting making the coat color beige, whereas E/– permits
genes A, C, and R to be represented by at least one the expression of B and b. Both genes are autosomal.
In the following pedigree, black symbols indicate A third pathway, whose compounds do not
a black coat, pink symbols indicate brown, and contribute pigment to petals, normally does not affect
unshaded symbols indicate beige. the blue and red pathways, but, if one of its interme-
diates (white3) should build up in concentration, it can
I
1 2 be converted into the yellow intermediate of the red
pathway.
In the diagram, the letters A through E represent
enzymes; their corresponding genes, all of which are
II
1 2 3 4 5 6 unlinked, may be symbolized by the same letters.
Assume that wild-type alleles are dominant and
encode enzyme function and that recessive alleles result
in a lack of enzyme function. Deduce which combi-
III
1 2 3 4 5 6 7 nations of true-breeding parental genotypes could be
crossed to produce F2 progeny in the following ratios:
a. What is the name given to the type of gene interac- a. 9 purple : 3 green: 4 blue
tion in this example?
b. 9 purple : 3 red : 3 blue : 1 white
b. What are the genotypes of the individual mice in
c. 13 purple: 3 blue
the pedigree? (If there are alternative possibilities,
state them.) d. 9 purple : 3 red : 3 green : 1 yellow
75. A researcher crosses two white-flowered lines of Antir- ( Note: Blue mixed with yellow makes green; assume
rhinum plants as follows and obtains the following that no mutations are lethal.)
results: 77. The flowers of nasturtiums ( Tropaeolum majus )
pure line 1× pure line 2 may be single (S), double (D), or superdouble (Sd).
Superdoubles are female sterile; they originated from
↓
a double-flowered variety. Crosses between varieties
F1 all white gave the progeny listed in the following table, in which
F1 × F1 pure means “pure breeding.”
↓ Cross Parents Progeny
F2 131 white 1 pure S × pure D All S
29 red 2 cross 1 F1 × cross 1 F1 78 S : 27 D
a. Deduce the inheritance of these phenotypes; use 3 pure D × Sd 112 Sd : 108 D
clearly defined gene symbols. Give the genotypes of
the parents, F1, and F2. 4 pure S × Sd 8 Sd : 7 S
b. Predict the outcome of crosses of the F1 with each 5 pure D × cross 4 Sd progeny 18 Sd : 19 S
parental line. 6 pure D × cross 4 S progeny 14 D : 16 S
76. Assume that two pigments, red and blue, mix to give the
Using your own genetic symbols, propose an explana-
normal purple color of petunia petals. Separate biochem-
tion for these results, showing
ical pathways synthesize the two pigments, as shown in
the top two rows of the accompanying diagram. “White” a. all the genotypes in each of the six rows.
refers to compounds that are not pigments. (Total lack b. the proposed origin of the superdouble.
of pigment results in a white petal.) Red pigment forms
78. In a certain species of fly, the normal eye color is
from a yellow intermediate that is normally at a concen-
red (R). Four abnormal phenotypes for eye color
tration too low to color petals.
were found: two were yellow (Y1 and Y2), one was
pathway I … › white1
E
→ blue brown (B), and one was orange (O). A pure line
pathway II … › white2
A
→ yellow
→ red B was established for each phenotype, and all possible
combinations of the pure lines were crossed. Flies of
c each F1 were intercrossed to produce an F2 . The F1 and
the F2 flies are shown within the following square; the
pathway III … → white3
D
→ white4 pure lines are given at the top and at the left-hand side.
a queen from the Brown line was mated with Van a. Explain the inheritance of these colors.
Scoy drones, all the F1 were nonhygienic. When drones b. Write the genotypes of the parents, the F1, and
from this F1 inseminated a queen from the Brown line, the F2.
the progeny behaviors were as follows:
84. Consider the following F1 individuals in different spe-
1 hygienic
4 cies and the F2 ratios produced by selfing:
1 uncapping but no removing of pupae
4
F1 Phenotypic ratio in the F2
1 nonhygienic
2 12 3 1
1 cream 16
cream 16
black 16
gray
However, when the compartment of dead pupae
9 7
was uncapped by the beekeeper and the nonhygienic 2 orange 16
orange 16
yellow
honeybees were examined further, about half the bees 13 3
3 black 16
black 16
white
were found to remove the dead pupae, but the other
half did not. 4 solid red 9 3 4
16
solid red 16
mottled red 16
small red dots
a. Propose a genetic hypothesis to explain these beha-
vioral patterns. If each F1 were testcrossed, what phenotypic ratios
would result in the progeny of the testcross?
b. Discuss the data in relation to epistasis, domi-
nance, and environmental interaction. 85. To understand the genetic basis of locomotion in
the diploid nematode Caenorhabditis elegans, reces-
(Note: Workers are sterile, and all bees from one line
sive mutations were obtained, all making the worm
carry the same alleles.)
“wiggle” ineffectually instead of moving with its
83. The normal color of snapdragons is red. Some pure usual smooth gliding motion. These mutations pre-
lines showing variations of flower color have been sumably affect the nervous or muscle systems. Twelve
found. When these pure lines were crossed, they gave homozygous mutants were intercrossed, and the F1
the following results (see the table): hybrids were examined to see if they wiggled. The
results were as follows, where a plus sign means that
Cross Parents F1 F2
the F1 hybrid was wild type (gliding) and “w” means
1 orange × yellow orange 3 orange : 1 yellow that the hybrid wiggled.
2 red × orange red 3 red : 1 orange
3 red × yellow red 3 red : 1 yellow
4 red × white red 3 red : 1 white
5 yellow × white red 9 red : 3 yellow : 4 white
6 orange × white red 9 red : 3 orange : 4 white
7 red × white red 9 red : 3 yellow : 4 white
1 2 3 4 5 6 7 8 9 10 11 12
1 w + + + w + + + + + + +
2 w + + + w + w + w + +
3 w w + + + + + + + +
4 w + + + + + + + +
5 w + + + + + + +
6 w + w + w + +
7 w + + + w w
8 w + w + +
9 w + + +
10 w + +
11 w w
12 w
a. Explain what this experiment was designed to test. b. What cross(es) would they do to make these
b. Use this reasoning to assign genotypes to all 12 worms?
mutants. c. What results would they expect in the F2 if
c. Explain why the phenotype of the F1 hybrids (1) it did act as a suppressor of bli-4?
between mutants 1 and 2 differed from that of the (2) it did not act as a suppressor of bli-4?
hybrids between mutants 1 and 5.
88. Six proline-requiring auxotrophic mutants were
86. A geneticist working on a haploid fungus makes a obtained in the haploid fungus Saccharomyces cerevisiae
cross between two slow-growing mutants called mossy (yeast). Each was crossed to wild type in order to
and spider (referring to the abnormal appearance of obtain the mutants in each mating type (“sex”), then
the colonies). Tetrads from the cross are of three types all combinations were crossed and the resultant ran-
(A, B, C), but two of them contain spores that do not dom ascospores were plated onto minimal medium. In
germinate. some cases, proline-independent colonies were obtained
Spore A B C (numbers are not shown); but in other cases, none of
the ascospores grew. The results are summarized in
1 wild type wild type spider
the following table, in which + indicates presence of
2 wild type spider spider c olonies, and 0 indicates no colonies. Formulate a
3 no germination mossy mossy hypothesis that explains the results of each cross.
4 no germination no germination mossy 1 2 3 4 5 6
1 0 + + 0 + 0
Devise a model to explain these genetic results, and
propose a molecular basis for your model. 2 + 0 0 + + +
87. In the nematode C. elegans, some worms have blistered 3 + 0 0 + + +
cuticles due to a recessive mutation in one of the bli 4 0 + + 0 + 0
genes. Someone studying a suppressor mutation that 5 + + + + 0 +
suppressed bli-3 mutations wanted to know if it would
6 0 + + 0 + 0
also suppress mutations in bli-4. They had a strain that
was homozygous for this recessive suppressor muta-
GENETICS AND SOCIETY
tion, and its phenotype was wild type.
a. How would they determine whether this recessive 1. How might a recessive lethal allele reveal itself in a
suppressor mutation would suppress mutations in human pedigree?
bli-4? In other words, what is the genotype of the 2. How has Beadle and Tatum’s pioneering work influ-
worms required to answer the question? enced the therapy of human disease?
6.5 TRANSDUCTION
LO 6.7 Predict the outcomes of transduction experiments using phages capable of
generalized or specialized transduction.
193
D
NA technology is responsible for the rapid advances recycling of nutrients such as nitrogen, sulfur, and carbon
being made in the genetics of all model organisms. It in ecosystems. Some are agents of human, animal, and plant
is also a topic of considerable interest in the public disease. Others live symbiotically inside our mouths and
domain. Examples are the highly publicized announcement intestines. In addition, many types of bacteria are useful for
of the full genome sequences of humans and chimpanzees the industrial synthesis of a wide range of organic products.
in recent years and the popularity of DNA-based forensic Hence, the impetus for the genetic dissection of bacteria
analysis in television shows and movies (Figure 6-1). Indeed, has been the same as that for multicellular o rganisms—to
improvements in technology have led to the sequencing of understand their biological function.
the genomes of many hundreds of species. Such dramatic Bacteria belong to a class of organisms known as
results, whether in humans, fish, insects, plants, or fungi, prokaryotes, which also includes the blue-green algae (clas-
are all based on the use of methods that permit small pieces sified as cyanobacteria). A key defining feature of prokaryotes
of DNA to be isolated, carried from cell to cell, and ampli- is that their DNA is not enclosed in a membrane-bounded
fied into large pure samples. The sophisticated systems nucleus. Like higher organisms, bacteria have genes composed
that permit these manipulations of the DNA of any organ- of DNA arranged in a long series on a chromosome. However,
ism are almost all derived from bacteria and their viruses. the organization of their genetic material is unique in several
Hence, the advance of modern genetics to its present state respects. The genome of most bacteria is a single molecule of
of understanding was entirely dependent on the develop- double-stranded DNA in the form of a closed circle. In addi-
ment of bacterial genetics, the topic of this chapter. tion, bacteria in nature often contain extra autonomous DNA
However, the goal of bacterial genetics has never been elements called plasmids. Most plasmids are also DNA circles
to facilitate eukaryotic molecular genetics. Bacteria are bio- but are much smaller than the main bacterial genome.
logically important in their own right. They are the most Bacteria can be parasitized by specific viruses called
numerous organisms on our planet. They contribute to the bacteriophages or, simply, phages. Phages and other viruses
are very different from the organisms that we have been
studying so far. Viruses have some properties in common
The fruits of DNA technology, made with organisms; for example, their genetic material can be
possible by bacterial genetics DNA or RNA, constituting a short “chromosome.” How-
ever, most biologists regard viruses as nonliving because
they are not cells and they have no metabolism of their
own. Hence, for the study of their genetics, viruses must be
propagated in the cells of their host organisms.
When scientists began studying bacteria and phages,
they were naturally curious about their hereditary systems.
Clearly, bacteria and phages must have hereditary systems
because they show a constant appearance and function
from one generation to the next (they are true to type). But
how do these hereditary systems work? Bacteria, like uni-
cellular eukaryotic organisms, reproduce asexually by cell
growth and division, one cell becoming two. This asexual
reproduction is quite easy to demonstrate experimentally.
However, is there ever a union of different types for the
purpose of sexual reproduction? Furthermore, how do the
much smaller phages reproduce? Do they ever unite for a
sex-like cycle? These questions are pursued in this chapter.
We will see that there is a variety of hereditary processes
FIGURE 6-1 The dramatic results of modern
DNA technology, such as sequencing the human in bacteria and phages. These processes are interesting
genome, were possible only because bacterial because of the basic biology of these forms, but they also
genetics led to the invention of efficient DNA act as models—as sources of insight into genetic processes
manipulation vectors. [Republished with at work in all organisms. For a geneticist, the attraction
permission of the American Association for the
of bacteria and phages is that they can be cultured in very
Advancement of Science, from Science vol. 291
16 February 2001 no. 5507 p. 1145–1434/image
large numbers because they are so small. Consequently, it
by Ann E. Cutting. Permission conveyed through is possible to detect and study very rare genetic events that
Copyright Clearance Center, Inc.] are difficult or impossible to study in eukaryotes.
194
Genome
Genome
Transduction
What hereditary processes are observed in bacteria? cell and inject it into another, where it can be incorporated
They can undergo both asexual and sexual reproduction. into the chromosome, in a process known as transduction.
Mutation occurs in asexual cells in much the same way as DNA transfer by conjugation, transformation, or trans-
it does in eukaryotes, and mutant alleles can be followed duction constitutes a process known as horizontal trans-
through both these processes in an approach analogous to mission, a type of gene transmission without the need for
that used in eukaryotes. We shall follow alleles in this way cell division. This term distinguishes this type of DNA
in the chapter ahead. transfer from that during vertical transmission, the pas-
When bacterial cells reproduce asexually, their genomic sage of DNA down thorough the bacterial generations.
DNA replicates and is partitioned into daughter cells, but Horizontal transmission can spread DNA rapidly through
the partitioning method is quite different from mitosis. a bacterial population by contact in much the same way
In sexual reproduction, two DNA molecules from dif- that a disease spreads. For bacteria, horizontal transmission
ferent sources are brought together. However, an impor provides a powerful method by which they can adapt rap-
tant difference from eukaryotes is that, in bacteria, rarely idly to changing environmental conditions.
are two complete chromosomes brought together; usually, Phages themselves can undergo recombination when
the union is of one complete chromosome plus a fragment two different genotypes both infect the same bacterial cell
of another. The ways in which bacteria exchange DNA are (phage recombination, not shown in Figure 6-2).
outlined in Figure 6-2. Before we analyze these modes of genetic exchange, let’s
The first process of gene exchange to be examined will consider the practical ways of handling bacteria, which are
be conjugation, which is the contact and fusion of two dif- much different from those used in handling multicellular
ferent bacterial cells. After fusion, one cell, called a donor, organisms.
sometimes transfers genomic DNA to the other cell. This
transferred DNA may be part or (rarely) all of the bacte-
rial genome. In some cases, one or more plasmids, if pres-
ent, are transferred. Some plasmids are capable of carrying 6.1 WORKING WITH
genomic DNA into the recipient cell. Any genomic frag- MICROORGANISMS
ment, transferred by whatever route, may recombine with
the recipient’s chromosome after entry. Bacteria are fast-dividing and take up little space; so they
A bacterial cell can also take up a piece of DNA from the are very convenient to use as genetic model organisms.
external environment and incorporate this DNA into its own They can be cultured in a liquid medium or on a solid
chromosome, a process called transformation. In addition, surface such as an agar gel, as long as basic nutrients
certain phages can pick up a piece of DNA from one bacterial are supplied. Each bacterial cell divides asexually from
1 → 2 → 4 → 8 → 16 cells, and so on, until the nutrients cannot. Figure 6-4 shows another way of distinguishing lac+
are exhausted or until toxic waste products accumulate to and lac− colonies by using a dye. In another mutant cate
levels that halt the population growth. A small amount of a gory, whereas wild types are susceptible to an inhibitor,
liquid culture can be pipetted onto a petri plate containing such as the antibiotic streptomycin, resistant mutants can
solid agar medium and spread evenly on the surface with divide and form colonies in the presence of the inhibitor.
a sterile spreader, in a process called plating (Figure 6-3). All these types of mutants allow the geneticist to distin-
The cells divide, but, because they cannot travel far on the guish different individual strains, thereby providing genetic
surface of the gel, all the cells remain together in a clump. markers (marker alleles) to keep track of genomes and cells
When this mass reaches more than 107 cells, it becomes in experiments. Table 6-1 summarizes some mutant bacterial
visible to the naked eye as a colony. Each distinct colony phenotypes and their genetic symbols.
on the plate has been derived from a single original cell. The following sections document the discovery of the
Members of a colony that have a single genetic ancestor are various processes by which bacterial genomes recombine.
known as cell clones. The historical methods are interesting in themselves but
Bacterial mutants are quite easy to obtain. Nutri- also serve to introduce the diverse processes of recombina-
tional mutants are a good example. Wild-type bacteria tion as well as analytical techniques that are still applicable
are p rototrophic, which means that they can grow and today.
divide on minimal medium—a substrate containing only
inorganic salts, a carbon source for energy, and water.
From a prototrophic culture, auxotrophic mutants can be
obtained: these mutants are cells that will not grow unless Distinguishing lac+ and lac− by using a red dye
the medium contains one or more specific cellular build-
ing blocks such as adenine, threonine, or biotin. Another
type of useful mutant differs from wild type in the ability
to use a specific energy source; for example, the wild type
(lac+ ) can use lactose and grow, whereas a mutant (lac− )
FIGURE 6-3 Bacterial phenotypes can be assessed in their gal− Cannot utilize galactose as a carbon source
colonies. A stock of bacterial cells can be grown in a liquid medium str r Resistant to the antibiotic streptomycin
containing nutrients, and then a small number of bacteria from the s
liquid suspension can be spread on solid agar medium. Each cell will str Sensitive to the antibiotic streptomycin
give rise to a colony. All cells in a colony have the same genotype and Note: Minimal medium is the basic synthetic medium for bacterial
phenotype. growth without nutrient supplements.
6.2 BACTERIAL CONJUGATION the elegantly simple experimental work of Joshua Lederberg
and Edward Tatum, who in 1946 discovered a sex-like pro-
LO 6.1 Distinguish between the three main ways by cess in what became the main model for bacterial genet-
which bacteria exchange genes, and describe the ics, Escherichia coli (see the Model Organism box on page
experimental procedures that reveal them. 198). They were studying two strains of E. coli with differ-
LO 6.2 Map bacterial genomes using interrupted conjugation. ent sets of auxotrophic mutations. Strain A− would grow
LO 6.3 Map bacterial genomes using recombinant frequency. only if the medium were supplemented with methionine
LO 6.4 In bacterial crosses, predict the inheritance of and biotin; strain B− would grow only if it were supple-
genetic elements carried on plasmids. mented with threonine, leucine, and thiamine. Thus, we can
designate the strains as
The earliest studies in bacterial genetics revealed the unex-
pected process of cell conjugation. strain A− : met − bio− thr + leu+ thi+
strain B− : met + bio+ thr− leu− thi−
Discovery of conjugation Figure 6-5a displays in simplified form the design of
Do bacteria possess any processes similar to sexual repro- their experiment. Strains A− and B− were mixed together,
duction and recombination? The question was answered by incubated for a while, and then plated on minimal medium,
(a)
A– + B
–
Mix
Some
WT progeny
(b)
A– B–
met – bio – thr + leu + thi + Mixture met + bio + thr – leu – thi –
Bacteria conjugate by using pili By accident, Hayes discovered a variant of his origi-
nal donor strain that would not produce recombinants on
crossing with the recipient strain. Apparently, the donor-
type strain had lost the ability to transfer genetic material
and had changed into a recipient-type strain. In work-
ing with this “sterile” donor variant, Hayes found that it
could regain the ability to act as a donor by association
with other donor strains. Indeed, the donor ability was
transmitted rapidly and effectively between strains during
conjugation. A kind of “infectious transfer” of some fac-
tor seemed to be taking place. He suggested that the abil-
ity to be a donor is itself a hereditary state, imposed by a
fertility factor (F). Strains that carry F can donate and are
designated F+. Strains that lack F cannot donate and are
recipients, designated F−.
We now know much more about F. It is a type of
small, nonessential circular DNA molecule called a
plasmid that can replicate in the cytoplasm independent
of the host chromosome. Figure 6-8 shows how bacteria
FIGURE 6-7 A donor cell extends one or more projections, or pili,
can transfer plasmids such as F. The F plasmid directs the
that attach to a recipient cell and pull the two bacteria together.
[Dr. L. Caro/Science Source.] synthesis of pili (sing., pilus), projections that initiate con-
tact with a recipient (see Figures 6-7 and 6-8) and draw
genuine recombinants had been produced. The physical it closer. The F plasmid DNA in the donor cell makes a
union of bacterial cells can be confirmed under an electron single-stranded version of itself in a peculiar mechanism
microscope and is now called conjugation (Figure 6-7). called rolling c ircle replication. The circular plasmid
“rolls,” and as it turns, it reels out a newly synthesized,
single-stranded “fishing line.” This single strand of DNA
Discovery of the fertility factor (F) passes through a pore into the recipient cell, where the
In 1953, William Hayes discovered that, in the types of other DNA strand is synthesized, forming a double helix.
“crosses” just described here, the conjugating parents Hence, a copy of F remains in the donor and another
acted unequally (later, we will see ways to demonstrate this appears in the recipient, as shown in Figure 6-8. Note that
unequal participation). One parent (and only that parent) the E. coli genome is depicted as a single circular chromo-
seemed to transfer some or all of its genome into another some in Figure 6-8. We will examine the evidence for it
cell. Hence, one cell acts as a donor, and the other cell acts later. Most bacterial genomes are circular, a feature quite
as a recipient. This “cross” is quite different from eukary- different from eukaryotic nuclear chromosomes. We will
otic crosses in which parents contribute nuclear genomes see that this feature leads to the many idiosyncrasies of
equally to a progeny individual. bacterial genetics.
KEY CONCEPT The transfer of genetic material in E. coli KEY CONCEPT F plasmids from F+ donor cells are transmit-
conjugation is not reciprocal. One cell, the donor, transfers part ted rapidly to F− recipient cells by rolling circle replication, but
of its genome to the other cell, which acts as the recipient. the bacterial chromosome is not transferred.
(a) (b)
Donor F +
Bacterial
chromosome
F+ F
Linear transmission of the Hfr genes from a fixed
point A clearer view of the behavior of Hfr strains was
obtained in 1957, when Elie Wollman and François Jacob
investigated the pattern of transmission of Hfr genes to F−
cells during a cross. They crossed
Hfr Hfr azi r ton r lac+ gal + str s × F− azi s tons lac− gal − str r
(Superscripts “r” and “s” stand for resistant and sensitive,
Integrated F
respectively.) At specific times after mixing, they removed
samples, which were each put in a kitchen blender for a few
FIGURE 6-9 In an F+ strain, the free F plasmid occasionally integrates seconds to separate the mating cell pairs. This procedure is
into the E. coli chromosome, creating an Hfr strain. called interrupted mating. The sample was then plated onto
Recombinant Lost
c+ b – a–
Double crossover
c– b+ a+ inserts
donor DNA
a medium containing streptomycin to kill the Hfr donor must have taken part in conjugation; such cells are called
cells, which bore the sensitivity allele strs. The surviving exconjugants. The results are plotted in Figure 6-12a, show-
strr cells then were tested for the presence of alleles from ing a time course of entry of each donor allele azir, tonr, lac+ ,
the donor Hfr genome. Any strr cell bearing a donor allele and gal +. Figure 6-12b portrays the transfer of Hfr alleles.
100
10 min
characters among str r exconjugants
azi r
Frequency (%) of Hfr genetic
80
ton r
40 lac + 17 min
20 gal +
0
0 10 20 30 40 50 60
Time (minutes)
25 min
KEY CONCEPT The Hfr chromosome, originally circular, 1. One end of the integrated F factor would be the origin,
unwinds a copy of itself that is transferred to the F cell in a − where transfer of the Hfr chromosome begins. The
linear fashion, with the F factor entering last. terminus would be at the other end of F.
2. The orientation in which F is inserted would determine
KEY CONCEPT Time of entry of Hfr alleles into an F recipient − the order of entry of donor alleles. If the circle contains
can be used to make a chromosome map. genes A, B, C, and D, then insertion between A and
D would give the order ABCD or DCBA, depending
on orientation. Check the different orientations of the
Inferring integration sites of F and chromosome
insertions in Figure 6-14.
circularity Wollman and Jacob went on to shed more light
on how and where the F plasmid integrates to form an Hfr How is it possible for F to integrate at different sites
cell and, in doing so, deduced that the chromosome is cir- and in different orientations? If F DNA had a region
cular. They performed interrupted-mating experiments with homologous to any of several regions on the bacterial chro-
different, separately derived Hfr strains. Significantly, the mosome, any one of them could act as a pairing region
order of transmission of the alleles differed from strain to at which pairing could be followed by a crossover. These
strain, as in the following examples: regions of homology are now known to be mainly segments
A single crossover inserts F at a specific locus, which then determines the order of gene transfer
O
Homologous
regions where
F pairing can
take place 2
1 2 1
a
a b
1 b c d a 2
Hfr
Transferred last Direction of transfer Transferred first
d c
E. coli chromosome d c
FIGURE 6-13 The insertion of F creates an Hfr cell. Hypothetical markers 1 and 2 are shown on F to
depict the direction of insertion. The origin (O) is the mobilization point where insertion into the E. coli
chromosome occurs; the pairing region is homologous with a region on the E. coli chromosome; a through
d are representative genes in the E. coli chromosome. Pairing regions (hatched) are identical in plasmid
and chromosome. They are derived from mobile elements called insertion sequences (see Chapter 16). In
this example, the Hfr cell created by the insertion of F would transfer its genes in the order a, d, c, b.
thi gly his gal pur lac pro thr lac pur gal his gly thi thr pro pro lac pur gal his gly thi thr
H 2 1
thr thr
thi pro thi pro
FIGURE 6-14 The five E. coli Hfr strains shown each have different F plasmid insertion points and
orientations. All strains have the same order of genes on the E. coli chromosome. The orientation of the F
factor determines which gene enters the recipient cell first. The gene closest to the terminus enters last.
F a+
F+ a +
Conjugation and
Insertion of F factor transfer of F factor
F a+ F+ a + F a+
Hfr a +
F– a – a–
Conjugation and
chromosome transfer
F a+ Hfr a + F a+
a+ F+ a +
a– F– a –
F – a +/ a – F a–
Recombination No recombination
F– a + F– a – F+ a –
based on crossovers, these linkage maps were originally A single crossover cannot produce a viable recombinant
purely genetic constructions. At the time they were originally
a+
devised, there was no way of testing their physical basis.
a– a+ a–
Fine-scale chromosome mapping by using recombi- Nonviable
nant frequency For an exconjugant to acquire donor genes
as a permanent feature of its genome, the donor fragment FIGURE 6-16 A single crossover between exogenote and
must recombine with the recipient chromosome. However, endogenote in a merozygote would lead to a linear, partly diploid
note that time-of-entry mapping is not based on recombinant chromosome that would not survive.
frequency. Indeed, the units are minutes, not RF. Neverthe-
less, recombinant frequency can be used for a more fine-scale an internal section of the exogenote. The other product of
type of mapping in bacteria, a method to which we now turn. the “double crossover,” the fragment, is generally lost in sub-
First, we need to understand some special features of sequent cell growth. Hence, only one of the reciprocal prod-
the recombination event in bacteria. Recall that recombi- ucts of recombination survives. Therefore, another unique
nation does not take place between two whole genomes, feature of bacterial recombination is that we must forget
as it does in eukaryotes. In contrast, it takes place between about reciprocal exchange products in most cases.
one complete genome, from the F− recipient cell, called the
endogenote, and an incomplete one, derived from the Hfr KEY CONCEPT Recombination during conjugation results
donor cell and called the exogenote. The cell at this stage from a double-crossover-like event, which gives rise to
has two copies of one segment of DNA: one copy is part reciprocal recombinants of which only one survives.
of the endogenote and the other copy is part of the exoge-
note. Thus, at this stage, the cell is a partial diploid, called a
merozygote. Bacterial genetics is merozygote genetics. A sin- With this understanding, we can examine recombi-
gle crossover in a merozygote would break the ring and thus nation mapping. Suppose that we want to calculate map
not produce viable recombinants, as shown in Figure 6-16. distances separating three close loci: met, arg, and leu. To
To keep the circle intact, there must be an even number of examine the recombination of these genes, we need “tri-
crossovers. An even number of crossovers produces a cir- hybrids,” exconjugants that have received all three donor
cular, intact chromosome and a fragment. Although such markers. Assume that an interrupted-mating experiment
recombination events are represented in a shorthand way as has shown that the order is met, arg, leu, with met trans-
double crossovers, the actual molecular mechanism is some- ferred first and leu last. To obtain a trihybrid, we need the
what different, more like an invasion of the endogenote by merozygote diagrammed here:
The goal now is to count the frequencies of crossovers nants and then examine them for the arg + and met + alleles,
at different locations. Note that we now have a different called the unselected markers. Figure 6-17 depicts the types
Hfr fragment
arg met
leu leu arg met
F chromosome
(b) Insertion of late marker and one early marker
arg met
leu
(d) Insertion of late and early markers, but not of marker in between
arg met
leu
FIGURE 6-17 The diagram shows how genes can be mapped by recombination in E. coli. In
exconjugants, selection is made for merozygotes bearing the leu+ marker, which is donated late. The
early markers (arg+ and met+ ) may or may not be inserted, depending on the site where recombination A N I M ATED A RT
between the Hfr fragment and the F− chromosome takes place. The frequencies of events diagrammed LO and mapping by
Bacterial conjugation
in (a) and (b) are used to obtain the relative sizes of the leu–arg and arg–met regions. Note that, in each recombination
case, only the DNA inserted into the F− chromosome survives; the other fragment is lost.
of double-crossover events expected. One crossover must exconjugants seemed to carry an F ′ plasmid with a piece of
be on the left side of the leu marker and the other must be the donor chromosome incorporated. The origin of this F ′
on the right side. Let’s assume that the leu+ exconjugants plasmid is shown in Figure 6-18. Note that the faulty exci-
are of the following types and frequencies: sion occurs because there is another homologous region
nearby that pairs with the original. The F ′ in our example
leu+ arg− met − 4%
is called F ′ lac because the piece of host chromosome that it
leu+ arg + met − 9% picked up has the lac gene on it. F ′ factors have been found
leu+ arg + met + 87% carrying many different chromosomal genes and have
been named accordingly. For example, F ′ factors carrying
The double crossovers needed to produce these genotypes
gal or trp are called F ′ gal and F ′ trp, respectively. Because
are shown in Figure 6-17a, b, and c. The first two types
F ′ lac+ /F− lac− cells are lac+ in phenotype, we know that
are the key because they require a crossover between leu
lac+ is dominant over lac−.
and arg in the first case and between arg and met in the
Partial diploids made with the use of F ′ strains are use-
second. Hence, the relative frequencies of these types corre-
ful for some aspects of routine bacterial genetics, such as
spond to the sizes of these two regions between the genes.
the study of dominance or of allele interaction. Some F ′
We would conclude that the leu –arg region is 4 m.u. and
strains can carry very large parts (as much as one-quarter)
that the arg –met region is 9 m.u.
of the bacterial chromosome.
In a cross such as the one just described, one type of
potential recombinants of genotype leu+ arg− met + requires KEY CONCEPT The DNA of an F′ plasmid is part F factor and
four crossovers instead of two (see Figure 6-17d). These part bacterial genome. Like F plasmids, F′ plasmids transfer
recombinants are rarely recovered because their frequency rapidly. They can be used to establish partial diploids for stud-
is very low compared with that of the other types of ies of bacterial dominance and allele interaction.
recombinants.
(a) Insertion F
IS1 lac
ton IS2
Integrated F factor
(b)
lac
Hfr chromosome
lac
(c) Excision
F' lac
FIGURE 6-18 An F plasmid can
lac
(d) pick up chromosomal DNA as it exits
a chromosome. (a) F is inserted in
an Hfr strain at a repetitive element
identified as IS1 (insertion sequence
1) between the ton and lac+ alleles.
(b) The integrated F factor. (c) Abnormal
“outlooping” by crossing over with a
different element, IS2, to include the
(e) F' lac /lac
lac lac locus. (d) The resulting F′ lac+
partial diploid
plasmid. (e) F′ lac+ /F− lac− partial diploid
produced by the transfer of the F′ lac+
lac plasmid to an F− lac− recipient. [Data
from G. S. Stent and R. Calendar, Molecular
Genetics, 2nd ed.]
characteristics that can be borne by plasmids. Figure 6-19 used as markers to keep track of the movement of the vec-
shows an example of a well-traveled plasmid isolated tors between cells.
from the dairy industry. On R plasmids, the alleles for antibiotic resistance
Engineered derivatives of R plasmids, such as pBR322 are often contained within a unit called a transposon
and pUC (see Chapter 10), have become the preferred vec- (Figure 6-20). Transposons are unique segments of DNA
tors for the molecular cloning of the DNA of all organisms. that can move around to different sites in the genome, a
The genes on an R plasmid that confer resistance can be process called transposition. (The mechanisms for transpo-
sition, which occurs in most species studied, will be detailed
in Chapter 16.) When a transposon in the genome moves
to a new location, it can occasionally embrace between its
TABLE 6-2 Genetic Determinants Borne by Plasmids
ends various types of genes, including alleles for drug resis-
Characteristic Plasmid examples
tance, and carry them along to their new locations as pas-
Fertility F, R1, Col sengers. Sometimes, a transposon carries a drug-resistance
Bacteriocin production Col E1 allele to a plasmid, creating an R plasmid. Like F plasmids,
Heavy-metal resistance R6 many R plasmids are conjugative; in other words, they are
Enterotoxin production Ent effectively transmitted to a recipient cell during conjuga-
tion. Even R plasmids that are not conjugative and never
Metabolism of camphor Cam
leave their own cells can donate their R alleles to a conju-
Tumorigenicity in plants T1 (in Agrobacterium tumefaciens) gative plasmid by transposition. Hence, through plasmids,
FIGURE 6-19 The diagram A plasmid with segments from many former bacterial hosts
shows the origins of genes of
the Lactococcus lactis plasmid Enterococcus Lactococcus
faecium lactis Mycoplasma
pK214. The genes are from
many different bacteria. [Data Listeria
from Table 1 in V. Perreten, monocytogenes
29 1 2
F. Schwarz, L. Cresta, M. Boeglin, 3
4
G. Dasen, and M. Teuber, Nature 5
28
389, 1997, 801–802.] 6
Lactococcus
Enterococcus 27
lactis
faecalis
7
26
25
8
Enterococcus
faecium 24
Listeria
monocytogenes
23 Plasmid pk214
Staphylococcus 22
9 Streptococcus
aureus
agalactiae
21
Lactobacillus
plantarum
10
Lactococcus
lactis
20 11
12
19 Streptococcus
13
14 pyogenes
Staphylococcus 18
aureus 17 15
16 Escherichia
coli
Enterococcus
faecium Lactococcus
Escherichia Escherichia lactis
coli coli
Staphylococcus
aureus
FIGURE 6-20 A transposon such as Tn5 can acquire several The nature of transformation
drug-resistance genes (in this case, those for resistance to the drugs
kanamycin and neomycin) and transmit them rapidly on a plasmid,
Transformation was discovered in the bacterium Strepto
leading to the infectious transfer of resistance genes as a package. coccus pneumoniae in 1928 by Frederick Griffith. Later, in
Insertion sequence 50 (IS50) forms the flanks of Tn5. 1944, Oswald T. Avery, Colin M. MacLeod, and Maclyn
Mechanism of DNA uptake by bacteria be carried on the same piece of transforming DNA. Hence,
both will be taken up, causing a double transformation.
Free DNA
DNA-binding Conversely, if genes are widely separated on the chromo-
complex
some, they will most likely be carried on separate trans-
Nucleotide forming segments. A genome could possibly take up both
Cell wall segments independently, creating a double transformant,
but that outcome is not likely. Hence, in widely separated
Cytoplasmic genes, the frequency of double transformants will equal the
membrane product of the single-transformant frequencies. Therefore,
testing for close linkage by testing for a departure from the
DNA-degrading product rule should be possible. In other words, if genes are
Free DNA enzyme
from dead linked, then the proportion of double transformants will be
bacterium
Transformed Transferred
greater than the product of single-transformant frequencies.
bacterium DNA Unfortunately, the situation is made more complex
Chromosome by several factors—the most important being that not all
cells in a population of bacteria are competent to be trans-
formed. Nevertheless, at the end of this chapter, you can
sharpen your skills in transformation analysis in one of the
(a) (b) problems, which assumes that 100 percent of the recipient
cells are competent.
FIGURE 6-21 A bacterium undergoing transformation (a) picks
up free DNA released from a dead bacterial cell. As DNA-binding
KEY CONCEPT Bacteria can take up DNA fragments from
complexes on the bacterial surface take up the DNA (inset), enzymes
the surrounding medium. Inside the cell, these fragments can
break down one strand into nucleotides; a derivative of the other
strand may integrate into the bacterium’s chromosome (b). integrate into the chromosome.
types are identified not by species names but by symbols— Electron micrograph of phage T4
for example, phage T4, phage λ, and so forth. Figures 6-22
and 6-23 show the structure of phage T4. During infection,
a phage attaches to a bacterium and injects its genetic
material into the bacterial cytoplasm, as diagrammed in
Figure 6-22. An electron micrograph of the process is
shown in Figure 6-24. The phage genetic information then
takes over the machinery of the bacterial cell by turning
off the synthesis of bacterial components and redirecting
the bacterial synthetic machinery to make phage compo-
nents. Newly made phage heads are individually stuffed
with replicates of the phage chromosome. Ultimately,
many phage descendants are made and are released when
the bacterial cell wall breaks open. This breaking-open
process is called lysis. The population of phage progeny is
called the phage lysate.
How can we study inheritance in phages when they
are so small that they are visible only under the electron
microscope? In this case, we cannot produce a visible
colony by plating, but we can produce a visible manifes-
tation of a phage by taking advantage of several phage
FIGURE 6-23 Enlargement of the E. coli phage T4 reveals details of
characters.
head, tail, and tail fibers. [Omikron/Science Source.]
Let’s look at the consequences of a phage infecting a sin-
gle bacterial cell. Figure 6-25 shows the sequence of events
in the infectious cycle that leads to the release of progeny
phages from the lysed cell. After lysis, the progeny phages infect neighboring bacteria. This cycle is repeated through
progressive rounds of infection, and, as these cycles repeat,
the number of lysed cells increases exponentially. Within
15 hours after one single phage particle infects a single
Structure and function of phage T4 bacterial cell, the effects are visible to the naked eye as a
Free phage T4 phage components clear area, or plaque, in the opaque lawn of bacteria cov-
ering the surface of a plate of solid medium (Figure 6-26).
DNA Such plaques can be large or small, fuzzy or sharp, and
so forth, depending on the phage genotype. Thus, plaque
Head morphology is a phage character that can be analyzed at
the genetic level. Another phage phenotype that we can ana-
Neck and collar lyze genetically is host range, because phages may differ in
the spectra of bacterial strains that they can infect and lyse.
Sheath
End plate
Cell wall
Injected
DNA
Fibers
Cycle of phage that lyses the host cells A plaque is a clear area in which all
bacteria have been lysed by phages
Clear areas, or plaques
Uninfected
cell
Lysis of Adsorption
host cell of phage FIGURE 6-26 Through repeated infection and production of progeny
Free to host cell
phages phage, a single phage produces a clear area, or plaque, on the opaque
lawn of bacterial cells. [D. Sue Katz, Rogers State University,
Claremore, OK.]
Plaques from recombinant and can take place between genetically similar phages as well as
parental phage progeny between different types. Thus, if we let P1 and P2 refer to
general parental genotypes, crosses of P1 × P1 and P2 × P2 take
place in addition to P1 × P2. For both these reasons, recombi-
nants from phage crosses are a consequence of a population
of events rather than defined, single-step exchange events.
Nevertheless, all other things being equal, the RF calculation
does represent a valid index of map distance in phages.
Because astronomically large numbers of phages can be
used in phage-recombination analyses, very rare crossover
events can be detected. In the 1950s, Seymour Benzer made
use of such rare crossover events to map the mutant sites
within the rII gene of phage T4, a gene that controls lysis.
For different rII mutant alleles arising spontaneously, the
mutant site is usually at different positions within the gene.
Therefore, when two different rII mutants are crossed, a
few rare crossovers may take place between the mutant
sites, producing wild-type recombinants, as shown here:
rII gene
Parent 1
Parent 2
a+
a+ b+
b+
a+
Donor bacterium b+
b+
Phages carrying
donor genes
a+
a+
a+ a– a+
a–
FIGURE 6-29 A newly forming phage may pick up DNA from its host cell’s chromosome (top) and
then inject it into a new cell (bottom right). The injected DNA may insert into the new host’s chromosome
by recombination (bottom left). In reality, only a very small minority of phage progeny (1 in 10,000) carry
donor genes.
21, 16
2.0
F– F–
gal gal
FIGURE 6-31 A λ prophage can be transferred to a recipient during conjugation, but the prophage
triggers lysis, a process called zygotic induction, only if the recipient has no prophage already—that is, in
the case shown in part a but not in part b.
Experiment 1 in Table 6-3 tells us that leu is relatively almost never recovered from this reciprocal cross. What is the
close to azi and distant from thr, leaving us with two explanation? The observations make sense if the λ prophage
possibilities: is behaving as a bacterial gene locus behaves (that is, as part
thr leu azi of the bacterial chromosome). Thus, in the Hfr(λ) × F− cross,
the prophage would enter the F− cell at a specific time cor-
or responding to its position in the chromosome. Earlier genes
thr azi leu are recovered because they enter before the prophage. Later
genes are not recovered because lysis destroys the recipi-
Experiment 2 tells us that leu is closer to thr than azi is, and ent cell. In interrupted-mating experiments, the λ prophage
so the map must be does in fact always enter the F− cell at a specific time, closely
thr leu azi linked to the gal locus.
In an Hfr(λ) × F− cross, the entry of the λ prophage
By selecting for thr + and leu+ together in the transduc into the cell immediately triggers the prophage into a lytic
ing phages in experiment 3, we see that the transduced cycle; this process is called zygotic induction (Figure 6-31).
piece of genetic material never includes the azi locus However, in the cross of two lysogenic cells Hfr(λ) × F− (λ),
because the phage head cannot carry a fragment of DNA there is no zygotic induction. The presence of any prophage
that big. P1 can only cotransduce genes less than approxi- prevents another infecting virus from causing lysis. This
mately 1.5 minutes apart on the E. coli chromosome map. is because the prophage produces a cytoplasmic factor
that represses the multiplication of the virus. (The phage-
directed cytoplasmic repressor nicely explains the immunity
Specialized transduction of the lysogenic bacteria, because a phage would immedi-
A generalized transducer, such as phage P22, picks up frag- ately encounter a repressor and be inactivated.)
ments of broken host DNA at random. How are other phages,
which act as specialized transducers, able to carry only cer- λ insertion The interrupted-mating experiments hereto-
tain host genes to recipient cells? The short answer is that a fore described showed that the λ prophage is part of the
specialized transducer inserts into the bacterial chromosome lysogenic bacterium’s chromosome. How is the λ prophage
at one position only. When it exits, a faulty outlooping occurs inserted into the bacterial genome? In 1962, Allan C ampbell
(similar to the type that produces F ′ plasmids). Hence, it can proposed that it inserts by a single crossover between a cir-
pick up and transduce only genes that are close by. cular λ phage chromosome and the circular E. coli chromo-
The prototype of specialized transduction was provided some, as shown in Figure 6-32. The crossover point would
by studies undertaken by Joshua and Esther Lederberg on a be between a specific site in λ, the λ attachment site, and
temperate E. coli phage called lambda (λ). Phage λ has become an attachment site in the bacterial chromosome located
the most intensively studied and best-characterized phage. between the genes gal and bio, because λ integrates at that
position in the E. coli chromosome.
Behavior of the prophage Phage λ has unusual effects An attraction of Campbell’s proposal is that from it fol-
when cells lysogenic for it are used in crosses. In the cross of low predictions that geneticists can test. For example, inte-
an uninfected Hfr with a lysogenic F− recipient [Hfr × F− (λ)], gration of the prophage into the E. coli chromosome should
lysogenic F− exconjugants with Hfr genes are readily recov- increase the genetic distance between flanking bacterial
ered, as expected. However, in the reciprocal cross Hfr(λ) × F−, genes, as can be seen in Figure 6-32 for gal and bio. In fact,
the early genes from the Hfr chromosome are recovered studies show that lysogeny does increase time-of-entry or
among the exconjugants, but recombinants for late genes are recombination distances between the bacterial genes. This
not recovered. Furthermore, lysogenic F− exconjugants are unique location of λ accounts for its specialized transduction.
2 2
3 1 3 1
dgal helper
1 2 gal 1 2 3
gal – bio
(i) Lysogenic transductants
1
2 gal
a bacterial gene, gal or bio. The abnormal DNA carrying 6.6 PHYSICAL MAPS AND
nearby genes can be packaged into phage heads to produce
phage particles that can infect other bacteria. These phages
LINKAGE MAPS COMPARED
are referred to as λdgal (λ-defective gal ) or λdbio. In the
LO 6.8 Explain how transposons can be used in insertional
presence of a second, normal phage particle in a double
mutagenesis to create and map new mutations.
infection, the λdgal can integrate into the chromosome at
the λ attachment site (Figure 6-33c). In this manner, the gal Some very detailed chromosomal maps for bacteria have
genes in this case are transduced into the second host. been obtained by combining the mapping techniques of
KEY CONCEPT Transduction occurs when newly forming
interrupted mating, recombination mapping, transforma-
phages acquire host genes and transfer them to other bacterial tion, and transduction. Today, new genetic markers are
cells. Generalized transduction can transfer any host gene. typically mapped first into a segment of about 10 to 15
It occurs when phage packaging accidentally incorporates map minutes by using interrupted mating. Then additional,
bacterial DNA instead of phage DNA. Specialized transduction closely linked markers can be mapped in a more fine-scale
is due to faulty outlooping of the prophage from the bacterial analysis with the use of P1 cotransduction or recombination.
chromosome, and so the new phage includes both phage By 1963, the E. coli map (Figure 6-34) already detailed
and bacterial genes. The transducing phage can transfer only the positions of approximately 100 genes. After 27 years
specific host genes. of further refinement, the 1990 map depicted the positions
D
A
B
I
C
thrA,D
pdxA
uvrA
malB
pyrA
(lex)
metA
*purH
(ace upM
ara
aceE
*purthl
pgl
A
leu
aceF
azi
Y
*s
Z
A,D
ftsA
D
P
pan
O
*cy H
arg B
80 0
arg C
1
ar gE
79
lac
pr cl
g
ph oC
ar c
lo hoR *
2
la
p oA
B
*serB
(mutT)
*thyR
*
pp rts
(trpR)
(tp p)
A 78 10
(gua pil
*valS
tonA
* d)
m
hsp
(ra
(ast)
D m etF *
argF
n
e in ,B
pyr
C)
glp tB m
A
pE
pro
77 11 u
B
fdp
rh K s s*
*p
P rn pG*
tfr oB
a
urA A
*a
B 11.5 su A*
pr
m
glt
p
C 76 0 / 90 15
ilv c
O 85 5 x su L*
A rbs 74 ts l p
su ,B*
D p h b
m E
o
(da S r tolA
E rA) 80 10 pu G K
me 16 aro T
*tna tE
R nicA E
*tna
A gal ) O
73
B
(chlB
75 15 (mglR
C bgl
) (phr)
A pyrE 17
xyl gltH* 2)
(att434.8
(gad) glyS aroA* att l
72
*gltC 70 20 pyrD chlA
bioA
mtl pyrC 18
urvB*
purB
(cat) 24 supC,O
71
65 25 supF*
aroE tdk*
A (chlC
67 *spc ) gaIU
argR 25 att f
as d A 80
*lin 60 30 pab tonB
yA B A
glpD *er sp aro trp
R 66 a H* B
g lp G aro C
arg np pp D 26 c ysB
lA p ) 55 35 D
ma m s* pyr
laS F E
B (a fda ph an* 36
aro ) * tC
e 50 40 eS
O
o B
(bi m
se gP
45
(ft gS
65 mo
rA
r
sB
*a
G
ar d)
y s m tB
c rel
*c 37
(e f)
ot
*m ysC
A
*
ab ) A
d
(zw
(da utS
(da pA)
*p
(so
D 64 uv
*pr )
pB
rg
(mg
*rec d
rC
A
(a
(tolC
m)
glpT
rA
56
purC
38
purF
*ctr
*nicB
supN
aroC
dsdA
dsdC
lP)
st
ha
)
pS
g
su
(tr
55 39
pH
(su
48
su E
ra
sh
(re T
pD
54 50 49
p
*a
his
)
iA
cB
)
gn
R
lysA
gal
d
thyA
argA
gua
fuc
purG*
glyA
tyrA
aroF*
pheA
uraP*
B
A
O
FIGURE 6-34 The 1963 genetic map of E. coli genes with mutant phenotypes. Units are minutes,
based on interrupted-mating and recombination experiments. Asterisks refer to map positions that are
not as precise as the other positions. [Data from G. S. Stent, Molecular Biology of Bacterial Viruses.]
mraA,B*
dapC*
cdsA*
gInD*
optA*
dapD
sefA*
polC
hlpA
IpxA
IpxB
rpsB
murG
murC
murE
firA
murF
erivA
mutT
secA
orf
orf
orf
polB
frsQ
tsf
ftsA
ftsZ
ddl
ftsl
orf
rpsT
lspA
ileS
ABC KJ AB
panBCD
acrC*
proS*
pyrH*
tadE*
popC*
sefA*
metD
mafB*
garB*
serR*
guaC
nadC
mrcB
pcnB
aceE
prlD*
dadB
fruR*
aceF
ftsM*
brnS*
ssyD
chlG*
aroP
(rimG) rimF*
apaH
gprA*
dapB
mafA
pdxA
ksgA
kefC
(popD) gprB
spe
Irs*
hpt
fhu
Ipd
folA
ant*
ara
leu
dna
orf
ilvJ
tolJ
car
ilv
orf
thr
(envN )
(tdi)
(toll)
0 1 2 3 4 5
FIGURE 6-35 A linear scale drawing of a sequenced 5-minute section of the 100-minute 1990
E. coli linkage map. The parentheses and asterisks indicate markers for which the exact location was
unknown at the time of publication. Arrows above genes and groups of genes indicate the direction of
transcription. [Data from B. J. Bachmann, “Linkage Map of Escherichia coli K-12, Edition 8,” Microbiol. Rev.
54, 1990, 130–197.]
of more than 1400 genes. Figure 6-35 shows a 5-minute process under study, and these phenotypes become the
section of the 1990 map (which is adjusted to a scale of focus of the analysis.
100 minutes). The complexity of these maps illustrates the
power and sophistication of genetic analysis. How well do Physical map of the E. coli genome
these maps correspond to physical reality? In 1997, the
DNA sequence of the entire E. coli genome of 4,632,221
base pairs was completed, allowing us to compare the exact Re
p lic
position of genes on the genetic map with the position of ho
re
1
the corresponding coding sequence on the linear DNA Or
igi
sequence (the physical map). The full map is represented in n
Figure 6-36. Figure 6-37 makes a comparison for a segment
of both maps. Clearly, the genetic map is a close match to
the physical map.
E. coli
KEY CONCEPT Generally, genetic maps of bacterial chromo-
somes show their genes in the same order and relative position
as physical bacterial maps derived from DNA sequencing.
Te
rm
Chapter 4 considered some ways in which the physi- i nu
Re s
cal map (usually the full genome sequence) can be useful p lic
in mapping new mutations. In bacteria, the technique of ho
re
insertional mutagenesis is another way to zero in rapidly 2
on a mutation’s position on a known physical map. The
FIGURE 6-36 This map was obtained from sequencing DNA and
technique causes mutations through the random insertion
plotting gene positions. Key to components from the outside in:
of “foreign” DNA fragments. The inserts inactivate any
• The DNA replication origin and terminus are marked.
gene in which they land by interrupting the transcriptional • The two scales are in DNA base pairs and in minutes.
unit. Transposons are particularly useful inserts for this • The orange and yellow histograms show the distribution of genes on
purpose in several model organisms, including bacteria. To the two different DNA strands.
map a new mutation, the procedure is as follows. The DNA • The arrows represent genes for rRNA (red) and tRNA (green).
of a transposon carrying a resistance allele or other select- • The central “starburst” is a histogram of each gene with lines of
length that reflect predicted level of transcription.
able marker is introduced by transformation into bacterial
[Republished with permission of the American Association for the
recipients that have no active transposons. The transposons
Advancement of Science, from F. R. Blattner et al., “The Complete Genome
insert more or less randomly, and any that land in the mid- Sequence of Escherichia coli K-12,” Science 277, 1997, 1453–1462,
dle of a gene cause a mutation. A subset of all mutants Figure 1. DOI: 10.1126/science.277.5331. Permission conveyed through
obtained will have phenotypes relevant to the bacterial Copyright Clearance Center, Inc. Image courtesy of Dr. Guy Plunkett III.]
Proportions of the genetic and physical maps are similar but not identical
(a)
cysC cysH eno relA argA recC mutH
60 ptr thyA 61
(b)
mutS rpoS pcm cysC iap cysH eno relA barA syd sdaC exo gcvA mltA argA ptr recC thyA ptsP mutH aas galR araE glyU
FIGURE 6-37 An alignment of the genetic and physical maps. (a) Markers on the 1990 genetic map
in the region near 60 and 61 minutes. (b) The exact positions of every gene, based on the complete
sequence of the E. coli genome. (Not every gene is named in this map, for simplicity.) The elongated
boxes are genes and putative genes. Each color represents a different type of function. For example, red
denotes regulatory functions, and dark blue denotes functions in DNA replication, recombination, and
repair. Lines between the maps in parts a and b connect the same gene in each map. [Data from F. R.
Blattner et al., The Complete Science 277, 1997, 1453–1462.]
The beauty of inserting transposons is that, because Transposon mutagenesis can be used to map a
their sequence is known, the mutant gene can be located mutation in the genome sequence
and sequenced. DNA replication primers are created
that match the known sequence of the transposon (see Wild-type
Chapter 10). These primers are used to initiate a sequenc- cell
ing analysis that proceeds outward from the transposon
into the surrounding gene. The short sequence obtained can
then be fed into a computer and compared with the com- Transposon
plete genome sequence. From this analysis, the position of
the gene and its full sequence are obtained. The function of
Mutant
a homolog of this gene might already have been deduced phenotype
in other organisms. Hence, you can see that this approach induced by
(like that introduced in Chapter 4) is another way of unit- transposon
insertion
ing mutant phenotype with map position and potential
function. Figure 6-38 summarizes the approach.
As an aside in closing, it is interesting that many of the
historical experiments revealing the circularity of bacterial
and plasmid genomes coincided with the publication and
popularization of J. R. R. Tolkien’s The Lord of the Rings.
Primed synthesis
Consequently, a review of bacterial genetics at that time led
off with the following quotation from the trilogy:
One Ring to rule them all, One Ring to find them,
One Ring to bring them all and in the darkness
bind them.
Indeed, plasmid and bacterial chromosome rings
Whole gene identified
turned out to be even more powerful than originally from genome sequence
realized. Discoveries made in bacteria were critical in
furthering the research of higher organisms and in the FIGURE 6-38 The insertion of a transposon inserts a mutation into
a gene of unknown position and function. The segment next to the
development of sophisticated genetic engineering tech-
transposon is replicated, sequenced, and matched to a segment in the
niques. Box 6-1 discusses some of the bacterial processes complete genome sequence.
that paved the way for these advancements.
BOX 6-1 Bacterial and Phage Genetics Harnessed for Manipulation of Eukaryotic DNA
The material in this box is a bridge between this chapter Most restriction enzymes produce what is referred
on bacterial genetics and later chapters on manipulation to as a staggered cut, such as the following hypothetical
of eukaryotic DNA. It illustrates a historical principle that example where each arrow represents a cut in a single
seemingly esoteric research can sometimes pave the way strand of DNA:
to unimagined important applications. Although research
on bacterial and phage genetics revealed systems that are
biologically fascinating in their own right, the basic prin-
ciples and processes revealed by this research produced
unexpected spinoffs that have revolutionized modern
produces
genetic analysis. Some of the genetic elements and pro-
cesses discovered in these simpler life forms have pro-
vided powerful approaches to the elucidation and genetic and
manipulation of the more complex eukaryotic genomes,
including the genome of our own species. Hence, the These single-stranded tails with their hydrogen-bonding
reader should treat this section as a relaxing appetizer sites exposed are called “sticky ends.” This stickiness is
for what is to come; that is, all the techniques mentioned important, as we shall see later.
here will be expanded into full analytical coverage in later However, the key element for now is that restriction
chapters (beginning with Chapter 10), which are where enzymes can be extracted from bacteria, purified, and
the key concepts are delineated in full. The material here used to cut eukaryotic DNA because their target sites will
will provide a useful springboard for jumping into the later also undoubtedly be present by chance in that DNA. Fur-
treatments. thermore, because the target sites are part of the genomic
Eukaryotic genomes are large, composed of tens of sequence for that particular eukaryote, the cuts will all be
thousands of genes and billions of nucleotide pairs of in the same positions in any homologous DNA molecule.
DNA. Hence, a direct frontal attack on the study of such Hence, starting with a sample of extracted eukaryotic
genomes is difficult to impossible. Consequently, the DNA (which in a test tube resembles an unappetizing blob
general approach devised for this type of genomic anal- of mucus), the bacterial restriction enzymes cut it into
ysis was to cut up the eukaryotic genome into defined defined segments, which can be sorted, identified, and
fragments and then, once characterized, the parts can be manipulated in various ways. This represents the begin-
reassembled into a complete genome or can be used indi- ning stage of eukaryotic genomics.
vidually for more specific types of intervention. Hence, the
first step is to cut the eukaryotic genome into defined frag- Bacterial vectors and hosts for
ments, and bacteria provided an excellent way of doing eukaryotic DNA
this, as shown in the next paragraph.
For detailed study, individual segments of restriction-
digested eukaryotic DNA need to be amplified to convert
Restriction enzymes in bacteria them into pure samples that are effectively chemical
Bacteria are constantly under threat from parasitic ele- reagents. Here, the sticky tails left by the restriction enzymes
ments, especially viruses. Therefore, they have evolved are particularly useful because if the DNA of a bacterial plas-
several types of defense systems, and one of these is mid or a phage is cut with the same restriction enzyme and
restriction modification. It was found in the 1950s that the two DNAs mixed, they join by virtue of complementary
phages produced in a lysate are often defective as a result hydrogen bonding of their identical sticky ends, and hence,
of their DNA having been cut by the previous host bacte- we get replicating bacterial molecules with defined eukary-
rium. This led to the discovery that bacteria have genes otic inserts. These hybrid molecules, which are called
encoding DNA-cutting endonucleases called restriction recombinant DNAs, can be inserted into a bacterial host
enzymes. Restriction enzymes find and cut a specific tar- where they will replicate and produce a large pure sample of
get sequence in the viral DNA; often the target sequence is recombinant DNA (including the eukaryotic insert) for subse-
around 6 to 10 base pairs long. These target sequences are quent study. The replicating molecules in such as a study are
not necessarily in functional regions, but nevertheless, the called vectors (carriers). In this way, eukaryotic DNA can be
likelihood of a genome containing such target sequences prepared in easily characterizable units called DNA clones.
by chance is high. (The target sequences in the bacterial The clones can be used in a number of ways. For exam-
genome are protected by the addition of methyl groups.) ple, the insert DNAs can be sequenced and assembled to
obtain a full sequence of the eukaryotic genome. Another The components of the bacterial CRISPR systems
approach is to use the clones to modify eukaryotic cells. have been harnessed to design very effective genetic
For example, a wild-type recombinant DNA insert can be engineering methods for eukaryotes. Remarkably, the
used to “correct” or reverse a mutation in a eukaryotic CRISPR system works in eukaryotic cells. One com-
recipient; it happens that many eukaryotic cells can be monly used system uses Cas9, derived from a species
transformed in a way similar to bacteria, so the entry of the of Streptococcus. DNA constructs bearing Cas9 plus a
corrective fragment is facilitated. Once inside their eukary- “guide RNA” homologous to a target gene to be modi-
otic host, transforming fragments often go through homol- fied are inserted into the eukaryotic cell. The guide RNA
ogous double recombination with host DNA, thus replacing (which is substituting for the phage immunity fragment)
the resident sequence. finds the target gene by base homology, and Cas9 cuts
DNA clones can also be used to tailor the DNA of the it, resulting in a double-strand break. At this point, the
eukaryotic recipient in a highly specific way, and an exam- eukaryotic cellular repair mechanisms take over and
ple of this is shown in the following section. mend the break. Such systems are inherently error-prone,
and the repair can sometimes result in a faulty sequence
Bacterial CRISPR systems for leading to a random mutation. A random mutation in a
engineering eukaryotic DNA specific target gene might in itself be very useful; how-
Many species of bacteria and archaea have an immune ever, the versatility of the gene modification process can
system that (in contrast to the general defensive action of be greatly expanded by adding a tailored piece of DNA
restriction enzymes) protects them against specific infec- flanked by sequences homologous to the target gene. In
tious viruses and plasmids. The basis of the system is loci such a situation, the repair system is tricked into inserting
composed of Clustered Regularly Interspersed Short Palin- the tailored sequence into the target gene at the location
dromic Repeats. These are generally referred to as CRIS- of the double-strand break. Hence, the CRISPR-based
PRs for short (and pronounced “crispers”). In language, technology has efficiently inserted an extra piece of spe-
palindromes are words that have the same spelling whether cially tailored DNA into the eukaryotic gene of interest.
read forward or backward; for example, RADAR. In DNA, a This provides immense scope for genetic modification of
palindromic sequence is one where the 5′-to-3′ sequence the eukaryotic genome, for example in correcting for del-
of one strand is identical to the 5′-to-3′ sequence of the eterious mutations or adding new functions.
complementary strand, such as The molecular details of all these types of techniques
and their applications to genetic engineering will be cov-
5′ AAGGCCTT 3′ ered in Chapter 10.
3′ TTCCGGAA 5′
At CRISPR loci, the short palindromic repeats are sep- How bacterial CRISPR loci act as agents of immunity
arated by several different unique sequences that initially
seemed mysterious until it was discovered that they were in
fact non-genomic sequences derived from various phages CRISPR
or plasmids. This observation led to the idea that they DNA
were part of an immune system against invasive DNAs. At CAS
one end of each repeated sequence, there is a sequence RNA
encoding one to several proteins called CRISPR-associated
proteins or Cas proteins, which are DNA-cutting nucleases.
The antiviral mechanism works in the following way. CAS
When viral DNA enters the cell, either it can either kill the cell Cut
or, in some cases, a fragment of its DNA can be inserted into
the CRISPR array, where it acts as a heritable immune pro-
tectant against that specific virus. If the cell is later infected
by this virus, the CRISPR array is transcribed and cut into
short specific RNAs, each of which are hooked up to Cas Phage
proteins. One of the RNAs will be homologous to a region of
the infecting virus DNA and will bind through base pairing. At
FIGURE 1 The black regions represent repetitive DNA and the col-
that point, the Cas protein cuts the viral DNA with a double- ored regions are fragments derived from parasitic elements such as
strand break, thus rendering it incapable of encoding for phage. The net outcome is that the Cas protein cuts the invading DNA
phage propagation. Figure 1 summarizes the process. recognized by the red RNA, rendering it inactive.
SUMMARY
Advances in bacterial and phage genetics within the past occur, DNA must be taken into a recipient cell, and recom-
50 years have provided the foundation for molecular biol- bination must then take place between a recipient chromo-
ogy and cloning (discussed in later chapters). Early in this some and the incorporated DNA.
period, gene transfer and recombination were found to take Bacteria can be infected by viruses called bacterio-
place between different strains of bacteria. In bacteria, how- phages. In one method of infection, the phage chromosome
ever, genetic material is passed in only one direction—for may enter the bacterial cell and, by using the bacterial
example, in Escherichia coli, from a donor cell (F+ or Hfr) metabolic machinery, produce progeny phages that burst
to a recipient cell (F− ). Donor ability is determined by the the host bacterium. The new phages can then infect other
presence in the cell of a fertility factor (F), a type of plasmid. cells. If two phages of different genotypes infect the same
On occasion, the F factor present in the free state in F+ cells host, recombination between their chromosomes can
can integrate into the E. coli chromosome and form an Hfr take place.
cell. When this occurs, a fragment of donor chromosome In another mode of infection, lysogeny, the injected
can transfer into a recipient cell and subsequently recom- phage lies dormant in the bacterial cell. In many cases, this
bine with the recipient chromosome. Because the F factor dormant phage (the prophage) incorporates into the host
can insert at different places on the host chromosome, early chromosome and replicates with it. Either spontaneously or
investigators were able to piece the transferred fragments under appropriate stimulation, the prophage can leave its
together to show that the E. coli chromosome is a single dormant state and lyse the bacterial host cell.
circle, or ring. Interruption of the transfer at different times A phage can carry bacterial genes from a donor to a
has provided geneticists with an unconventional method recipient. In generalized transduction, random host DNA is
(interrupted mating) for constructing a linkage map of incorporated alone into the phage head during lysis. In spe-
the single chromosome of E. coli and other similar bacte- cialized transduction, faulty excision of the prophage from
ria, in which the map unit is a unit of time (minutes). In an a unique chromosomal locus results in the inclusion of spe-
extension of this technique, the frequency of recombinants cific host genes as well as phage DNA in the phage head.
between markers known to have entered the recipient can Today, a physical map in the form of the complete
provide a finer-scale map distance. genome sequence is available for many bacterial species.
Several types of plasmids other than F can be found. With the use of this physical genome map, the map posi-
R plasmids carry antibiotic-resistance alleles, often within a tion of a mutation of interest can be precisely located.
mobile element called a transposon. Rapid plasmid spread First, appropriate mutations are produced by the insertion
causes population-wide resistance to medically important of transposons (insertional mutagenesis). Then, the DNA
drugs. Derivatives of such natural plasmids have become sequence surrounding the inserted transposon is obtained
important cloning vectors, useful for gene isolation and and matched to a sequence in the physical map. This tech-
study in all organisms. nique provides the locus, the sequence, and possibly the
Genetic traits can also be transferred from one bacte- function of the gene of interest.
rial cell to another in the form of pieces of DNA taken into As we will see in subsequent chapters, many of the
the cell from the extracellular environment. This process of bacterial genetic mechanisms discovered historically have
transformation in bacterial cells was the first demonstration proved to be the basis for powerful approaches to engineer-
that DNA is the genetic material. For transformation to ing DNA in general.
KEY TERMS
auxotroph (p. 196) exconjugant (p. 201) Hfr (high frequency of recombina-
bacteriophage (phage) (p. 194) exogenote (p. 204) tion) (p. 200)
cell clone (p. 196) F+ (donor) (p. 199) horizontal transmission (p. 195)
colony (p. 196) F− (recipient) (p. 199) insertional mutagenesis (p. 218)
conjugation (p. 199) F ′ plasmid (p. 206) interrupted mating (p. 200)
cotransductant (p. 213) fertility factor (F) λ attachment site (p. 215)
donor (p. 199) (p. 199) lysate (p. 210)
double (mixed) infection (p. 211) generalized transduction lysis (p. 210)
double transformation (p. 209) (p. 213) lysogen (lysogenic bacterium)
endogenote (p. 204) genetic marker (p. 196) (p. 213)
SOLVED PROBLEMS
of recombinants is a function of the order of entry for each c. Refer to the following diagram:
marker. Therefore, we can immediately determine the order
of the genetic markers simply by looking at the percentage met 15.4 m.u. pur 1.8 m.u. thi
of recombinants for any marker among the leu+ recombi- Hfr
nants. Because the inheritance of thr + is the highest, thr +
must be the first marker to enter after leu. The complete met pur thi
order is leu, thr, ile, mal, trp.
F
SOLVED PROBLEM 4
A cross is made between an Hfr that is met + thi+ pur + To compute the distance between met and pur, we
and an F− that is met − thi− pur−. Interrupted-mating stud- compute the percentage of met + pur − thi− , which is
ies show that met + enters the recipient last, and so met + 52/338 = 15.4 m.u. Similarly, the distance between pur and
recombinants are selected on a medium containing sup- thi is 6/338 = 1.8 m.u.
plements that satisfy only the pur and thi requirements.
These recombinants are tested for the presence of the thi+ SOLVED PROBLEM 5
and pur + alleles. The following numbers of individuals are
Compare the mechanism of transfer and inheritance
found for each genotype:
of the lac+ genes in crosses with Hfr, F+ , and F ′ lac+
met + thi+ pur + 280 strains. How would an F− cell that cannot undergo nor-
+
met thi pur+ −
0 mal homologous recombination (rec− ) behave in crosses
+ − + with each of these three strains? Would the cell be able to
met thi pur 6
inherit the lac+ gene?
+ − −
met thi pur 52
SOLUTION
a. Why was methionine (Met) left out of the selection
Each of these three strains donates genes by conjugation. In
medium?
the Hfr and F+ strains, the lac+ genes on the host chromo-
b. What is the gene order? some are donated. In the Hfr strain, the F factor is integrated
c. What are the map distances in recombination units? into the chromosome in every cell, and so chromosomal
markers can be efficiently donated, particularly if a marker
SOLUTION
is near the integration site of F and is donated early. The F+
a. Methionine was left out of the medium to allow selec- cell population contains a small percentage of Hfr cells, in
tion for met + recombinants because met + is the last which F is integrated into the chromosome. These cells are
marker to enter the recipient. The selection for met + responsible for the gene transfer displayed by cultures of F+
ensures that all the loci that we are considering in the cells. In the Hfr and F+-mediated gene transfer, inheritance
cross will have already entered each recombinant that requires the incorporation of a transferred fragment by
we analyze. recombination (recall that two crossovers are needed) into
b. Here, a diagram of the possible gene orders is help- the F− chromosome. Therefore, an F− strain that cannot
ful. Because we know that met enters the recipient undergo recombination cannot inherit donor chromosomal
last, there are only two possible gene orders if the first markers even though they are transferred by Hfr strains or
marker enters on the right: met, thi, pur or met, pur, Hfr cells in F+ strains. The fragment cannot be incorporated
thi. How can we distinguish between these two orders? into the chromosome by recombination. Because these frag-
Fortunately, one of the four possible classes of recom- ments do not possess the ability to replicate within the F−
binants requires two additional crossovers. Each possi- cell, they are rapidly diluted out during cell division.
ble order predicts a different class that arises by four Unlike Hfr cells, F ′ cells transfer genes carried on the F ′
crossovers rather than two. For instance, if the order factor, a process that does not require chromosome trans-
was met, thi, pur, then met + thi− pur + recombinants fer. In this case, the lac+ genes are linked to the F ′ factor
would be very rare. On the other hand, if the order and are transferred with it at a high efficiency. In the F−
was met, pur, thi, then the four-crossover class would cell, no recombination is required because the F ′ lac+ strain
be met + pur − thi+ . From the information given in the can replicate and be maintained in the dividing F− cell pop-
table, the met + pur − thi+ class is clearly the four-cross- ulation. Therefore, the lac+ genes are inherited even in a
over class and therefore the gene order met, pur, thi is rec− strain.
correct.
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a Problem Solving Video. Problems with the icon have an Unpacking the Problem exercise.
www
www
21. Regarding Figure 6-27, outline an experimental proto- 36. A microbial geneticist isolates a new mutation in
col to set up double infection in the lab. E. coli and wishes to map its chromosomal location.
22. In Figure 6-28, what are the physical features of the She uses interrupted-mating experiments with Hfr
plaques of recombinant phages? strains and generalized-transduction experiments with
phage P1. Explain why each technique, by itself, is
23. a. Outline an experimental lab protocol that insufficient for accurate mapping.
would allow the experiment in Figure 6-29 to be
accomplished. 37. In E. coli, four Hfr strains donate the following mark-
ers, shown in the order donated:
b. In Figure 6-29 , do you think that b+ could be
transduced instead of a+? As well as a+? Strain 1: M Z X W C
Strain 2: L A N C W
24. a. In Figure 6-30, which genes show the highest fre- Strain 3: A L B R U
quencies of cotransduction? Strain 4: Z M U R B
b. What is the cotransduction frequency of narC and All these Hfr strains are derived from the same F+
purB? strain. What is the order of these markers on the circu-
25. In Figure 6-31, what essential event on a petri dish is lar chromosome of the original F+?
observed in one experiment and not in the other? 38. You are given two strains of E. coli. The Hfr strain is
26. a. In Figure 6-32, what do the half-red, half-blue seg- arg + ala+ glu+ pro+ leu+ T s ; the F− strain is arg− ala−
ments represent? glu− pro− leu− T r . All the markers are nutritional
b. Would you say that integration is by one or by two except T, which determines sensitivity or resistance
crossovers? to phage T1. The order of entry is as given, with arg +
entering the recipient first and Ts last. You find that
27. a. In Figure 6-33 , which is the rarest λ genotype the F− strain dies when exposed to penicillin (pens),
produced in the initial lysate? but the Hfr strain does not (penr). How would you
b. Draw a diagram to show how the λdgal and λ locate the locus for pen on the bacterial chromosome
helper lysate could have been produced. with respect to arg, ala, glu, pro, and leu? Formulate
28. In Figure 6-34, if F is inserted at minute 45, which your answer in logical, well-explained steps, and draw
two genes would tell you in which orientation it was explicit diagrams where possible.
inserted and in what kind of experiment? 39. A cross is made between two E. coli strains: Hfr
29. Look at Figures 6-34 and 6-35, find the ara gene in arg + bio+ leu+ × F− arg− bio− leu−. Interrupted mat-
each, and compare the gene landscape in its vicinity. ing studies show that arg + enters the recipient last,
Why is there a difference? and so arg + recombinants are selected on a medium
containing bio and leu only. These recombinants are
30. Looking at Figure 6-37, which region would you say tested for the presence of bio+ and leu+ . The fol-
shows the best proportionality between the DNA and lowing numbers of individuals are found for each
the recombination maps? genotype:
31. In Figure 6-38, precisely which gene is eventually iden- arg + bio+ leu+ 320 arg + bio− leu+ 0
tified from the genome sequence? + + − + − −
arg bio leu 8 arg bio leu 48
a. What is the gene order?
BASIC PROBLEMS
b. What are the map distances in recombination
32. Describe the state of the F factor in an Hfr, F+, and F− percentages? www
strain.
40. Linkage maps in an www
Hfr bacterial strain are calculated
+
33. How does a culture of F cells transfer markers from in units of minutes (the number of minutes between
the host chromosome to a recipient? genes indicates the length of time that it takes for the
34. With respect to gene transfer and the integration of the second gene to follow the first in conjugation). In mak-
transferred gene into the recipient genome, compare ing such maps, microbial geneticists assume that the
bacterial chromosome is transferred from Hfr to F− at a
a. Hfr crosses by conjugation and generalized
constant rate. Thus, two genes separated by 10 minutes
transduction.
near the origin end are assumed to be the same physical
b. F′ derivatives such as F ′ lac and specialized distance apart as two genes separated by 10 minutes
transduction. near the F− attachment end. Suggest a critical experi-
35. Why is generalized transduction able to transfer any ment to test the validity of this assumption.
gene, but specialized transduction is restricted to only 41. A particular Hfr strain normally transmits the pro+
a small set? marker as the last one in conjugation. In a cross of
this strain with an F− strain, some pro+ recombinants 46. Recall that, in Chapter 4, we considered the possibil-
are recovered early in the mating process. When these ity that a crossover event may affect the likelihood
pro+ cells are mixed with F− cells, the majority of of another crossover. In the bacteriophage T4, gene a is
the F− cells are converted into pro+ cells that also carry 1.0 m.u. from gene b, which is 0.2 m.u. from gene c.
the F factor. Explain these results. The gene order is a, b, c. In a recombination experi-
ment, you recover five double crossovers between a and
42. F′ strains in E. coli are derived from Hfr strains. In
c from 100,000 progeny viruses. Is it correct to con-
some cases, these F′ strains show a high rate of integra-
clude that interference is negative? Explain your answer.
tion back into the bacterial chromosome of a s econd
strain. Furthermore, the site of integration is often the 47. E. coli cells were infected with two strains of T4 virus.
site occupied by the fertility factor in the original Hfr One strain is minute (m), rapid lysis (r), and turbid (t);
strain (before production of the F′ strains). Explain the other is wild type for all three markers. The lytic
these results. products of this infection were plated and classified.
The resulting 10,342 plaques were distributed among
43. You have two E. coli strains, F− str s ala− and Hfr
eight genotypes as follows:
str s ala+, in which the F factor is inserted close to
ala+. Devise a screening test to detect strains carrying mrt 3469 m++ 521
F′ ala+ . +++ 3727 +r t 475
mr + 854 +r + 171
44. Five Hfr strains A through E are derived from a sin-
m+t 163 ++t 963
gle F+ strain of E. coli. The following chart shows the
entry times of the first five markers into an F− strain a. What are the linkage distances between m and r,
when each is used in an interrupted-c onjugation between r and t, and between m and t.
experiment: b. Determine the linkage order for the three genes.
A B C D E c. What is the coefficient of coincidence (see
mal +
(1) ade +
(13) pro +
(3) + +
pro (10) his (7) Chapter 4) in this cross? What does it signify?
str s +
(11) his +
(28) met (29) gal + (16) gal + (17) 48. With the use of P22 as a generalized transducing phage
grown on a pur + pro+ his+ bacterial donor, a recipient
ser + (16) gal + (38) xyl + (32) his+ (26) pro+ (23) strain of genotype pur− pro− his− was infected and
ade+ (36) pro+ (44) mal + (37) ade+ (41) met + (49) incubated. Afterward, transductants for pur + , pro+ ,
and his+ were selected individually in experiments I, II,
his+ (51) met + (70) str s (47) ser + (61) xyl + (52)
and III, respectively.
a. Draw a map of the F+ strain, indicating the a. What medium is used in each of these selection
positions of all genes and their distances apart in experiments?
minutes. b. The transductants were examined for the presence
b. Show the insertion point and orientation of the of unselected donor markers, with the following
F plasmid in each Hfr strain. results:
c. In the use of each of these Hfr strains, state which I II III
allele you would select to obtain the highest pro- − − − −
portion of Hfr exconjugants. pro his 86% pur his 44% pur pro− 20%
−
45. Streptococcus pneumoniae cells of genotype str s mtl − pro+ his− 0% pur + his− 0% pur + pro− 14%
are transformed by donor DNA of genotype str r mtl + pro− his+ 10% pur− his+ 54% pur− pro+ 61%
and (in a separate experiment) by a mixture of two
pro+ his+ 4% pur + his+ 2% pur + pro+ 5%
DNAs with genotypes str r mtl − and str s mtl + . The
accompanying table shows the results.
What is the order of the bacterial genes?
Percentage of cells transformed into
c. Which two genes are closest together?
Transforming DNA str r mtl − str s mtl + str r mtl +
d. Based on your answer to part c, explain the relative
str r mtl + 4.3 0.40 0.17 proportions of genotypes observed in experiment II.
str r mtl − + str s mtl + 2.8 0.85 0.0066 49. Although most λ-mediated gal + transductants are
inducible lysogens, a small percentage of these trans-
a. What does the first row of the table tell you? ductants in fact are not lysogens (that is, they contain
Why? no integrated λ). Control experiments show that these
b. What does the second row of the table tell you? transductants are not produced by mutation. What is
Why? the likely origin of these types?
50. An ade+ arg + cys+ his+ leu+ pro+ bacterial strain is d. Are nad and pdx on the same side or on opposite
known to be lysogenic for a newly discovered phage, sides of pur? Explain.
but the site of the prophage is not known. The bacte- (Draw the exchanges needed to produce the various
rial map is transformant classes under either order to see which
arg
requires the minimum number to produce the results
his obtained.)
cys
leu 52. In a generalized-transduction experiment, phages are
collected from an E. coli donor strain of genotype
ade
pro cys+ leu+ thr + and used to transduce a recipient of
genotype cys− leu− thr− . Initially, the treated recipient
The lysogenic strain is used as a source of the phage, population is plated on a minimal medium supple-
and the phages are added to a bacterial strain of gen- mented with leucine and threonine. Many colonies are
otype ade− arg− cys− his− leu− pro− . After a short obtained.
incubation, samples of these bacteria are plated on six
a. What are the possible genotypes of these colonies?
different media, with the supplementations indicated in
the following table. The table also shows whether colo- b. These colonies are then replica plated onto three
nies were observed on the various media. different media: (1) minimal plus threonine only;
(2) minimal plus leucine only; and (3) minimal.
Nutrient supplementation in medium Presence of
What genotypes could, in theory, grow on these
Medium Ade Arg Cys His Leu Pro colonies three media?
1 − + + + + + N c. Of the original colonies, 56 percent are observed
2 + − + + + + N to grow on medium 1, 5 percent on medium 2, and
3 + + − + + + C no colonies on medium 3. What are the actual gen-
otypes of the colonies on media 1, 2, and 3?
4 + + + − + + N
d. Draw a map showing the order of the three genes
5 + + + + − + C
and which of the two outer genes is closer to the
6 + + + + + − N middle gene.
(In this table, a plus sign indicates the presence of a 53. Deduce the genotypes of the following E. coli strains 1
nutrient supplement, a minus sign indicates that a sup- through 4:
plement is not present, N indicates no colonies, and C
Minimal Minimal plus
indicates colonies present.) arginine
a. What genetic process is at work here?
1 2 1 2
b. What is the approximate locus of the prophage? 3 4 3 4
51. In a generalized-transduction system using P1 phage,
the donor is pur + nad + pdx− and the recipient is
Minimal plus Minimal plus
pur− nad − pdx+ . The donor allele pur + is initially methionine arginine and methionine
selected after transduction, and 50 pur + transductants
1 2 1 2
are then scored for the other alleles present. Here are the
3 4 3 4
results:
Genotype Number of colonies
+
nad pdx + 3
54. In an interrupted-conjugation experiment in E. coli,
nad + pdx− 10 the pro gene enters after the thi gene. A pro+ thi+
nad − pdx+ 24 Hfr is crossed with a pro− thi− F− strain, and excon-
jugants are plated on medium containing thiamine
nad − pdx− 13 but no proline. A total of 360 colonies are observed,
50 and they are isolated and cultured on fully supple-
mented medium. These cultures are then tested for
a. What is the cotransduction frequency for pur and their ability to grow on medium containing no pro-
nad? line or thiamine (minimal medium), and 320 of the
b. What is the cotransduction frequency for pur and cultures are found to be able to grow but the remain-
pdx? der cannot.
c. Which of the unselected loci is closest to pur? a. Deduce the genotypes of the two types of cultures.
a. What donor genes are being selected on each type Number of Number of
of agar? Drugs added colonies Drugs added colonies
b. The following table shows the number of colonies None 10,000 BC 50
on each type of agar for samples taken at various
A 1155 BD 48
times after the strains are mixed. Use this informa-
tion to determine the order of genes a, b, and c. B 1147 CD 785
C 1162 ABC 31
Time of sampling Number of colonies on agar of type
(minutes) D 1140 ABD 43
1 2 3
AB 47 ACD 631
0 0 0 0
AC 641 BCD 35
5 0 0 0
AD 941 ABCD 29
7.5 102 0 0
10 202 0 0 a. One of the genes is distant from the other three,
12.5 301 0 74 which appear to be closely linked. Which is the dis-
tant gene?
15 400 0 151
17.5 404 49 225 b. What is the likely order of the three closely linked
genes?
20 401 101 253
61. You have two strains of λ that can lysogenize E. coli;
25 398 103 252
their linkage maps are as follows:
c. From each of the 25-minute plates, 100 colonies are Strain X Strain Y
picked and transferred to a petri dish containing agar c2 b2 c1 b1
with all the nutrients except D. The numbers of colo-
nies that grow on this medium are 90 for the sample
from agar type 1, 52 for the sample from agar type
2, and 9 for the sample from agar type 3. Using these d1 a1 d2 a2
1 2 3 1 2 3
data, fit gene d into the sequence of a, b, and c.
d. At what sampling time would you expect colonies The segment shown at the bottom of the chromosome,
to first appear on agar containing C and strepto- designated 1–2–3, is the region responsible for pairing
mycin but no A or B? and crossing over with the E. coli chromosome. (Keep
59. In the cross Hfr aro+ arg + ery r str s × F− aro− arg− erys the markers on all your drawings.)
str r , the markers are transferred in the order given a. Diagram the way in which λ strain X is inserted
(with aro+ entering first), but the first three genes into the E. coli chromosome (so that the E. coli is
are very close together. Exconjugants are plated on a lysogenized).
medium containing Str (streptomycin, to kill Hfr cells),
b. The bacteria that are lysogenic for strain X can be
Ery (erythromycin), Arg (arginine), and Aro (aromatic
superinfected by using strain Y. A certain percent-
amino acids). The following results are obtained for 300
age of these superinfected bacteria become “dou-
colonies isolated from these plates and tested for growth
bly” lysogenic (that is, lysogenic for both strains).
on various media: on Ery only, 263 strains grow; on
Diagram how it will take place. (Don’t worry
Ery + Arg, 264 strains grow; on Ery + Aro, 290 strains
about how double lysogens are detected.)
grow; on Ery + Arg + Aro, 300 strains grow.
c. Diagram how the two λ prophages can pair.
a. Draw up a list of genotypes, and indicate the num-
ber of individuals in each genotype. d. Crossover products between the two prophages
can be recovered. Diagram a crossover event and
b. Calculate the recombination frequencies.
the consequences.
c. Calculate the ratio of the size of the arg -to- aro
62. You have three strains of E. coli. Strain A is F ′ cys+
region to the size of the ery-to-arg region.
trp1/cys+ trp1 (that is, both the F ′ and the chromosome
60. A bacterial transformation is performed with a carry cys+ and trp1, an allele for tryptophan require-
donor strain that is resistant to four drugs, A, B, C, ment). Strain B is F− cys – trp2 Z (this strain requires
and D, and a recipient strain that is sensitive to all cysteine for growth and carries trp2, another allele
four drugs. The resulting recipient cell population is causing a tryptophan requirement; strain B is lysogenic
divided and plated on media containing various com- for the generalized transducing phage Z). Strain C is
binations of the drugs. The following table shows the F− cys+ trp1 (it is an F− derivative of strain A that has
results. lost the F ′ ). How would you determine whether trp1
and trp2 are alleles of the same locus? (Describe the chromosome, thereby allowing the F ′ lac to be main-
crosses and the results expected.) tained at high temperatures. By combining this selec-
63. A generalized transducing phage is used to transduce tion with a simultaneous selection for resistance to
an a− b− c− d − e− recipient strain of E. coli with an T1 phage infection, they found that the only survivors
a+ b+ c+ d + e+ donor. The recipient culture is plated were cells in which the F ′ lac had integrated into the
on various media with the results shown in the follow- tonB locus, as shown here:
ing table. (Note that a− indicates a requirement for A F9
tonB att80
as a nutrient, and so forth.) What can you conclude lac
about the linkage and order of the genes?
Compounds added Presence (+) or absence (−) This result placed the lac region near the integration
to minimal medium of colonies site for phage φ80. Describe the subsequent steps that
CDE − the researchers must have followed to isolate the spe-
cialized transducing particles of phage φ80 that carried
BDE − the lac region.
BCE +
65. Wild-type E. coli takes up and concentrates a certain
BCD + red food dye, making the colonies blood red. Trans-
ADE − poson mutagenesis was used, and the cells were plated
on food dye. Most colonies were red, but some colo-
ACE −
nies did not take up dye and appeared white. In one
ACD − white colony, the DNA surrounding the transposon
ABE − insert was sequenced, with the use of a DNA repli-
ABD + cation primer identical with part of the end of the
transposon sequence, and the sequence adjacent to
ABC −
the transposon was found to correspond to a gene
64. In 1965, Jon Beckwith and Ethan Signer devised a of unknown function called atoE, spanning positions
method of obtaining specialized transducing phages 2.322 through 2.324 Mb on the map (numbered from
carrying the lac region. They knew that the integra- an arbitrary position zero). Propose a function for
tion site, designated att80, for the temperate phage φ80 atoE. What biological process could be investigated in
(a relative of phage λ ) was located near tonB, a gene this way, and what other types of white colonies might
that confers resistance to the virulent phage T1: be expected?
M
CHAPTER 7 olecular genetics, also called molecular biology, is the study of how DNA, RNA,
DNA: STRUCTURE AND REPLICATION
and protein molecules store, transmit, and exchange the information that deter-
CHAPTER 8 mines the phenotypes of organisms. The foundation of molecular biology is the
RNA: TRANSCRIPTION, PROCESSING, central dogma (DNA makes RNA makes protein) and its integral molecular processes
AND DECAY
(DNA replication, transcription, and translation) (see Figure 1-10). Molecular biology
CHAPTER 9 provides mechanistic explanations for why mutations in the sequence of a gene alter its
PROTEINS AND THEIR SYNTHESIS
expression and function and why cells that contain the same genome sequence express
CHAPTER 10 different genes in response to developmental and environmental signals. This mechanistic
GENE ISOLATION AND MANIPULATION understanding can be used in innumerable ways, including the treatment of diseases and
CHAPTER 11 the improvement of crop plants. As an example, in Chapter 1 (pages 16–18), you read
REGULATION OF GENE EXPRESSION IN about a quantitative trait locus (QTL) that allows rice plants to survive submergence in
BACTERIA AND THEIR VIRUSES
deep water for up to two weeks, making the plants tolerant to floods. Researchers used
CHAPTER 12 molecular biology principles and techniques to determine that a gene called submergence
REGULATION OF TRANSCRIPTION IN tolerant (SUB1) contains the information that confers flood tolerance and that the encoded
EUKARYOTES
SUB1 protein functions by regulating the expression of other genes. Researchers were then
CHAPTER 13 able to use this understanding to increase the yield of rice and other plants.
THE GENETIC CONTROL OF DEVELOPMENT
The large number of molecules and processes in molecular biology is related to the
CHAPTER 14 huge variety of forms and functions of organisms. As an example, in eukaryotic but not
GENOMES AND GENOMICS prokaryotic cells, the physical separation of DNA from ribosomes by the nuclear mem-
brane necessitates mechanisms that transport RNAs and proteins between the nucleus and
the cytoplasm. Despite differences among organisms, there are principles in molecular bio
logy that apply to all organisms from single-celled bacteria and yeasts to multi-celled plants
and animals because all cells are largely the same; they grow, divide, and respond to devel-
opmental and environmental signals.
The following catalog of core principles in molecular and developmental genetics will
help guide your learning of the molecules and molecular processes that are described in
Chapters 7–14. Knowing these principles should make it easier for you to understand the
unique details of each molecule and molecular process and to apply the information to
solve present-day problems. We suggest that you think of the principles as shelves to orga-
nize molecular biology information. As you work through the chapters, return to the prin-
ciples and place the information that you learn on the shelves.
232
In this example, a noncoding DNA sequence called an enhancer contains information that regulates transcription of the
coding DNA sequence of a gene that contains information to make an RNA.
Each molecular process in the central dogma of molecular biology (indicated in parentheses) involves DNA-DNA,
RNA-DNA, or RNA-RNA base pairing.
+ No interaction
Structural change
+ Functional interaction
In this example, a change in the structure of the circular protein converts it from nonfunctional to functional by enabling
it to interact with the hexagon protein.
234
RNA synthesis
(RNA polymerase)
A
U G AACUGCACCGU
C (RNA nucleases)
RNA decay
In this example, an RNA is synthesized from nucleotides (A, C, G, and U) by RNA polymerase in the process of
transcription, and, in the reverse reaction, the RNA is taken apart into nucleotides by RNA nucleases.
Cell surface
Nutrient level Transcription level
Signaling
Receptor
pathway
Low High Low High
In this example, the amount of nutrients outside of a cell (i.e., an environmental signal) is sensed by a receptor on the
surface of the cell. The receptor then turns on a signaling pathway that activates the transcription of genes whose
function is to respond to the amount of nutrients.
235
Off Off On
In this example, the gene is transcribed in cell type C but not in A and B because the combined activities of two
transcription factors (i.e., binding of the green and purple proteins to the green and purple regulatory sequences in the
gene) is required to activate transcription. In cell types A and B, only one of the transcription factors is expressed, so
transcription is not activated.
A A
F B B
E C E
D D
236
b. Loss-of-function experiments
Loss-of-function experiments are used to determine whether molecules are necessary
for a molecular process or phenotype to occur. A common type of loss-of-function
experiment is a gene knockout in a whole organism. If a molecular process or pheno-
type differs between knockout and wild-type organisms, it can be concluded that the
gene is necessary for the event. In the following example, knocking out gene C changed
the color of cell type X into that of cell type Y, but knocking out gene F had no effect
on the cell color. Thus, gene C is necessary for the molecular mechanism that generates
the color of cell type X, but gene F is not.
Cell type X
Knock out Knock out
A A A
gene C gene F
F B F B B
E E C E C
D D D
c. Gain-of-function experiments
Gain-of-function experiments are used to determine whether molecules are sufficient
for a molecular process or phenotype to occur. A common type of gain-of-function
experiment is to express a gene in a cell that normally does not express the gene. If a
process or phenotype differs between cells that misexpress the gene and wild-type cells,
it can be concluded that the gene is sufficient for these events. In the following example,
misexpression of gene C changed the color of cell type Y, but misexpression of gene F
had no effect on the cell color. Thus, gene C is sufficient to activate the mechanism that
generates the color of cell type X, but gene F is not.
Cell type Y
Misexpress Misexpress
A A A
gene F gene C
F B B B
E E E C
D D D
237
239
I
n this chapter, we describe the structure of DNA and the An artistic representation of DNA
process of DNA replication, which makes an identical
copy of DNA every time a cell divides. The story begins
in the early 1900s, when results of several experiments led
scientists to conclude that DNA, rather than another bio-
logical molecule such as carbohydrate, protein, or lipid,
is the genetic material. DNA is a simple molecule made
up of only four building blocks called nucleotides. It was
thus necessary to understand how this very simple mole-
cule could be the blueprint for the incredible diversity of
organisms on Earth.
A large part of this understanding came from the
structure of DNA, which was determined in 1953 by
James Watson and Francis Crick through modeling
based on the data of others. Their model of the struc-
ture of DNA was revolutionary because it defined genes
in chemical terms and, in doing so, paved the way for
understanding gene action and heredity at a molecular
level. A measure of the importance of their discovery is
that the double-helical structure of DNA has become a
cultural icon that is seen more and more frequently in
various forms of art (Figure 7-1).
The model of DNA proposed by Watson and Crick was
built upon the results of scientists before them. They relied
on earlier discoveries of the chemical composition of DNA
and the ratios of its nucleotide bases. In addition, pictures
of DNA fibers produced by X-ray diffraction revealed to
the trained eye that DNA is a helix of precise dimensions.
Watson and Crick concluded that DNA is a double helix
composed of two strands of linked nucleotides that wind
around each other.
The proposed structure of DNA immediately sug-
gested that the sequence of nucleotides composing the FIGURE 7-1 A 15-meter-tall sculpture of DNA is housed at the Prince
two DNA strands of the helix could serve as a blueprint Felipe Museum of Science in Valencia, Spain. [Peter Blixt/Alamy.]
for constructing an organism. In addition, the structure
hinted at how the blueprint could be copied into all cells
in an organism. Because of the rules of base complemen- DNA copies in a process called DNA replication. This
tarity discovered by Watson and Crick, the sequence of information is essential for understanding the molecular
one strand determines the sequence of the other strand. basis of genes and genetic inheritance. Precisely how DNA
In this way, genetic information in the sequence of is replicated is still an active area of research more than
DNA can be passed from a mother cell to each daugh- 65 years after the discovery of the double-helix structure.
ter cell by having each of the separated strands of Our current understanding of the mechanism of replica-
DNA serve as a template for producing new copies of tion gives a central role to a protein machine called the
double-stranded DNA. replisome. This complex of proteins coordinates numer-
In summary, this chapter focuses on the structure of ous reactions that are necessary for rapid and accurate
DNA and the molecules and mechanisms that produce replication of DNA.
240
7.1 DNA IS THE GENETIC strain can be changed, that is, “transformed,” by mixing
it with a different, heat-killed bacterial strain. His stud-
MATERIAL ies used the bacterium Streptococcus pneumoniae, which
causes pneumonia in humans and is normally lethal in
LO 7.1 Describe evidence demonstrating that DNA is the
mice. However, some strains of this bacterial species have
genetic material.
evolved to be less virulent (less able to cause disease or
death). In experiments summarized in Figure 7-2, Griffith
Before we see how Watson and Crick solved the structure
used two strains that are distinguishable by the appear-
of DNA, let’s review what was known about genes and
ance of their colonies when grown in laboratory cultures.
DNA at the time they began their historic collaboration:
One strain was a normal virulent type, deadly to most
1. Genes—the hereditary “factors” described by Mendel— laboratory animals. The cells of this strain are enclosed
were known to be associated with specific traits, but their in a polysaccharide capsule, giving colonies a smooth
physical nature was not understood. Similarly, mutations appearance; hence, this strain is identified as S. Griffith’s
were known to alter gene function, but the precise chemi- other strain was a mutant, nonvirulent type that grows in
cal nature of a mutation was not understood. mice but is not lethal. In this strain, the polysaccharide
2. The one-gene–one-enzyme hypothesis (described in coat is absent, giving colonies a rough appearance; this
Chapter 5) postulated that genes determine the struc- strain is called R.
ture of proteins. Griffith killed some virulent S cells by boiling them. He
then injected the heat-killed cells into mice. The mice sur-
3. Genes were known to be carried on chromosomes.
vived, showing that the carcasses of the cells do not cause
4. Chromosomes were known to consist of DNA and death. However, mice injected with a mixture of heat-killed
protein. virulent S cells and live nonvirulent R cells did die. Fur-
5. As described next, experiments beginning in the 1920s thermore, live cells could be recovered from the dead mice;
revealed that DNA is the genetic material. these cells gave smooth colonies and were virulent on sub-
sequent injection. Somehow, the cell debris of the boiled
S cells converted some of the live R cells into live S cells.
The discovery of bacterial transformation:
That is, the live R cells were transformed into S cells by
the Griffith experiment picking up some chemical component of the dead S cells.
In 1928, Frederick Griffith made the puzzling observa- The process, already discussed in Chapter 6, is called
tion that the genotype and phenotype of a live bacterial transformation.
S strain R strain
live cells live cells FIGURE 7-2 The presence of
heat-killed S cells transforms live
R cells into live S cells. (a) Mice
(c) (d) die after injection with virulent
S cells. (b) Mice survive after
injection with R cells. (c) Mice
survive after injection with heat-
+ killed S cells. (d) Mice die after
injection with a mixture of heat-
Mouse lives Mouse dies killed S cells and live R cells. Live
S cells were isolated from dead
mice, indicating that heat-killed
S strain R strain S strain Live S strain S cells somehow transform
heat-killed live cells heat-killed recovered
nonvirulent R cells into virulent
S cells.
Evidence that DNA is the genetic material KEY CONCEPT The demonstration that DNA is the trans-
in bacteria: the Avery, Macleod, and forming agent was the first evidence that genes (the hereditary
material) are composed of DNA.
McCarty experiments
The next step was to determine which chemical component
of dead S cells caused transformation. This molecule had
changed the genotype of the recipient strain and therefore
Evidence that DNA is the genetic material
was a candidate for the hereditary material. The problem in phage: the Hershey–Chase experiment
was solved by experiments conducted in 1944 by Oswald The experiments conducted by Avery and his colleagues
Avery, Colin MacLeod, and Maclyn McCarty (Figure 7-3). were definitive, but many scientists were reluctant to accept
Their approach to the problem was to destroy all of the DNA (rather than proteins) as the genetic material. After
major categories of chemicals in an extract of dead S cells all, how could such a low-complexity molecule as DNA
one at a time, and to find out if the extract had lost the encode the diversity of all living things? In 1952, Alfred
ability to transform. Virulent S cells had a smooth poly- Hershey and Martha Chase provided additional evidence
saccharide coat, whereas nonvirulent R cells did not; in an experiment that made use of bacteriophage T2 (or
hence, polysaccharides were an obvious candidate for the phage T2 for short), a virus that infects bacteria. They rea-
transforming agent. However, when polysaccharides were soned that the infecting phage must inject into the bacte-
destroyed, the mixture could still transform. Lipids, RNAs, rium the specific information that directs the production of
and proteins were all similarly shown not to be the trans- new viral particles. If they could find out what material the
forming agent. In contrast, the mixture lost its transforming phage was injecting into the bacterial host, they would have
ability when the donor mixture was treated with the enzyme determined the genetic material of phages.
deoxyribonuclease (DNase), which destroys DNA. These The phage is relatively simple in molecular composition.
results strongly implicated DNA as the genetic material. The T2 structure is similar to T4 shown in Figures 6-22 to
It is now known that fragments of the transforming DNA 6-24. Most of its structure is protein, with DNA contained
that confer virulence enter the bacterial chromosome and inside the protein sheath of its “head.” Hershey and Chase
replace their counterparts that confer nonvirulence. used radioisotopes to give DNA and protein distinct labels
S strain
extract
R strain
Mouse dies Mouse dies Mouse dies Mouse dies Mouse dies Mouse lives
FIGURE 7-3 DNA is the genetic material that transforms nonvirulent R cells into virulent S cells. Mice
survive when injected with a mixture of heat-killed S cells with destroyed DNA and live nonvirulent R
cells. However, destroying polysaccharides, lipids, RNAs, or proteins does not allow the mice to survive.
Thus, DNA, but not polysaccharide, lipid, RNA, or protein, is necessary for transformation.
Blend
and
centrifuge
Pellet
that they could track during infection. Phosphorus is not packaging that is discarded after delivering the viral DNA
found in the amino acid building blocks of proteins but is to the bacterial cell.
found in DNA; conversely, sulfur is not in the nucleotide
building blocks of DNA but is in proteins. Hershey and
Chase incorporated a radioisotope of phosphorus ( 32P) 7.2 DNA STRUCTURE
into DNA and that of sulfur (35S) into proteins of separate
phage cultures. Radioisotopes are unstable (i.e., radioac- LO 7.2 Describe evidence used to construct the double
tive) isotopes of an element that emit radiation to transform helix model of DNA.
into a more stable form. Emitted radiation can be measured LO 7.3 Draw the chemical structure of a DNA double helix.
using instruments such as a scintillation counter or a Geiger
counter or by autoradiography (described in Section 7.3). Even before the structure of DNA was elucidated, genetic
As shown in Figure 7-4, after labeling the phage DNA studies indicated that the hereditary material must have
and proteins, they then infected two E. coli cultures with three key properties:
many phage particles per cell: one E. coli culture received
1. Because essentially every cell in the body of an organism
phage labeled with 32 P, and the other received phage
has the same genetic makeup, accurate replication of the
labeled with 35S. After allowing sufficient time for infec-
genetic material at every cell division is crucial. Thus,
tion to take place, they sheared the empty phage carcasses
structural features of the genetic material must allow
(called ghosts) off the bacterial cells in a kitchen blender.
accurate replication. The structural features of DNA will
They separated the bacterial cells from the phage ghosts
be covered in this section of the chapter.
in a centrifuge and then measured the radioactivity in
the solid pellet of bacteria and the liquid supernatant of 2. Because it must encode the collection of proteins
phage ghosts. When the 32P-labeled phages were used to expressed by an organism, structural features of the
infect E. coli, the radioactivity ended up inside the bacte- genetic material must have informational content. How
rial cells, indicating that phage DNA entered the cells. In information coded in DNA is deciphered to produce
contrast, when 35S-labeled phages were used, the radioac- proteins is the subject of Chapters 8 and 9.
tive material ended up in the phage ghosts, indicating that 3. Because hereditary changes, called mutations, provide the
phage proteins did not enter the bacterial cell. In addi- raw material for evolutionary selection, genetic material
tion, the progeny of 32P-labeled phages remained labeled, must be able to change on rare occasion. Nevertheless, the
but the progeny of 35S-labeled phages were not labeled. structure of the genetic material must be stable enough for
These data once again indicated that DNA is the hered- an organism to rely on its encoded information. The mech-
itary material. The phage proteins are mere structural anisms of DNA mutations will be covered in Chapter 16.
DNA structure before Watson and Crick Two of the bases, adenine and guanine, have a double-ring
structure characteristic of a class of chemicals called
Consider the discovery of the double-helical structure of
purines. The other two bases, cytosine and thymine, have a
DNA by Watson and Crick as the solution to a complicated
single-ring structure characteristic of another class of chem-
three-dimensional puzzle. To solve this puzzle, Watson and
icals called pyrimidines. Carbon and nitrogen atoms in the
Crick used a process called “model building,” in which they
rings of the bases are assigned numbers for ease of refer-
assembled the results of earlier and ongoing experiments
ence. Carbon atoms in the sugar group also are assigned
(the puzzle pieces) to form the three-dimensional puzzle
numbers—in this case, each number is followed by a prime
(the double helix model). To understand how they built the
(1′, 2′, and so forth).
DNA model, we first need to know what pieces of the puz-
zle were available to them.
KEY CONCEPT DNA contains four bases—two purines
(adenine and guanine) and two pyrimidines (cytosine and
The building blocks of DNA The first piece of the puz-
thymine).
zle was knowledge of the basic building blocks of DNA.
As a chemical, DNA is quite simple. It contains three com-
ponents: (1) phosphate, (2) a sugar called deoxyribose, The chemical subunits of DNA are nucleotides or, more
and (3) four nitrogenous bases—adenine, guanine, cyto- specifically, deoxynucleotides, each composed of a phos-
sine, and thymine (Figure 7-5). The sugar in DNA is called phate group, a deoxyribose sugar molecule, and one of the
“deoxyribose” because it contains ribose sugars that are four bases (Figure 7-5). It is convenient to refer to each
missing an oxygen atom. Deoxyribose has a hydrogen atom nucleotide by the first letter of the name of its base: A, G, C,
(H) at the 2′-carbon atom, unlike ribose (a component of or T. The nucleotide with the adenine base is called deoxy-
RNA), which has a hydroxyl (OH) group at that position. adenosine 5′-monophosphate and abbreviated dAMP,
Pyrimidine nucleotides
NH2 O
H3C
4
3NH
4
5 3N 5
Cytosine (C) 6 2
Thymine (T)
6 2 1
1
N O O N O
O
–O –O P O
P O CH2 O CH2 O
–O H H –O H H
H H H H
OH H OH H
Deoxycytidine 5′-monophosphate (dCMP) Deoxythymidine 5′-monophosphate (dTMP)
FIGURE 7-5 Nucleotides are the fundamental building blocks of DNA. All nucleotides have a phos-
phate, a sugar, and a base. The sugar is called deoxyribose because it is a variant of ribose that lacks
an oxygen atom, indicated by the red arrow. There are two purine bases (adenine and guanine) and two
pyrimidine bases (cytosine and thymine). Note that each of the bases contains nitrogen atoms.
where the 5′ refers to the position of the carbon atom in the KEY CONCEPT DNA contains an equal amount of A and T
sugar ring to which the single (mono) phosphate group is nucleotides and G and C nucleotides. Organisms vary in the
attached. The other nucleotides are named using the same relative amount of A + T versus G + C, but different tissues
convention. in the same organism have the same relative amount of A + T
versus G + C.
KEY CONCEPT DNA nucleotides are known as deoxynucle-
otides and are composed of a phosphate, a deoxyribose, and
a purine or pyrimidine base. X-ray diffraction analysis of DNA: Rosalind Franklin
The third piece of the puzzle came from the X-ray diffrac-
tion pattern of DNA fibers (Figure 7-6a) that was collected
Chargaff’s rules of base composition The second piece by Rosalind Franklin (Figure 7-6b). In this experiment,
of the puzzle used by Watson and Crick came from work X rays were fired at DNA fibers that were collected from
done several years earlier by Erwin Chargaff. Studying a cells, as shown in the opening photograph of this chap-
large selection of DNAs from different organisms (Table 7-1), ter. The scatter of the X rays from the fibers is detected
C hargaff established certain empirical rules about the as spots on photographic film (Figure 7-6a). The angle of
amounts of each type of nucleotide found in DNA: scatter represented by each spot on the film gives infor-
1. The total amount of purine nucleotides (A + G) always mation about the position of an atom or certain groups
equals the total amount of pyrimidine nucleotides (T + C). of atoms in DNA. Darker spots are where the film was
hit multiple times by X rays from repeated parts of DNA
2. The amount of A always equals the amount of T, and
such as nucleotide bases. This procedure is not simple
the amount of G always equals the amount of C; that is,
to carry out (or to explain), and interpretation of the
A/T and G/C is close to 1.0, regardless of the source of
spot patterns requires complex mathematical treatment
DNA (Table 7-1).
that is beyond the scope of this text. The available data
3. The amount of A + T is not necessarily equal to the suggested that DNA is long and skinny and that it has
amount of G + C, as can be seen in the last column of two similar parts that are parallel to each other and run
Table 7-1. The (A + T)/(G + C) ratio varies among dif- along the length of the molecule. The X-ray data showed
ferent organisms. For example, in sea urchins, the ratio that DNA is helical, like a spiral staircase. Unknown to
is 1.85, indicating that the sea urchin genome has almost Franklin, her best X-ray picture (Figure 7-6a) was shown
twice as much A + T than G + C; it is said to be AT-rich. to Watson and Crick, and this crucial piece of the puzzle
In contrast, the Mycobacterium tuberculosis genome is allowed them to deduce the three-dimensional structure
GC-rich, with about twice as much G + C than A + T. of DNA (Figure 7-7).
However, the ratio is virtually the same in different tis-
sues of the same organism (as seen for human tissues
KEY CONCEPT The X-ray diffraction pattern of DNA
in the last three rows of Table 7-1), supporting the idea
showed that it is a long and skinny, two-stranded helix (that is,
that all cells of an organism have the same genomic
a double helix).
DNA sequence.
Rosalind Franklin’s critical experimental result FIGURE 7-6 (a) The X-ray diffrac-
tion pattern of DNA determined by
(a) (b)
(b) Rosalind Franklin. [(a) OMIKRON/
Science Source; (b) © World History
Archive/Alamy Stock Photo.]
The DNA double helix structure: main requirements for a hereditary molecule: the ability to
Watson and Crick store information, the ability to be replicated, and the ability
to mutate.
A 1953 paper by Watson and Crick in the journal Nature By studying models that they made of the structure,
began with two sentences that ushered in a new age of biol- Watson and Crick realized that the observed diameter of
ogy: “We wish to suggest a structure for the salt of deoxyri- the double helix (known from the X-ray data) would be
bose nucleic acid (D.N.A.). This structure has novel features explained if a purine base always pairs (by hydrogen bond-
which are of considerable biological interest.”1 The struc- ing) with a pyrimidine base (Figure 7-8). Such pairing would
ture of DNA had been a subject of great debate since the account for the (A + G) = (T + C) regularity observed by
experiments of Avery and co-workers in 1944. The gen- Chargaff, but it would predict four possible pairings: A-T,
eral composition of DNA was known, but how the parts G-T, A-C, and G-C. However, Chargaff’s data indicate that
fit together was not known. The structure had to fulfill the G pairs only with C, and A pairs only with T. Watson and
Crick concluded that each base pair consists of one purine
Pyrimidine + pyrimidine:
DNA too skinny
Purine + purine:
DNA too wide
Purine + pyrimidine:
DNA width is consistent
with X-ray data
Length of
FIGURE 7-7 James Watson (left) and Francis Crick (right) with their
hydrogen
three-dimensional DNA model and a two-dimensional drawing of DNA bond
on the wall. [BARRINGTON BROWN/Science Source.]
FIGURE 7-8 Pairing of purines with pyrimidines accounts exactly for
1
J. Watson and F. Crick, Nature 171:737, 1953. the diameter of the DNA double helix determined from X-ray data.
base and one pyrimidine base, paired according to the strand is formed by alternating phosphate and deoxyribose
following rule: G pairs with C (G-C), and A pairs with T sugar units that are connected by phosphodiester linkages
(A-T). These are called complementary bases. The double (Figure 7-9b). These linkages are used to describe how a
helix accounted nicely for Franklin’s X-ray data as well as nucleotide chain is organized. As already mentioned, the
Chargaff’s base composition data. carbon atoms of the sugar groups are numbered 1′ through
5′. A phosphodiester linkage connects the 5′-carbon atom
KEY CONCEPT The two strands of DNA contain complemen- of one deoxyribose to the 3′-carbon atom of the adjacent
tary base pairs—G base pairs with C and A base pairs with T.
deoxyribose. Thus, each sugar–phosphate backbone is said
to have a 5′-to-3′ polarity, or direction. Understanding this
The three-dimensional structure derived by Watson and polarity is essential in understanding how DNA fulfills its
Crick is composed of two side-by-side chains (“strands”) of roles. In double-stranded DNA, the two backbones are in
nucleotides twisted into the shape of a double helix with 10 opposite, or antiparallel, orientation; one is oriented 5′-to-
base pairs in each complete turn of the helix (Figure 7-9a). 3′ and the other is oriented 3′-to-5′ (Figure 7-9b).
DNA is a right-handed helix; in other words, it has the same
structure as that of a screw that would be screwed into place
using a clockwise turning motion. The two strands are held KEY CONCEPT The base-paired strands of DNA are oriented
together by hydrogen bonds between purine and pyrimidine antiparallel to one another—one strand is oriented in the 5′-to-
3′ direction and the other strand is oriented in the 3′-to-5′
bases of each strand, forming the stairs of a spiral staircase.
direction.
On the outside of the double helix, the backbone of each
G C linkage –O O O O O–
H N
A T CH2 P
T N H O O
O NA
T A
O
O
O O
C G P CH2
–O O
O H N O O–
T A
CH2 P
O G N H N C O O
G C
N H O
A T
O
O
CH2
H
T A O O–
3′
P
C G O O
5′
FIGURE 7-9 (a) A simplified model showing the helical structure of orange). The backbones run in opposite directions. The strand on the
DNA. Horizontal sticks represent base pairs; blue ribbons represent left is oriented 5′-to-3′ from top to bottom, and the strand on the right
the sugar–phosphate backbones of the two antiparallel DNA chains. is oriented 3′-to-5′ from top to bottom. Each base pair has one purine
(b) A chemical diagram of the DNA double helix, unrolled to show base, adenine (A) or guanine (G), and one pyrimidine base, thymine (T)
the sugar–phosphate backbones (blue) and base-pair rungs (purple, or cytosine (C), connected by hydrogen bonds (red dots).
Each base is attached to the 1′-carbon atom of a deoxy- the base pairs. A single strand of nucleotides is not helical;
ribose sugar in the backbone of each strand and faces the helical shape of DNA depends entirely on pairing and
inward toward a base on the other strand. Hydrogen bonds stacking of bases in the antiparallel strands. The most stable
between pairs of purine and pyrimidine bases (indicated by form that results from base stacking is a double helix with
dots in Figure 7-9b) hold the two strands of DNA together. two distinct sizes of grooves running in a spiral: shallow
Note that G–C base pairs have three hydrogen bonds, major grooves occur where the sugar–phosphate backbones
whereas A–T base pairs have only two. We would predict are far apart, and deep minor grooves occur where they are
that DNA containing many G–C base pairs would be more close together. Both types of grooves can be seen in ribbon
stable than DNA containing many A–T base pairs. In fact, (Figure 7-10a) and space-filling (Figure 7-10b) models. Pro-
this prediction is confirmed. Heat causes the two strands of teins that bind DNA interact specifically with either major
the DNA double helix to separate (a process called DNA or minor grooves.
melting or DNA denaturation); DNAs with higher G + C
content require higher temperatures to melt because of KEY CONCEPT The geometry of base pairs creates shallow,
wide major grooves and narrow, deep minor grooves along the
greater attraction of G–C base pairs.
DNA helix; features that are recognized for protein binding.
H
3′ Minor
groove
Sugar–phosphate O
5′
backbone
5′
P
3′ Minor
Major groove
groove
C and N
Base pairs
Major
groove
FIGURE 7-10 (a) The ribbon diagram highlights the stacking of base pairs, whereas (b) the space-filling
model emphasizes the major and minor grooves. In both models, the sugar–phosphate backbone is
blue and the bases are gold.
sequence of amino acids in the protein specified by that DNA is copied by semiconservative replication
gene. In other words, some sort of genetic code may write
information in DNA as a sequence of nucleotides and (a) Predictions of semiconservative model
then translate it into a different language of amino acid Parental 1st generation 2nd generation
sequence in protein. Just how this is done is the subject of
Chapter 9.
2. As Watson and Crick stated in the concluding words of
their 1953 Nature paper that reported the double-helical
structure of DNA: “It has not escaped our notice that the
specific pairing we have postulated immediately suggests
a possible copying mechanism for the genetic material.”2 14N/14N
(light)
To geneticists at the time, this statement proposed that
14N/15N
DNA is replicated by a semiconservative mechanism, as
(intermediate)
described in the next section.
15N/15N
3. If the nucleotide sequence of DNA specifies the amino (heavy)
acid sequence, mutations are possible by the substitu-
tion of one nucleotide for another at one or more posi-
tions. Mutations will be discussed in Chapter 16.
with semiconservative replication, when cells grown in heavy 15N (intermediate) (intermediate)
are transferred to light 14N medium, the first generation produces a 15N/15N
single DNA band of intermediate density and the second generation (heavy)
produces two bands: one intermediate and one light. (b and c) In
contrast, the data do not match results predicted for conservative and
dispersive replication.
2
J. Watson and F. Crick, Nature 171:737, 1953.
(a) Chromosome after one round of replication (b) Chromosome during second round of replication
Replication forks
FIGURE 7-12 A replicating bacterial chromosome has two replication forks. (a) Left: An autoradio-
graph of a bacterial chromosome after one generation of replication in tritiated thymidine. Right: An
interpretation of the autoradiograph. The gold helix represents the tritiated strand. In agreement with
the semiconservative model of replication, one of the two strands should be radioactive. (b) Left: An
autogradiograph of a bacterial chromosome in the second generation of replication in tritiated thymi-
dine. Right: An interpretation of the autoradiograph. Replication forks occur at the two sites of ongoing,
bidirectional DNA replication. Once again, in accord with the semiconservative model of replication, the
newly replicated double helix that crosses the circle consists of two radioactive strands (if the parental
strand was the radioactive one).
3 DNA rotates
to remove
Replication fork the supercoils.
SSB
FIGURE 7-13 (a) Twisted and supercoiled regions accumulate ahead of the replication fork as paren-
tal DNA strands separate for replication, analogous to what happens when a rope is separated into
individual strands. (b) Helicase encircles one DNA strand at each replication fork and uses the energy
of ATP hydrolysis to break hydrogen bonds between bases, unzipping the two strands and causing
twisting and supercoiling. SSB molecules bind the single-stranded DNA to prevent their reannealing.
A topoisomerase called DNA gyrase counteracts twisting and supercoiling by cutting the DNA strands,
allowing them to rotate, and then rejoining the strands.
prevent overwinding: they are helicases and topoisomerases, KEY CONCEPT Helicases, topoisomerases, and
respectively. single-strand-binding proteins generate and maintain
Helicases are enzymes that disrupt hydrogen bonds single-stranded DNA that is used as a template for DNA
that hold the two strands of the double helix together. The replication.
DNA replication helicase is a ring-shaped homohexamer
of DnaB proteins (that is, a complex of six copies of DnaB)
that encircles single-stranded DNA at the replication forks.
Assembling the replisome:
From this position, helicases use energy from ATP hydro-
lysis to rapidly unzip the double helix ahead of DNA syn- replication initiation
thesis (Figure 7-13b, step 1). Unwound DNA is stabilized Assembly of the replisome is an orderly process that begins
by single-strand DNA-binding (SSB) proteins, which bind at a precise site on the chromosome called the origin of
to single-stranded DNA and prevent the duplex from replication , or simply origin. E. coli replication begins
re-forming. from a single origin (a locus called oriC ) and then pro-
Unwinding of DNA by helicases causes extra twisting ceeds in both directions (with moving forks at both ends,
to occur ahead of replication forks, and supercoils form to as shown in Figure 7-12b) until the forks merge. OriC is
release the strain of the extra twisting (Figure 7-13b, steps 1). 245 base pairs long and contains five copies of 9-base-
Twists and supercoils must be removed (commonly called pair sequences called DnaA boxes and an adjacent DNA
relaxed) to allow replication to continue. This is done by unwinding element that is AT-rich (Figure 7-14). The first
enzymes termed topoisomerases, of which an example is step in replisome assembly is binding of a protein called
DNA gyrase (Figure 7-13b, steps 2 and 3). Topoisomerases DnaA to the DnaA boxes, which helps other copies of
relax supercoiled DNA by breaking either a single DNA DnaA bind at the origin in a process called oligomeriza-
strand or both strands, which allows DNA to rotate into tion. Subsequent binding of DnaA to the AT-rich region
a relaxed molecule. Topoisomerases finish their job by reli- promotes unwinding to form a single-stranded DNA bub-
gating the strands of the now-relaxed DNA (Figure 7-13b, ble. Recall that A–T base pairs are held together with only
step 4). two hydrogen bonds, whereas G–C base pairs are held
Initiation of DNA replication in bacteria for the initiation of replication. As replication progresses
through the origin, the replisome displaces DnaA from
AT-rich the DNA.
DNA unwinding
element DnaA boxes
KEY CONCEPT Where and when replication takes place are
carefully controlled by ordered assembly of the replisome at a
precise site called the origin.
DnaA proteins Origin (oriC) recognition
and unwinding
O O
–O P O –O P O
O O
H2C O G C H2C O G C
H H H H
H H H H
3′
HO
••
H PPi O H
–O P O
O O
O
–O P O P O P O
γ β α
O O– H2C O C G
O– O–
H2C 5′ O C G H H
H H
H H
H H HO H
3′
HO H
T T
5′ 5′
FIGURE 7-15 DNA polymerases catalyze the chain-elongation reaction. A dNTP, in this case dCTP, A N I M ATED A RT
base pairs to the single-stranded DNA template, the free 3′-OH group at the end of the growing DNA
The nucleotide polymerization
chain is activated to attack the alpha (α) phosphate of the dNTP, resulting in attachment of dNMP to the
3′ end and release of linked β and γ phosphates (PPi). Thus, the DNA chain is elongated in the 5′-to-3′ process
direction by DNA polymerases.
DNA replication is semidiscontinuous on the other template is also in the 5′-to-3′ direction, but
because it is in the direction opposite to that of replication
Another problem in DNA replication arises because
fork movement it must be in short segments. Reinitiation
DNA polymerases can extend a chain but cannot start a
of synthesis occurs for each segment as the growing fork
chain. Therefore, synthesis must be initiated by a primer,
exposes new DNA template. The 1000–2000 nucleotide
a short chain of nucleotides that forms a segment of
stretches of newly synthesized DNA are called Okazaki
duplex nucleic acid (Figure 7-16). Primers are synthesized
fragments, in honor of their discoverer Reiji Okazaki. As
by a set of proteins called a primosome, of which a cen-
with synthesis of the leading strand, each Okazaki frag-
tral component is an RNA polymerase called primase
ment is primed at the 5′ end by an RNA primer synthesized
( Figure 7-17 ). Primase copies the template DNA in the
by primase. Thus, for this strand, called the lagging strand,
5′-to-3′ direction, producing a short RNA of about 11
synthesis takes place in a discontinuous manner. Because
nucleotides. DNA pol III then takes over and continues
DNA replication is continuous for the leading strand and
to copy the template DNA, extending off the 3′ end of
discontinuous for the lagging strand, the overall process is
the RNA primer.
described as semidiscontinuous.
Because DNA polymerases synthesize DNA only in the
5′-to-3′ direction, only one of the two DNA template strands
KEY CONCEPT DNA replication is described as semidis-
can serve as a template for replication in the same direction
continuous because one DNA template strand is synthesized
as movement of each replication fork (see Figure 7-16).
continuously and the other strand is synthesized as a series of
For this strand, called the leading strand , synthesis
discontinuous fragments.
takes place in a smooth continuous manner. Synthesis
1. Primase synthesizes short RNA primers using DNA as a template. 3. DNA polymerase I removes RNA primers and fills the gap.
3′
5′ RNA primer
3′ 5′ 5′
3′
3′ 5′ 3′
5′
2. DNA polymerase III synthesizes DNA, starting at the 3′ end 4. DNA ligase connects adjacent DNA fragments.
of RNA primers.
FIGURE 7-16 DNA replication takes place at the replication fork, where the double helix is unwound
and the two strands are separated. DNA replication proceeds continuously in the direction of the
unwinding replication fork for the leading strand. In contrast, DNA is synthesized in short segments in
the direction away from the replication fork for the lagging strand. DNA polymerase requires a primer, a
short chain of nucleotides, to be in place to begin synthesis. Additional details are provided in the text.
A different DNA polymerase, DNA pol I, removes the RNA primer is more likely than DNA to contain errors.
the RNA primers with its 5′-to-3′ exonuclease activity The need to maintain the high fidelity of replication is one
and fills in the gaps with its 5′-to-3′ polymerase activ- reason why RNA primers at the ends of Okazaki fragments
ity (Figure 7-17). As mentioned earlier, DNA pol I is the must be removed and replaced with DNA. Only after the
enzyme originally purified by Kornberg. Another enzyme, RNA primer is gone does DNA pol I catalyze DNA syn-
DNA ligase, joins the 3′ end of the gap-filling DNA to the thesis to replace the primer. Mismatches that escape proof-
5′ end of the downstream Okazaki fragment. In general, reading are corrected by DNA repair mechanisms that will
DNA ligases join broken pieces of DNA by catalyzing be covered in detail in Chapter 15.
formation of a phosphodiester bond between a 5′ phos-
phate of one fragment and a 3′ OH group of an adjacent KEY CONCEPT DNA polymerases I and III have proofreading
fragment. activity but primase does not.
Origin of replication
Leading strand Lagging strand
Overall direction
of replication
Leading strand
template
DNA
3′
polymerase III 5′
Leading strand Continuous synthesis
Parental
DNA SSB
β clamp
5′ 3′ Replication
3′ 5′ fork Okazaki Okazaki Okazaki
β clamp fragment #3 fragment #2 fragment #1
DNA Helicase 3′
gyrase
Primase 5′ 3′ 5′
RNA 3′
primer 5′
Lagging Lagging strand
strand discontinuous
template DNA synthesis DNA DNA
polymerase III polymerase I ligase
O– O–
OH –O P O DNA O P O
ligase
O O
FIGURE 7-17 DNA replication occurs in both directions. The same protein factors are required for A N I M ATED A RT
synthesis of both the leading and lagging strands, but synthesis occurs in a single continuous stretch for
Leading and lagging strand
the leading strand and in short stretches for the lagging strand. The β clamp is required for processive
DNA synthesis. Helicase separates the strands of parental DNA and gyrase (a topoisomerase) removes synthesis
the twists and supercoils in DNA. Single-strand DNA-binding proteins (SSBs) prevent the separated DNA
strands from reannealing. Primase makes short RNA primers for continuous synthesis by DNA pol III of
the leading strand as well as discontinuous synthesis of the lagging strand as Okazaki fragments. DNA pol
I removes RNA primers and fills in the resulting gaps. Lastly, DNA ligase joins the DNA fragments together.
of cells—replication, transcription, and translation, for proteins (not visible in Figure 7-19) form a connection that
example—are carried out by large multisubunit complexes bridges the two catalytic cores, thus coordinating synthesis
has changed the way that we think about cells. To begin to of the leading and lagging strands.
understand why, let’s look at the replisome more closely.
Some of the interacting components of the replisome KEY CONCEPT A molecular machine called the replisome
in E. coli are shown in Figure 7-19. At the replication fork, carries out DNA synthesis. It includes two DNA polymerase
the catalytic core of DNA pol III is part of a much larger units to handle synthesis on each strand, and it coordinates
complex, called the DNA pol III holoenzyme, which con- the activity of accessory proteins required for unwinding the
sists of two catalytic cores and several accessory proteins. double helix, stabilizing the single strands and processing
RNA primers.
One of the catalytic cores handles synthesis of the leading
strand while the other handles the lagging strand. The lag-
ging strand is shown looping around so that the replisome Attachment of DNA pol III to the DNA template is
can coordinate the synthesis of both strands and move in maintained by other accessory proteins, the β clamp (also
the direction of the replication fork. Some of the accessory known as the sliding clamp), which encircles the DNA like a
FIGURE 7-18 DNA polymerases I and III use their 3′-to-5′ exonucle-
Proofreading removes mispaired bases
ase activity to remove the A–C mismatch.
DNA polymerase I and III
A
5′ G 3′
T G G A C T donut, and the clamp loader (also called τ complex), which
A C C T G A C G G assembles β clamps onto DNA. The β clamp transforms
3′ 5′
DNA pol III from an enzyme that can add only 10 nucleo-
tides before falling off the template (termed a distributive
Extension:
incorrect base enzyme) into an enzyme that stays at the moving fork and
A bonded to C adds tens of thousands of nucleotides (termed a processive
enzyme). In sum, through the action of accessory proteins,
synthesis of both the leading and lagging strands is rapid
5′ 3′ G
T G G A C T A and highly coordinated.
A C C T G A C G G Note that primase, the enzyme that synthesizes the
3′ 5′
RNA primer, is not touching the clamp protein. Therefore,
primase acts as a distributive enzyme—it adds only a few
Proofreading:
incorrect base ribonucleotides before dissociating from the template. This
detected mode of action makes sense because primers only need to
and removed
be long enough to form a suitable duplex starting point for
A
DNA pol III.
5′ 3′ G
T G G A C T
KEY CONCEPT The β clamp converts DNA polymerase III
A C C T G A C G G
3′ 5′ from a distributive to a processive enzyme.
Extension:
correct base
G added
FIGURE 7-19 A dimer of DNA pol III enzymes coordinates replica-
tion of the leading and lagging DNA strands. Looping of the template
for the lagging strand orients it for synthesis by DNA pol III in the
5′ 3′
T G G A C T G 5′-to-3′ direction. DNA pol III releases the lagging strand template
A C C T G A C G G after synthesizing 1000–2000 nucleotides, a new loop is formed,
3′ 5′ and primase synthesizes an RNA primer to initiate another Okazaki
fragment.
Looping permits dimeric DNA polymerase III to simultaneously replicate both strands
β clamp Leading
strand
5′
Replication
fork movement 3′
5′ Primase
DNA polymerase III dimer
3′
DNA Helicase
gyrase
RNA primer
Next Okazaki
fragment will
start here.
3′
5′
SSB
7.5 DNA REPLICATION IN replicate the much larger eukaryotic genomes. Approxi-
mately 400 replication origins are dispersed throughout
EUKARYOTES the 16 chromosomes of yeast, and humans have 40,000
to 80,000 origins among the 23 chromosomes. Thus, in
LO 7.6 Explain why and how DNA replication differs
eukaryotes, replication proceeds in both directions from
between bacteria and eukaryotes.
multiple points of origin (Figure 7-20). Double helices that
are produced at each origin elongate and eventually join
DNA replication in both bacteria and eukaryotes uses a one another. When replication of the two strands is com-
semiconservative mechanism and employs leading and lag- plete, two identical daughter molecules of DNA result.
ging strand synthesis. For this reason, it should not come
as a surprise that components of the bacterial and eukary- KEY CONCEPT Replication proceeds in both directions
otic replisomes are very similar (see Table 7-2). However, from hundreds or thousands of origins on linear eukaryotic
because eukaryotic genomes are larger and have linear, chromosomes.
not circular, chromosomes, there are additional mechanistic
complexities and associated factors. Bacteria such as E. coli
usually complete replication in about 40 minutes, but
in eukaryotes, the amount of time to complete replication DNA replication and the yeast cell cycle
can vary from a few minutes to many hours depending on
DNA synthesis takes place only in S (synthesis) phase of
many features, including the size of the genome, the num-
the eukaryotic cell cycle (Figure 7-21). How is the onset of
ber of origins, and the type of cell. Eukaryotes also have to
DNA synthesis limited to this single phase? In yeast, the
solve the problem of coordinating the replication of more
method of control is to link replisome assembly to the cell
than one chromosome.
cycle. Figure 7-22 shows the process. In yeast, three proteins
are required to begin assembly of the replisome. The origin
Eukaryotic origins of replication recognition complex (ORC) first binds to sequences in
To understand eukaryotic replication origins, we will yeast origins, much as DnaA protein does in E. coli. ORC
first turn our attention to the simple eukaryote yeast then acts as a landing pad to recruit Cdc6 to origins early
(Saccharomyces cerevisiae). Many eukaryotic proteins hav- in gap 1 (G1) phase of the cell cycle. Together, ORC and
ing roles at replication origins were first identified in yeast Cdc6 then load a complex of Cdt1 and helicase. A second
because of the ease of genetic analysis (see the yeast Model helicase-Cdt1 complex is recruited through association
Organism box in Chapter 12 ). Origins of replication in with the already assembled helicase-Cdt1 complex. Once
yeast are referred to as autonomously replicating sequences the helicases are on the DNA in early S phase, Cdc6 and
(ARSs) and are very much like oriC in E. coli. ARSs are Cdt1 are released and DNA polymerases are loaded onto
about 100 to 200 base pairs long and contain several con- the DNA. Replication is linked to the cell cycle through
served DNA sequence elements, including an AT-rich ele- the availability of Cdc6 and Cdt1. In yeast, these proteins
ment that melts when an initiator protein binds to adjacent are synthesized during late mitosis (M) and G1 and are
elements. Unlike bacterial chromosomes, each eukary- destroyed by proteolysis at the beginning of S phase. In this
otic chromosome has many replication origins to quickly way, the replisome can only be assembled before S phase.
Sister chromatids
DNA copies (daughter molecules)
2n 11-bp consensus
2n sequence
AT-rich
Original 4n Daughter
2n 2n
cell cells
ORC
Origin recognition
Phases of the cell cycle
(4n) (2n)
M = mitosis
S = DNA synthesis M
G1 = gap 1 ORC
G2 = gap 2
M phase
(4n) G2 G1 (2n)
Loading of Cdc6
Cdc6
S
(4n) (2n)
FIGURE 7-21 The cell cycle is composed of G1, S, G2, and Recruitment of
M phases. DNA replication occurs during S phase, and chromosome second helicase/Cdt1
segregation and cell division occur during M phase. Numbers in
parentheses indicate the ploidy (that is, the number of sets of chro-
mosomes) at each phase of the cell cycle in a diploid organism.
G1 phase
Once replication has begun, new replisomes cannot form
at origins, because Cdc6 and Cdt1 are no longer available.
Unwinding of helix
KEY CONCEPT DNA replication in eukaryotes requires Cdc6 and sliding of helicase
and Cdt1, proteins that are only available during late mitosis
(M) and G1 phase, ensuring that the genome is only replicated
once per cell cycle. Early in
S phase
Recruitment of
Replication origins in higher eukaryotes DNA polymerase
As already stated, most of the approximately 400 origins of
replication in yeast are composed of similar DNA sequence
motifs (100–200 base pairs in length) that are recognized S phase
by ORC subunits. Although all characterized eukary-
otes have similar ORC proteins, origins of replication in
higher eukaryotes such as humans are much longer, possi-
bly as long as tens of thousands or hundreds of thousands FIGURE 7-22 This eukaryotic example from yeast shows initiation
of base pairs. Significantly, they have limited sequence of DNA synthesis at an origin of replication. Similar to initiation in
similarity. Thus, although yeast ORC recognizes specific bacteria (see Figure 7-14), sequences at the origin of replication are
DNA sequences in yeast chromosomes, what the related bound by a protein complex, in this case the origin recognition com-
plex (ORC), which recruits helicases to separate the two strands of
ORCs of higher eukaryotes recognize is not clear at this the double helix. The cell cycle-regulated factors Cdc6 and Cdt1 are
time, but the feature recognized is probably not a specific essential for recruitment of the helicases, thereby linking DNA replica-
DNA sequence. In practical terms, this uncertainty means tion to the cell cycle.
that it is much harder to isolate origins from humans and The DNA replication problem at chromosome ends
other higher eukaryotes, because scientists cannot use an
Origin of replication
isolated DNA sequence of one human origin, for example,
to perform a computer search of the entire human genome
sequence to find other origins.
If the ORCs of higher eukaryotes do not interact with Lagging Leading
3′ strand strand 5′
a specific sequence scattered throughout the chromosomes,
then how do they find the origins of replication? These 5′ Leading Lagging 3′
ORCs are thought to interact indirectly with origins by strand strand
Replication
associating with other protein complexes that are bound fork
to chromosomes. Such a recognition mechanism may have Primer
evolved so that higher eukaryotes can regulate the timing Leading strand Lagging strand
3′ 5′
of DNA replication during S phase. Gene-rich regions of 5′ 3′
the chromosome (the euchromatin) have been known for Primer degraded
some time to replicate early in S phase, whereas gene-poor Internal Terminal
regions, including the densely packed heterochromatin, gap gap
replicate late in S phase (see Chapter 12 for more about
3′ 5′
euchromatin and heterochromatin). 5′ 3′
All internal gaps filled,
KEY CONCEPT Yeast origins of replication, like origin in bac- terminal gap not filled
teria, contain a conserved DNA sequence that is recognized 3′ 5′
by ORC and other proteins needed to assemble the replisome. 5′ 3′
In contrast, origins of higher eukaryotes have been difficult to
3′ overhang
isolate and study because they are long and complex and do
not contain a conserved DNA sequence. FIGURE 7-23 Top: The replication of each Okazaki fragment for
the lagging strand begins with a primer. Bottom: The fate of the
bottom strand in the transcription bubble. When the primer for the
last Okazaki fragment of the lagging strand is removed, there is no
way to fill the gap by conventional replication. A shortened chro-
Telomeres and telomerase: replication mosome would result when the chromosome containing the gap is
termination replicated.
DNA polymerase
fills in the gap
In addition to preventing the erosion of genetic mate-
DNA polymerase
rial after each round of replication, telomeres preserve
chromosomal integrity by associating with proteins such 3′ AACCC CAAC C CCAAC CCCAACCCCAAC 5′
as WRN, TRF1, and TRF2, to form a protective structure 5′ 3′
called a telomeric loop (t-loop) (Figure 7-25). These struc-
tures sequester the 3′ single-stranded overhang, which can
The primer is removed
be as much as 100 nucleotides long. Without t-loops, the and DNA ligase seals the gap
ends of chromosomes would be mistaken for double-strand DNA ligase
breaks by the cell and dealt with accordingly. As you
3′ AACCC CAACCCCAAC 5′
will see in Chapter 15, double-strand breaks are poten-
5′ 3′
tially very dangerous because they can result in chromo-
somal instability that can lead to cancer and a variety of
261
TRF2
TRF1
5′
3′
(b)
FIGURE 7-26 A woman with Werner syndrome at ages 15 (left)
and 48 (right). [International Registery of Werner Syndrome, www
.wernersyndrome.org.]
SUMMARY
Experimental work on the molecular nature of hereditary RNA primer (synthesized by primase) that provides a 3′ end
material has demonstrated conclusively that DNA (not for deoxyribonucleotide addition.
protein, lipids, or carbohydrates) is the genetic material. The multiple events that have to occur accurately and
Using data obtained by others, Watson and Crick deduced a rapidly at the replication fork are carried out by a biological
double-helical model with two DNA strands, wound machine called the replisome. This protein complex includes
around each other, running in antiparallel fashion. Binding two DNA polymerase units, one to produce the leading
of the two strands together is based on the fit of adenine (A) strand and the other to produce the lagging strand. In this
to thymine (T) and guanine (G) to cytosine (C). The former way, the more complex synthesis and joining of Okazaki
pair is held by two hydrogen bonds; the latter, by three. fragments into a continuous strand can be temporally coor-
The Watson–Crick model shows how DNA can be dinated with the less complex synthesis of the leading strand.
replicated in an orderly fashion—a prime requirement for Where and when replication takes place is carefully con-
genetic material. Replication is accomplished semiconser- trolled by the ordered assembly of the replisome at certain
vatively in both bacteria and eukaryotes. One double helix sites on chromosomes called origins. Eukaryotic genomes can
is replicated to form two identical helices, each with their have tens of thousands of origins. Assembly of replisomes at
nucleotides in the identical linear order; each of the two origins takes place only at a specific time in the cell cycle.
new double helices is composed of one old and one newly The ends of linear chromosomes (telomeres) in eukary-
polymerized strand of DNA. otes present a problem for the replication system because
The DNA double helix is unwound at a replication fork, there is always a short stretch on one strand that cannot
and the two single strands serve as templates for polymer- be primed. The enzyme telomerase adds numerous short,
ization of free nucleotides. Nucleotides are polymerized by repetitive sequences to maintain the length of telomeres.
the enzyme DNA polymerase, which adds new nucleotides Telomerase carries a small RNA that acts as the template
only to the 3′ end of a growing DNA chain. Because addi- for synthesis of telomeric repeats. These noncoding telo-
tion is only at 3′ ends, polymerization on one template is meric repeats associate with proteins to form a telomeric
continuous, producing the leading strand; and on the other, loop that protects against DNA damage. Telomeres shorten
it is discontinuous in short stretches (Okazaki fragments), with age because telomerase is not produced in somatic
producing the lagging strand. Synthesis of the leading cells. Individuals who have defective telomeres experience
strand and of every Okazaki fragment is primed by a short premature aging.
KEY TERMS
adenine (p. 244) DNA pol III holoenzyme primer (p. 254)
antiparallel (p. 247) (p. 256) primosome (p. 254)
bases (p. 244) DNA replication (p. 240) processive enzyme (p. 257)
β-clamp (sliding clamp) (p. 256) double helix (p. 240) purine (p. 244)
cell cycle (p. 258) genetic code (p. 249) pyrimidine (p. 244)
complementary bases (p. 247) guanine (p. 244) radioisotope (p. 243)
conservative replication (p. 249) helicase (p. 252) replication fork (p. 250)
cytosine (p. 244) lagging strand (p. 254) replisome (p. 240)
daughter molecule (p. 249) leading strand (p. 254) reverse transcriptase (p. 261)
deoxynucleotide (p. 244) major groove (p. 248) ribose (p. 244)
deoxyribose (p. 244) minor groove (p. 248) semiconservative replication (p. 249)
dispersive replication nucleoside (p. 250) semidiscontinuous (p. 254)
(p. 249) nucleotide (p. 240) single-strand DNA-binding (SSB)
distributive enzyme (p. 257) Okazaki fragment (p. 254) protein (p. 252)
DNA gyrase (p. 252) origin of replication (origin) (p. 252) telomerase (p. 260)
DNA ligase (p. 255) Origin recognition complex (ORC) telomere (p. 260)
DNA polymerase I (DNA pol I) (p. 258) telomeric loop (t-loop) (p. 261)
(p. 253) parental molecule (p. 249) template (p. 240)
DNA polymerase III (DNA pol III) phosphate (p. 244) thymine (p. 244)
(p. 253) primase (p. 254) topoisomerase (p. 252)
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the www
WORKING WITH THE FIGURES 17. In Figure 7-16, draw an analogous diagram for the
(The first 27 questions require inspection of text figures.) other replication fork.
1. In Table 7-1, complete the table for a genome that is 18. In Figure 7-17, which factors are involved in lagging
20 percent adenine. strand synthesis but not leading strand synthesis?
2. In Figure 7-1, what features of the sculpture of DNA 19. In Figure 7-18, draw the phosphodiester linkage between
are correct or incorrect? T and the misincorporated A in the strand being synthe-
sized, and place an arrow at the bond that is broken by
3. In Figure 7-2 , speculate as to why Griffith did not
the 3′-to-5′ exonuclease activity of DNA polymerase.
conduct the experiment the other way around, that is,
with heat-killed R cells and live S cells. 20. In Figure 7-19, why is the DNA looped for one strand
but not for the other strand when they are both
4. In Figure 7-3 , what types of enzymes could Avery,
serving as templates for DNA synthesis by the DNA
Macleod, and McCarty have used to destroy proteins
polymerase III dimer?
and RNAs?
21. In Figure 7-20 , could the spacing of origins affect
5. In Figure 7-4 , what part of the DNA structure is
the amount of time that it takes to replicate a
labeled by 32P?
chromosome?
6. In Figure 7-5 , draw 7-methylguanine and 5-
22. In Figure 7-21, how much DNA would a cell contain if
methylcytosine. A methyl group is CH3.
it went through two cell cycles that did not include an
7. In Figure 7-6 , what information did Rosalind M phase?
Franklin’s X-ray diffraction data provide that was key
23. In Figure 7-22, why does replication not initiate at ori-
to determining the structure of DNA?
gins in G2 phase?
8. In Figure 7-7 , why do you think that Watson and
24. In Figure 7-23, analogous to the last diagram in the
Crick built a three-dimensional model of DNA rather
figure, draw the top DNA strand in the bubble after
than only a two-dimensional model?
the primers are degraded and the gaps are filled. Based
9. In Figure 7-8, would the diameter of DNA change if on this drawing, is telomerase required for the replica-
the pyrimidine was on the left and the purine on the tion of both ends of chromosomes?
right? Justify your answer.
25. In Figure 7-24a, the telomerase RNA template contains
10. In Figure 7-9a, why are there two rows of dots between one and a half copies of the repeat sequence. Circle the
A–T base pairs but three rows between G–C base pairs? full copy and put a box around the half copy.
11. In Figure 7-10a, is a purine or a pyrimidine on the left 26. In Figure 7-25b, fluorescent antibodies to what protein
in the bottom base pair? may have been used to detect the telomeres?
12. In Figure 7-11, draw cesium chloride gradients for a 27. In Figure 7-26 , if Figure 7-25b represents chromo-
Meselson–Stahl experiment in which cells are first somes from this individual at a young age, how might
grown in 14N and then in 15N for two generations. the image in Figure 7-25b differ at the older age?
13. In Figure 7-12, draw an autoradiograph for a chromo-
some during the second round of replication in which BASIC PROBLEMS
the DNA that crosses the circle has one blue parental
28. Does the Hershey-Chase experiment definitively
strand.
demonstrate that DNA is the genetic material, or just
14. In Figure 7-13a , what would happen in the rope that it is consistent with being the genetic material?
demonstration if you cut one of the two strands in the Justify your answer.
supercoiled region.
29. Does the Avery, MacLeod, and McCarty experiment
15. In Figure 7-14, in the second to last diagram, why do definitively demonstrate that DNA is the genetic mate-
the arrows show the two helicase molecules moving in rial or just that it is consistent with being the genetic
opposite directions? material? Justify your answer.
16. In Figure 7-15, draw the chemical reaction that occurs 30. Draw a cesium chloride gradient for the third genera-
to add the next nucleotide in the DNA chain. tion of DNA produced by semiconservative replication
in the Meselson–Stahl experiment.
31. Write the sequence of the telomerase RNA that 45. Would the Meselson–Stahl experiment have worked if
serves as a template for the telomere repeat sequence diploid eukaryotic cells had been used instead?
5′-TTAGGG-3′. 46. Consider the following segment of DNA, which is part
32. Why might Werner syndrome increase the chances of of a much longer molecule constituting a chromosome:
getting cancer?
5′ . . . ATTCGTACGATCGACTGACTGACAGTC . . . 3′
33. Draw 2′,3′ dideoxyadenosine and predict what would
happen if this nucleotide was incorporated into the 3′ . . . TAAGCATGCTAGCTGACTGACTGTCAG . . . 5′
growing DNA chain during replication. If the DNA polymerase starts replicating this segment
34. Explain how DNA fulfills the three main requirements from the right,
for a hereditary molecule: (1) the ability to store infor- a. which will be the template for the leading strand?
mation, (2) the ability to be replicated, and (3) the
ability to mutate. b. draw the molecule when the DNA polymerase is
halfway along this segment.
35. Match the protein with its function.
c. draw the two complete daughter molecules.
A. DNA polymerase creates RNA primers
47. The DNA polymerases are positioned over the fol-
B. Helicase links short DNA chains
lowing DNA segment (which is part of a much larger
C. Ligase helps hold polymerase on DNA molecule) and moving from right to left. If we assume
D. Primase separates DNA strands that an Okazaki fragment is made from this segment,
E. Gyrase prevents reannealing of DNA what will be the fragment’s sequence? Label its 5′ and
3′ ends.
F. Sliding clamp extends DNA strand
G. SSB removes supercoils in DNA 5′ . . . CCTTAAGACTAACTACTTACTGGGATC . . . 3′
36. Why is telomerase not required for replication of the 3′ . . . GGAATTCTGATTGATGAATGACCCTAG . . . 5′
bacterial genome?
37. Explain what is meant by the terms conservative and CHALLENGING PROBLEMS
semiconservative replication.
48. If you extract the DNA of the coliphage φX174, you
38. Describe two pieces of evidence indicating that DNA will find that its composition is 25 percent A, 33 per-
polymerase I is not the chromosomal replicase. cent T, 24 percent G, and 18 percent C. Does this
39. What is meant by a primer, and why are primers nec- composition make sense in regard to Chargaff’s rules?
essary for DNA replication? How would you interpret this result? How might such
a phage replicate its DNA?
40. A molecule of composition
49. Given what you know about the structure and func-
5′-AAAAAAAAAAAAA-3′
tion of telomerase, provide a plausible model to
3′-TTTTTTTTTTTTT-5′ explain how a species could exist with a combination
is replicated in a solution containing unlabeled (not of two different repeats (for example, TTAGGG and
radioactive) dGTP, dCTP, and dTTP plus dATP with all TTGTGG) on each of their telomeres.
its phosphorus atoms in the form of the radioactive iso- 50. Why is it unlikely that continuous replication of both
tope 32P. Will both daughter molecules be radioactive? DNA strands occurs but is yet to be discovered?
Explain. Then repeat the question for the molecule
5′-ATATATATATATAT-3′
GENETICS AND SOCIETY
3′-TATATATATATATA-5′
In this chapter, you learned that the shortening of chro-
41. Why is DNA synthesis continuous on one strand and mosome telomeres due to diminished telomerase activity
discontinuous on the opposite strand? is associated with aging. This raises the possibility that
42. Explain why cutting one strand of supercoiled DNA gene therapy aimed at overexpression of telomerase will
removes the supercoiling. increase longevity. Do you think that it is ethical to use this
43. Describe how the enzymatic activities of DNA poly- approach to increase the longevity of normal, healthy peo-
merases I and III are similar and different. ple? Does your answer change if you consider that there
are nongenetic means such as calorie restriction that may
44. If the GC content of a DNA is 48 percent, what are the increase longevity, or that gene therapy is being pursued to
percentages of the four bases (A, T, G, and C) in this treat numerous diseases?
molecule? www
www
RNA: Transcription,
Processing, and Decay
267
I
n this chapter, we describe how the information stored in eliminated from cells by decay mechanisms that disassem-
DNA is transferred to RNA. The key event in this transfer ble RNAs into individual nucleotides. Figure 8-1 provides
is transcription, which copies the information from one an overview of the chapter by illustrating the timeline
strand of DNA into a strand of RNA. In bacteria, the infor- of events that occur in the life cycle of protein-coding
mation in protein-coding RNAs is almost immediately con- RNAs (mRNAs) in bacteria and eukaryotes. Every process
verted into protein by a process called translation (the focus described in this chapter relies on molecular interactions
of Chapter 9). In contrast, in eukaryotes, transcription and that are specified by nucleic acid sequences. RNAs inter-
translation are spatially separated: transcription takes place act with DNA and other RNAs by base pairing of comple-
in the nucleus and translation in the cytoplasm. Furthermore, mentary sequences, and proteins interact with DNA and
in eukaryotes, before RNAs are ready to be exported to the RNA by binding specific sequences. Therefore, mutations
Griffiths12e_08_01ab
cytoplasm for translation, they undergo extensive processing, in DNA and RNA that disrupt molecular interactions can
including deletion of internal nucleotides and addition of spe- affect the expression of proteins.
cial nucleotide structures to the 5′ and 3′ ends. In summary, this chapter focuses on the molecules and
Both bacteria and eukaryotes also produce other types mechanisms that produce and destroy RNAs. The mole-
of RNA that are not translated into protein but instead cules and mechanisms are important to geneticists because
perform a variety of roles in cells by base pairing to mutations that affect them change which proteins are
other RNAs, binding proteins, and performing enzymatic expressed, their sequence, and their abundance, and lead to
reactions. Lastly, the chapter describes how RNAs are altered phenotypes.
Nucleotides
tion
tion
on
ing
ati
Initia
Initia
3′
5′
ng
pp
de
de
n
De
io
n
Elo
Int
ca
ca
Ca
ca at
io
ern
ng
at
pp
y
al
iti
ing
o
cle
El
In
av ion
ag gat 3′ d
e
E lon eca
y
Decay Transcription cing
5′ pro
cessin Deade Spli
g nylatio
n
Cytoplasm
Transcription Editing
Nucleus
Translation
tion Decay tion Nucleoti
Termina Termina modificade
tion
Translation Processing
tio n
mina
Ter
n
io
Po
t
ga
lya
Term
tion
Nuclear export
on
de
El
Initia
ny
inatio
la
tio
n
n
FIGURE 8-1 In both (a) bacteria and (b) eukaryotes, the transfer of information from DNA to
protein involves a sequential series of molecular events dedicated to mRNA. Note the similarities
and differences in the life cycles of mRNA in bacteria and eukaryotes. Transcription, translation, and
decay of mRNA occur in both bacteria and eukaryotes. In contrast, transcription, translation, and
decay occur concurrently only in bacteria, and various mRNA processing events and nuclear export
are unique to eukaryotes.
268
KEY CONCEPT The life cycle of an mRNA in bacteria and in Volkin and Astrachan demonstrated the rapid turnover
eukaryotes includes transcription, translation, and decay. The of RNA using a protocol called a pulse–chase experiment.
life cycle of an mRNA in eukaryotes also includes RNA To conduct a pulse–chase experiment, the infected bacte-
processing and nuclear export. ria are first fed (pulsed with) radioactive uracil, a mole-
cule needed for the synthesis of RNA but not DNA. Any
RNA synthesized in the bacteria from then on is “labeled”
KEY CONCEPT Central to the transfer of information from
with the readily detectable radioactive uracil. After a short
DNA to RNA in both bacteria and eukaryotes are molecular
period of incubation, the radioactive uracil is washed away
interactions that involve base pairing of complementary nucleic
acids (DNA and RNA) and binding of proteins to specific
and replaced (chased) by uracil that is not radioactive. This
nucleic acid sequences. procedure “chases” the label out of RNA because, as the
pulse-labeled RNA breaks down, only the unlabeled uracil
is available to synthesize new RNA molecules. Volkin and
8.1 RNA STRUCTURE Astrachan found that the RNA recovered shortly after the
pulse was labeled, but RNA recovered just a few minutes
LO 8.1 Describe how the structure of RNA enables it to later was unlabeled, indicating that the RNA has a very
function differently from DNA. short lifetime in bacteria.
RNAs carry out an amazing variety of biological functions, KEY CONCEPT Unlike DNA, RNA has a high rate of turnover
including providing the information for making proteins, within cells.
regulating translation, processing RNA, and maintaining
chromosome ends. The versatility of RNA relative to DNA
A similar experiment can be done with eukaryotic cells.
is due to the ability of single-stranded RNA to form an
Cells are first pulsed with radioactive uracil and, after a
immense variety of elaborate three-dimensional structures
short time, they are transferred to medium (the liquid they
that scaffold the binding of proteins, base pair with other
grow in) with unlabeled uracil. In samples taken imme-
RNAs, and carry out enzymatic reactions. Furthermore,
diately after the pulse, most of the labeled RNA is in the
the versatility of RNA as a biomolecule is enhanced by the
nucleus. However, in samples taken a few minutes later, the
variety of ways in which RNA function can be regulated,
labeled RNA is also found in the cytoplasm (Figure 8-2).
including mechanisms that alter RNA structure, abun-
This indicates that, in eukaryotes, RNA is synthesized in
dance, and cellular localization.
the nucleus and then moves to the cytoplasm, where pro-
teins are synthesized. These data along with other data led
RNA is the information-carrying to the conclusion that RNA is the information-transfer
intermediate between DNA and proteins intermediary between DNA and protein.
Early investigators had good reason for thinking that infor-
KEY CONCEPT In eukaryotic cells, RNA can move from the
mation is not transferred directly from DNA to protein. In nucleus to the cytoplasm.
eukaryotic cells, DNA is located in the nucleus, which phys-
ically separates it from the protein
synthesis machinery in the cytoplasm.
Thus, an intermediate is needed that Eukaryotic RNA moves from the nucleus to the cytoplasm
carries the DNA sequence informa- Cytoplasm
tion from the nucleus to the cyto- Pulse with Chase with
plasm. That intermediate is RNA. radioactive nonradioactive
uracil uracil
In 1957, Elliot Volkin and Law-
rence Astrachan made an observation
suggesting that RNA was the inter-
mediate molecule. They found that Nucleus RNA is synthesized Newly synthesized RNA is
one of the most striking molecular in the nucleus exported to the cytoplasm
changes that takes place when E. coli
is infected with the bacteriophage T2 FIGURE 8-2 The pulse–chase experiment showed that RNA moves from the nucleus to the
cytoplasm in eukaryotic cells. Cells are grown briefly in medium with radioactive uracil to label
is a rapid burst of RNA synthesis. Fur- newly synthesized RNA (pulse). Cells are washed to remove the radioactive uracil and then grown
thermore, this bacteriophage-induced in medium with nonradioactive uracil (chase). The red dots indicate the location of RNAs containing
RNA “turns over”; that is, the amount radioactive uracil over time. The location of radioactive RNAs was not determined by microscopy;
of time it spends in the cell is brief, on rather, it was inferred using a molecular approach. Cells were collected immediately after the
the order of minutes. Its rapid appear- pulse as well as after the chase and broken open, nuclei were separated from the cytoplasm
by centrifugation, RNA was purified from the nuclear and cytoplasmic fractions (leaving behind
ance and disappearance suggested that radioactive uracil that was not incorporated into RNA), and the amount of radioactivity in the nuclear
RNA might play some role in the syn- and cytoplasmic fractions was determined using a scintillation counter (an instrument that measures
thesis of more T2 phage particles. ionizing radiation).
Consequences of the distinct chemical DNA and can form a greater variety of three-dimensional
properties of RNA structures. Base pairing between regions within an RNA
(i.e., intramolecular base pairing) is an important deter-
Although both RNA and DNA are nucleic acids, the build- minant of RNA structure. For example, the stem-loop is
ing blocks of RNA differ from those of DNA in three the fundamental structural element of RNA (Figure 8-3d).
important ways that allow RNA to have greater structural Stem-loops are made up of a double-stranded stem of
and functional diversity. complementary regions of an RNA and a single-stranded
1. RNA has ribose sugar in its nucleotides, rather than the loop at the end of the stem.
deoxyribose found in DNA (Figure 8-3a). As the names sug-
gest, the sugars differ in the presence or absence of just one KEY CONCEPT Unlike DNA, RNA contains 2′-hydroxyls on the
oxygen atom at the 2′ carbon. The 2′-OH in RNA reduces ribose sugars, uracil replaces thymine, and it is single-stranded
the stability of single-stranded RNA relative to single- but base pairs to itself to form double-stranded regions.
stranded DNA. RNA cleavage can occur when a deproton-
ated 2′-OH acts as a nucleophile at a nearby phosphodi- As exemplified by the terms ribose sugar and uracil
ester bond. Furthermore, the 2′-OH provides an extra site base, there is specific terminology for describing the build-
for hydrogen bonding as well as an extra site for chemical ing blocks of RNA. There are terms for each of the four
modification such as methylation. Properties of the 2′-OH RNA nucleobases (i.e., the base itself), the nucleosides (i.e.,
contribute to the ability of RNAs, which are u sually single- the base with a ribose sugar), and the nucleotides (i.e., the
stranded, to fold into complex three-dimensional structures. base with a ribose sugar and one, two, or three phosphates)
2. RNA contains the pyrimidine base uracil (U), instead of (Figure 8-4). For example, uracil is a nucleobase, uridine is a
thymine (T) found in DNA (Figure 8-3b). Uracil forms nucleoside, and uridine triphosphate is a nucleotide.
two hydrogen bonds with adenine just like thymine does
in DNA (Figure 8-3c). In addition, uracil is capable of Classes of RNA
base pairing with G in helices of a folded RNA or between RNA molecules can be grouped into two general classes.
two separate RNAs, but not with G in DNA during tran- One class of RNA is messenger RNA (mRNA) because,
scription. The ability of U to base pair with both A and G like a messenger, it serves as an intermediary that carries
is a major reason why RNA can form intricate structures. information. The information from DNA is transferred to
Throughout this chapter, we revisit the chemical and mRNA through the process of transcription, and mRNA
structural properties of the ribose 2′-OH and the uracil passes that information on to proteins through the process
base because they are critical to the folding, function, of translation. The other class of RNA is noncoding RNA
and recognition of RNAs in RNA processing events. (ncRNA) because it does not encode proteins. Instead, the
3. RNA is usually single-stranded, not double-stranded like ncRNA is the final product whose function is determined
DNA. As a consequence, RNA is much more flexible than by its sequence and three-dimensional structure.
NH2 O NH2 O
N N N NH N NH
Base N N O O
N N NH2 N N
H H H H
Adenine Guanine Cytosine Uracil
NH2 O
NH2 O
N N N NH N NH
N N N N N O N O
NH2
H H H H H H H H
H H H H H H H H
OH OH OH OH OH OH OH OH
NH2 O NH2 O
N N N NH N NH
N N N N NH2 N O N O
Nucleotide H H H H H H H H
H H H H H H H H
OH OH OH OH OH OH OH OH
FIGURE 8-4 Each of the four building blocks of RNA has a distinct name for its nucleobase,
nucleoside, and nucleotide.
ncRNAs vary in length and function, although they RNAs, about 21 nucleotides in length, that suppress the
often act similarly by base pairing with another RNA, by expression of genes and help maintain genome stability.
serving as a scaffold for binding proteins, or by catalyzing
a chemical reaction. In both bacteria and eukaryotes, some KEY CONCEPT There are two general classes of RNAs,
ncRNA such as transfer RNAs (tRNAs) and r ibosomal those that encode proteins (mRNA) and those that do not
RNAs (rRNAs) function in translation. Bacteria and encode proteins (ncRNA). ncRNAs participate in a variety
eukaryotes also express a variety of ncRNAs that are about of cellular processes, including protein synthesis (tRNA and
rRNA), RNA processing (snRNA), the regulation of gene
50–300 nucleotides in length and function at multiple
expression (siRNA and miRNA), and genome defense (siRNA).
steps in gene expression. In eukaryotes, ncRNAs are cat-
egorized based on their location in the cell: small nuclear
RNAs (snRNAs), small nucleolar RNAs (snoRNAs), and
small cytoplasmic RNAs (scRNAs). The nucleolus is a non– 8.2 TRANSCRIPTION AND DECAY
membrane-bound region in the nucleus where ribosomes OF mRNA IN BACTERIA
are produced. Eukaryotes also express long n oncoding
RNAs (lncRNAs) that are typically greater than 300 nucle- LO 8.2 Explain how RNA polymerases are directed to begin
otides in length. Thousands of lncRNAs have been identi- and end transcription at specific places in genomes.
fied in humans, but only a few have been assigned functions,
and these are mostly regulators of gene expression. Some The first step in the transfer of information from DNA
eukaryotes also encode microRNAs (miRNAs) and gener- to protein is to produce an RNA strand whose nucleo-
ate small interfering RNAs (siRNAs) and other very small tide sequence matches the nucleotide sequence of a DNA
segment. Because this process is reminiscent of transcrib- Overview: DNA as transcription template
ing (copying) written words, the synthesis of RNA is Information encoded in DNA is transferred to the RNA
called transcription. The DNA is said to be transcribed transcript by the complementary pairing of DNA and
into RNA, and the RNA is called a transcript. Volkin and RNA bases. Consider the transcription of a chromosomal
Astrachan showed that RNA is transcribed and degraded segment that constitutes a gene. First, the two strands of
rapidly within the cell, and later experiments showed that the DNA double helix separate locally to form a transcrip-
the abundance of a given RNA is regulated by controlling tion bubble. One of the separated strands acts as a tem-
its rate of transcription and its rate of decay. These pro- plate for RNA synthesis and is called the template strand
cesses, though chemically simple, are controlled by a vari- (or noncoding strand) and the other strand is called the
ety of factors. non-template strand (or coding strand) (Figure 8-5a). The
H H H H
H H H H
O OH O OH
O P O− O P O−
Base Base
O CH2 O O CH2 O
H H O O H H
H H −O P O P O H H
RNA polymerase 3′
OH OH O− O− O OH
O O O
Base O P O−
−O Base
P O P O P O CH2 O
5′ O CH2 O
O− O− O− H H
H H
H H
H H
OH OH
OH OH
RNA
polymerase I
DNA
rRNAs
DNA
resulting RNA sequence is complementary to the tem- from DNA to RNA: base complementarity and protein-
plate strand and identical (except for the use of uracil in nucleic acid interactions.
place of thymine) to the non-template strand. When DNA
sequence is cited in the scientific literature, the sequence of KEY CONCEPT RNA is transcribed in the 5′-to-3′ direction
the non-template strand is almost always given because it is using a single-stranded DNA template oriented in the 3′-to-5′
the same as the RNA sequence. direction. Thus, RNAs start with a 5′-triphosphate (5′-ppp) and
end with a 3′ hydroxyl (3′-OH).
KEY CONCEPT RNA sequence is complementary to the
template strand and is the same as the coding (non-template) KEY CONCEPT One gene can be transcribed by multiple
strand, except it contains U in place of T. RNA polymerase molecules at the same time.
5 UTR
AUG Transcription
(a) 5
Promoter Coding sequence of the gene
+1
ATG
(b) Strong E. coli promoters
FIGURE 8-7 (a) The promoter lies “upstream” (i.e., 5′) of the transcription start site (+1) and protein- A N I M ATED A RT
coding sequences. (b) Promoters have regions of similar sequences, as indicated by the yellow
shading in seven different promoter sequences in E. coli. Spaces (dots) are inserted in the sequence Transcription in bacteria
to optimize the alignment of common sequences. Numbers refer to the number of bases before (−)
or after (+) the transcription start site.
Transcription initiation in bacteria How does RNA different genes do not have to be identical to perform a simi-
polymerase find the correct starting point for transcrip- lar function. Nonetheless, it is possible to arrive at a sequence
tion? In bacteria, RNA polymerase usually binds to a spe- of nucleotides, called a consensus sequence, that is in agree-
cific DNA sequence called a promoter, located close to the ment with most sequences. The E. coli promoter consensus
start of the transcribed region. Promoters are an impor- sequence is shown at the bottom of Figure 8-7. An RNA
tant part of the transcriptional regulatory region of a gene polymerase holoenzyme (see the next paragraph) binds to
(Figure 8-7). The first transcribed base is called the initiation the DNA at this point, then unwinds the DNA double helix
site or the transcription start site. The promoter is referred and begins the synthesis of an RNA molecule. Note in Figure
to as upstream of the initiation site because it is located 8-7 that the protein-coding part of the gene usually begins at
ahead of the initiation site (5′ of the gene). A downstream an AUG sequence in the mRNA (i.e., ATG in DNA), but the
site is located later in the direction of transcription. Nucle- transcription start site, where transcription begins, is usually
otide positions upstream of the initiation site are indicated well upstream of this sequence. The region between the start
by a negative (−) sign and those downstream by a posi- of transcription and the start of translation is referred to as
tive (+) sign. By convention, the first DNA base to be tran- the 5′ untranslated region (5′ UTR).
scribed is numbered +1. The bacterial RNA polymerase that scans the DNA for
a promoter sequence is called the RNA polymerase holoen-
KEY CONCEPT Nucleotides in genes are numbered relative zyme (Figure 8-8a). This multi-protein complex is composed of
to the transcription start site; those before the start site have a five-subunit core enzyme (two subunits of α, one of b, one
negative numbers and are called upstream and those after the of b′, and one of ω) plus a subunit called sigma factor (σ). The
start site have positive numbers and are called downstream. two α subunits help assemble the enzyme and promote interac-
tions with regulatory proteins, the b subunit is active in catal-
Figure 8-7 shows the promoter sequences of seven dif- ysis, the b′ subunit binds DNA, and the ω subunit has roles in
ferent genes in E. coli. Because the same RNA polymerase assembly of the holoenzyme and the regulation of gene expres-
binds to the promoter sequences of these different genes, sion. The σ subunit binds to the −10 and −35 regions, thus
similarities among the promoters are not surprising. In par- positioning the holoenzyme to initiate transcription correctly
ticular, two regions of great similarity appear in virtually at the start site (Figure 8-8a). The σ subunit also has a role in
every case. These regions have been termed the −35 (minus separating the DNA strands around the −10 region so that the
35) and −10 (minus 10) regions because they are located core enzyme can bind tightly to the DNA in preparation for
35 base pairs and 10 base pairs, respectively, upstream RNA synthesis. After the holoenzyme is bound, transcription
of the first transcribed base. They are shown in yellow in initiates and the σ subunit dissociates (Figure 8-8b). The core
Figure 8-7. As you can see, the −35 and −10 regions from enzyme then elongates through the gene (Figure 8-8c).
β
α Rewinds Unwinds
DNA
α 5′ 3′ 3′
β′ 3′
AUCG
TAG GGUA U 5′
70 CC C A T A
σ
ω
5′ 5′
–35 –10
FIGURE 8-8 (a) Binding of the σ subunit to the −10 and −35 regions positions the RNA polymerase A N I M ATED A RT
holoenzyme for correct initiation. (b) Shortly after RNA synthesis begins, the σ subunit dissociates from
the core enzyme, which continues transcription. (c) Synthesis of an RNA strand complementary to the Transcription in bacteria
single-strand region of the DNA template is in the 5′-to-3′ direction. DNA that is unwound ahead of
RNA polymerase is rewound after it is transcribed.
E. coli, like most other bacteria, has several different σ the transcript. Elongation proceeds until RNA polymerase
factors. One, called σ70 because its mass in kilodaltons (kDa) recognizes special nucleotide sequences that act as a signal
is 70, is the primary σ factor used to initiate transcription of to stop transcription and release RNA polymerase and the
the vast majority of E. coli genes. Alternative σ factors rec- nascent (i.e., newly synthesized) RNA from the template.
ognize promoters with different consensus sequences. Thus, There are two major types of termination mechanisms in E.
by associating with different σ factors, the same core RNA coli (and other bacteria), factor-independent termination
polymerase enzyme can transcribe different genes to respond (also called intrinsic or rho-independent) and Rho-dependent
to stresses, changes in cell shape, and nitrogen uptake. This is termination (also called factor-dependent) (Figure 8-9).
discussed in greater detail in Chapter 11.
DNA AAAAA
of it and rewinds the DNA that has already been tran- CA G C
A G C
scribed (Figure 8-8c). In this way, it maintains a region of A
U
A
C G
A
U
single-stranded DNA, called a transcription bubble, within A C G
RNA polymerase
RN U A
Factor-independent termination occurs after the tran- the long rRNA precursor into individual functional rRNAs.
scription of a GC-rich stretch followed by an A-rich stretch The measure of decay is half-life (also denoted t1/2), which
in the template strand. In the RNA, the GC-rich sequence is the amount of time it takes for half of the pool of an
is self-complementary and forms a 7- to 20-base-pair stem- RNA molecule to be decayed. In bacteria, mRNAs typically
loop followed by a 7- to 8-nucleotide U-rich tract. Stem- have a half-life of less than two minutes. Rapid mRNA
loops are also called hairpins because they resemble the decay is thought to allow bacteria to quickly alter gene
metal clips used to hold a person’s hair in place (Figure expression in response to changing nutritional and environ-
8-9a). Normally, in the course of transcription elongation, mental conditions.
RNA polymerase will pause if the short RNA–DNA hybrid
in the transcription bubble is weak and will backtrack to KEY CONCEPT The abundance of an RNA in cells is
stabilize the hybrid. The strength of the hybrid is deter- determined by transcription and decay.
mined by the relative number of 3-hydrogen bond G−C
and C−G base pairs compared with 2-hydrogen bond A−T
and U−A base pairs. In the factor-independent mechanism, Decay of mRNA in bacteria is commonly initiated by an
the polymerase is believed to pause after synthesizing the endonuclease that cuts the mRNA into pieces, followed by
U’s (U−A forms a weak RNA–DNA hybrid). However, the digestion of the RNA pieces by exonucleases that remove
backtracking polymerase encounters the hairpin. This road- nucleotides one at a time from the 3′ end (Figure 8-10). In
block sets off the release of RNA from the polymerase and E. coli, decay often begins with conversion of the triphosphate
the polymerase from the DNA template. at the 5′ end of the RNA to a monophosphate via removal
In the Rho-dependent termination mechanism, a pro- of pyrophosphate (PPi) by an RNA pyrophosphohydrolase.
tein called Rho factor recognizes nucleotide sequences in The 5′-monophosphate serves as a binding site for the main
the RNA that act as a termination signal for RNA poly- endonuclease RNase E, which cuts single-stranded RNA. The
merase. RNAs with Rho-dependent termination signals do RNase E products are then digested by 3′-to-5′ exonucleases.
not have the string of U residues at their 3′ end and usually Because access to the mRNA by RNase E is critical to decay,
do not have a hairpin (Figure 8-9b). Instead, they have a the presence of ribosomes can affect the half-life of an mRNA.
sequence of about 50–90 nucleotides that is rich in C res- Remember that translation in bacteria occurs while mRNAs are
idues and poor in G residues and includes an upstream being transcribed. Inefficient translation initiation presumably
segment called a Rut (Rho utilization) site. Rut sites are increases the distance between translating ribosomes, providing
located upstream (recall that upstream means 5′ of) from greater opportunity for cleavage by RNase E and decreasing
sequences at which the RNA polymerase tends to pause. the half-life of an mRNA.
Rho is a homo-hexamer consisting of six identical sub-
units that has helicase activity. Helicases use energy from
ATP hydrolysis to move along a nucleic acid and unwind
mRNA decay in E. coli
nucleic acid helices. Once bound at the Rut site, Rho travels
toward the 3′ end of the transcript. When Rho encounters a Pyrophosphohydrolase
paused RNA polymerase, it unwinds the RNA-DNA hybrid
within the transcription bubble, dissociating the RNA and pp p
terminating transcription. Thus, Rho-dependent termina-
tion entails binding of Rho to the Rut site, pausing of RNA RNase E
polymerase, and Rho-mediated release of the RNA from
RNA polymerase.
p p
KEY CONCEPT Special sequences within the 3′ UTR of an
mRNA direct transcription termination using mechanisms that
are either factor-independent or Rho-dependent. 3′ exonuclease
KEY CONCEPT The first step of decay in bacteria is carried promoters and recruit the RNA polymerase to the
out by an endonuclease, which cuts an RNA into two pieces, transcription start site. The roles of GTFs and their
and the next step is carried out by exonucleases, which digest interactions with RNA polymerases will be described in
the RNA pieces into single nucleotides starting at their 3′ end. the section on eukaryotic transcription initiation.
3. Transcription in eukaryotes takes place in the nucleus.
An important cellular difference between bacteria and
eukaryotes is the presence of a nucleus in eukaryotes (see
8.3 TRANSCRIPTION IN Figure 8-2). Because bacteria lack a nucleus, the infor-
EUKARYOTES mation in RNA is almost immediately translated into
protein, as described in Chapter 9. In eukaryotes, the
LO 8.3 Describe how mRNA transcription and decay nuclear membrane spatially separates transcription and
mechanisms in eukaryotes are similar to those in translation—transcription takes place in the nucleus and
bacteria. translation in the cytoplasm. It also means that mecha-
nisms exist to export RNAs from the nucleus to the cyto-
Transcription in eukaryotic organisms, including humans, plasm. Additionally, before RNAs leave the nucleus, they
is similar to transcription in bacteria in that they retain are modified in several ways. Both ends of an mRNA
many of the events associated with initiation, elongation, are chemically modified to protect against degradation:
and termination. However, transcription in eukaryotes is capping at the 5′ end and polyadenylation at the 3′ end.
more complex in four main ways. These modifications and others are collectively referred
to as RNA processing. Newly synthesized RNAs that are
1. Eukaryotes have many genes that are spaced far apart.
not yet processed are called primary transcripts or pre-
The larger eukaryotic genomes have many more genes
cursor RNAs (pre-RNAs), for example, pre-mRNA and
to be recognized and transcribed. Whereas bacteria usu-
pre-rRNA. RNA processing often occurs co-transcrip-
ally have a few thousand genes, eukaryotes can have tens
tionally, that is, while the RNA is being transcribed. Thus,
of thousands of genes. Furthermore, there is much more
RNA polymerases synthesize RNA while simultaneously
non-transcribed DNA in eukaryotes. Non-transcribed
coordinating a variety of processing events.
DNA originates by a variety of mechanisms that will be
discussed in Chapter 16. So, even though eukaryotes have 4. DNA in eukaryotes is packaged with proteins into chroma-
more genes than bacteria, their genes are, on average, far- tin. The template for transcription, genomic DNA, is tightly
ther apart. For example, whereas the average gene density wrapped around proteins to form chromatin in eukaryotes,
in the bacteria E. coli is 1 gene per 1400 base pairs, that whereas DNA is less compacted in bacteria. The structure
number drops in eukaryotic organisms to 1 gene per 9000 of chromatin can affect transcription initiation, elongation,
base pairs for the fruit fly Drosophila melanogaster and and termination by all three RNA polymerases as well as
1 gene per 100,000 base pairs for humans. the processing of their transcripts. These chromatin-based
mechanisms will be covered in Chapter 12.
2. Eukaryotes have three RNA polymerases. In contrast to
bacteria that have one RNA polymerase that transcribes
all genes in the cellular genome, eukaryotes have three KEY CONCEPT Differences in transcription between eukary-
RNA polymerases that work with distinct initiation, otes and bacteria are related to (1) larger eukaryotic genomes
elongation, and termination factors. RNA polymerase I with genes that are spaced further apart, (2) the division of
transcribes rRNA, excluding 5S rRNA. RNA polymerase transcription in eukaryotes among three RNA polymerases,
II transcribes all mRNAs and some ncRNAs, including (3) the nuclear membrane in eukaryotic cells that decouples
miRNAs and some snRNAs. RNA polymerase III tran- transcription and translation and necessitates nuclear RNA
export, and (4) the tight packaging of eukaryotic genomic DNA
scribes a variety of ncRNAs, including tRNAs, 5S rRNA,
into chromatin.
and some snRNAs. Each RNA polymerase is a multi-
subunit enzyme composed of about 12 proteins, some of
which are identical or similar among the polymerases and
others of which are unique to one polymerase. Transcription initiation in eukaryotes
Unique features of the RNA polymerases carry RNA polymerases I, II, and III cannot recognize promoter
out polymerase-specific functions. For example, the sequences on their own. However, unlike bacteria, where
largest protein in RNA polymerase II contains a promoters are recognized by σ factor as an integral part of
unique sequence called the carboxy-terminal domain the RNA polymerase holoenzyme, eukaryotic promoters
(CTD) that helps coordinate RNA processing events are recognized by GTFs that first bind specific sequences
that are exclusive to its transcripts. In addition, since in the promoter and then bind the RNA polymerase.
eukaryotic RNA polymerases cannot bind promot- Nevertheless, the mechanisms in bacteria and eukary-
ers or initiate transcription on their own, each poly- otes are conceptually similar. In both cases, the informa-
merase functions along with its own set of general tion that defines a promoter is provided by short DNA
transcription factors (GTFs) that are required to bind sequences located near the transcription start site, and
KEY CONCEPT RNA polymerase I, II, and III genes have a molecule’s size that is based on its rate of sedimentation
unique promoters that direct transcription initiation. Promoters upon centrifugation.
are first recognized by RNA polymerase-specific general tran- Between the tandemly repeated rDNA genes are intergenic
scription factors (GTFs). One of the main functions of GTFs is spacers (IGSs) that contains two promoter elements impor
to recruit a specific RNA polymerase and position it to begin tant for transcription initiation. A Core element is located at
RNA synthesis at the transcription start site. the transcription start site, and an Upstream Control Element
(UCE) is located 100 to 150 base pairs upstream of the tran-
scription start site (Figure 8-11, bottom). In humans, the Core
the sequences are bound by proteins that associate with element is bound by a multi-protein complex called Selec-
RNA polymerase and position it at the correct site to start tivity Factor 1 (SL1), which contains TATA-binding protein
transcription. (TBP), and the UCE is bound by Upstream Binding Factor
(UBF). In addition to SL1 and UBF, the protein TIF-1A (Tran-
RNA polymerase I promoters and GTFs In eukary- scription Initiation Factor 1A) is also required for recruitment
otic organisms, ribosomal RNA (rRNA) is transcribed by of RNA polymerase I to the transcription start site of rDNA
RNA polymerase I from hundreds of near-identical copies genes. Unlike SL1 and UBF, TIF-1A does not bind DNA, but
of rDNA genes that are tandemly repeated in the genome instead it functions through protein-protein interactions and
and reside in the nucleolus, a non–membrane-bound region forms a bridge between SL1 and RNA polymerase I.
in the nucleus where rRNA transcripts are synthesized, pro- Every time a cell divides, the number of ribosomes must
cessed, and assembled with proteins into ribosomes (see be doubled to keep the number of ribosomes in the two
Figure 8.6). Each rDNA gene encodes a single rRNA tran- daughter cells equal to that in the parent cell. Mammalian
script that contains 18S, 5.8S, and 28S rRNAs along with an cells have one to two million ribosomes. Therefore, every
external transcribed spacer (ETS) and internal transcribed cell division involves the production of one to two million
spacers (ITSs) (Figure 8-11, top). After transcription, spacer rRNA transcripts by RNA polymerase I. Consequently, the
regions are removed by processing enzymes, 18S rRNA is RNA polymerase I transcription mechanism has evolved to
assembled with ribosomal proteins to form the 40S small be exquisitely sensitive to environmental conditions that pro-
ribosomal subunit, and 5.8S, 28S, and 5S rRNAs are assem- mote or inhibit cell proliferation (i.e., an increase in the num-
bled with ribosomal proteins to form the 60S ribosomal ber of cells). Conditions that affect cell proliferation can act
subunit. RNA polymerase III transcribes 5S rRNA from on transcription initiation by RNA polymerase I by altering
tandem arrays of hundreds of gene copies located at dif- the activities of SLI, UBF, and TIF-1A. For example, condi-
ferent places in the genome from those of the RNA poly- tions that promote cell proliferation lead to phosphorylation
merase I genes. Note that the “S” in 18S, 5.8S, 28S, 5S, 40S, of TIF-1A on a specific serine residue, increasing its ability to
and 60S stands for Svedberg units, which is a measure of recruit RNA polymerase I and trigger transcription initiation.
SL1
TIF-1A RNA Pol I
UBF
TBP
5′ 3′
UCE Core
3′ 5′
−150 −100 −45 +1 +20
FIGURE 8-11 The information that controls transcription initiation by not directly bind DNA but is important for the recruitment and function of
RNA polymerase I is contained in rDNA promoter sequences (UCE and RNA polymerase I. UCE = Upstream Control Element, UBF = Upstream
Core) that are located upstream of the transcription start site (+1) and Binding Factor, TBP = TATA Binding Protein, SL1 = Selectivity Factor 1,
are bound by protein factors (UBF and SL1). A third factor, TIF-1A, does TIF-1A = Transcription Initiation Factor 1A.
KEY CONCEPT In the nucleolus, initiation of rDNA gene a TATA box, an sequence element so-named because the
transcription by RNA polymerase I is regulated by UCE and nucleotide sequence TATA appears in the consensus sequence
Core promoter elements and UBF, SL1, and TIF-1A regulatory TATAAAA. In animals, the TATA box is located about 30
factors. nucleotides upstream of the transcription start site (i.e., −30),
but in yeast its location is more variable (between −50 and
−125). Another common promoter element is the initiator
RNA polymerase II promoters and GTFs RNA poly- (Inr), which is located right at the transcription start site in
merase II is responsible for transcribing all mRNAs as well about 40 percent of genes.
as numerous ncRNAs, including snRNAs involved in splic- Collectively, only about 50 percent of RNA polymerase
ing and miRNAs involved in mRNA decay and inhibition II genes contain a TATA box and/or an Inr. This predicts
of translation. Transcription by RNA polymerase II is the the existence of other promoter elements. Computational
most complex transcriptional system in eukaryotes because analyses that searched for common sequences surrounding
of the large number of gene targets with unique expression the transcription start sites of RNA polymerase II genes
patterns. have identified additional promoter elements, including
RNA polymerase II promoters, which are somewhat arbi- the downstream promoter element (DPE), which is located
trarily defined as sequences located within 100 base pairs of at about +25, and the TFIIB recognition element (BRE),
the transcription start site, contain a variety of promoter ele- which is located at about −40. However, since many genes
ments, a few of which are relatively common (Figure 8-12). lack all of the known promoter elements, it is likely that
About 25 percent of promoters in yeast and humans contain promoter elements remain to be discovered.
TFIIA
5′ 3′
BRE TATA INR DPE
3′ 5′
−40 −30 +1 +25
CTD P
P Initiation
P CTD phosphorylation
P
P
Capping CTD P
TFIIB TFIIE TFIIH Elongation P P P
P P
P P NELF
P
m7G P
P
P
RNA Pol II
TFIIA DSIF
TFIIF
TFIID P-TEFb
FIGURE 8-12 Transcription initiation by RNA polymerase II is directed by a variety of promoter A N I M ATED A RT
elements, including the BRE, TATA box, Inr, and DPE, located within 100 base pairs upstream or
downstream of the transcription start site (+1). Assembly of the PIC occurs in a sequential manner, Transcription in eukaryotes
starting with TFIID, which contains proteins (TBP and TAFs) that bind promoter elements. Assembly of
TFIID leads to recruitment of the other GTFs and RNA polymerase II. TFIIH is required during initiation
for opening the transcription bubble and phosphorylation of the CTD. Shortly after initiation, the pre-
mRNA is capped and elongation is promoted by the P-TEFb kinase, which phosphorylates the RNA
polymerase II CTD, DSIF, and NELF.
KEY CONCEPT About half of RNA polymerase II genes to the promoter by TFIIE, contains proteins with helicase
contain TATA box and/or Inr promoter elements. The other activity that unwind the DNA into two strands to form
half contain less common promoter elements, some of which the transcription bubble. Lastly, TFIIF places the promoter
remain to be defined. DNA in a position in RNA polymerase II that is appropri-
ate for DNA unwinding and initiation of transcription at
the start site. After transcription has been initiated, RNA
All of the proteins that bind RNA polymerase II pro- polymerase II dissociates from most of the GTFs to elon-
moter elements are subunits of GTFs. Transcription fac- gate the RNA transcript. Some GTFs, including TFIID,
tor IIB (TFIIB) binds the BRE, and TFIID binds the other remain at the promoter to attract the next RNA polymerase
promoter elements. The TFIID complex contains TBP (the II. In this way, multiple RNA polymerase II molecules can
same protein involved in RNA polymerase I transcription) simultaneously synthesize transcripts from a single gene.
and about 15 TBP-associated factors (TBP). TBP binds
the TATA box, and TAFs bind the Inr and DPE. Binding KEY CONCEPT The general transcription factors (GTFs)
of TFIID at a promoter is the first step in the sequential TFIIB and TFIID recruit RNA polymerase II and other GTFs to
assembly of other GTFs and RNA polymerase II. TFIID the promoter, forming a pre-initiation complex (PIC) and a tran-
binding instructs assembly of TFIIA and TFIIB, followed scription bubble.
by TFIIF and RNA polymerase II as a pre-assembled com-
plex and ending with addition of TFIIE and TFIIH. The RNA polymerase III promoters and GTFs RNA poly-
assemblage of GTFs and RNA polymerase II constitutes the merase III transcribes noncoding RNAs (ncRNAs) shorter than
preinitiation complex (PIC), which serves to position RNA 300 nucleotides. The RNA polymerase III gene targets are clas-
polymerase II at the transcription start site, generate the sified into three types based on their promoter elements. The
transcription bubble, and position the DNA in the active 5S rRNA genes are Type 1 and contain three promoter ele-
site of RNA polymerase II. TFIIA stabilizes the binding of ments, Box A, intermediate element (IE), and Box C, that are all
TFIIB and TFIID at the promoter. TFIIH, which is recruited located downstream of the transcription start site (Figure 8-13a).
5′ 3′
Box A IE Box C
3′ 5′
+1 +60 +85
TBP TFIIIC
5′ 3′
Box A Box B
3′ 5′
+1 +20 +60
5′ 3′
DSE PSE TATA
3′ 5′
−200 −60 −25 +1
FIGURE 8-13 RNA polymerase III transcribes genes with three types of promoters, Types 1, 2, and
3 (a–c). TFIIIA, TFIIIC, and SNAPc bind promoter elements (IE, Box B, and PSE, respectively) that are
unique to each type of RNA polymerase III gene and can be viewed as the specificity factors for the
gene type. The main function of these factors is to recruit TFIIIB to the promoter, which then leads to
recruitment of RNA polymerase III and transcription initiation.
tRNA genes are Type 2 and contain the same Box A element RNA polymerase II CTD
as 5S genes, but instead of a Box C they have a Box B (Figure
8-13b). Lastly, Type 3 genes such as the U6 snRNA gene that is Interacting
involved in pre-mRNA splicing contain promoter elements that Transcription protein factor
initiation
are located upstream of the transcription start site and look
YSPTSPS CTD repeat
very similar to promoter elements found in RNA polymerase II 2 5 sequence
genes, including a TATA box (Figure 8-13c).
The GTFs for RNA polymerase III transcription are des- Capping
P
ignated TFIIIA, TFIIIB, TFIIIC (transcription factor for RNA
YSPTSPS
polymerase III), and SNAPc (snRNA activating protein com-
Transcription cycle
plex). TFIIIB is required for the transcription of all three types
of RNA polymerase III genes and functions to recruit RNA Transcription
polymerase III as well as open the transcription bubble. TFIIIA elongation P P
binds promoter elements in Type 1 genes and helps recruit YSPTSPS
TFIIIC, which binds promoter elements in Type 1 and 2 genes. Splicing
Note that since TFIIIA and TFIIIC bind downstream of the P
transcription start site, they need to be temporarily displaced YSPTSPS
as RNA polymerase III transcribes through the promoter DNA.
For Type 3 genes, recruitment of RNA polymerase III is assisted
Polyadenylation
by binding of both the TBP subunit of TFIIIB to the TATA box P
and of SNAPc to the proximal sequence element (PSE). Binding YSPTSPS
of SNAPc to the PSE is helped by transcription factors that bind Transcription
to the distal sequence element (DSE). Transcription factors are termination P
discussed in detail in Chapter 12. YSPTSPS
KEY CONCEPT Genes transcribed by RNA polymerase III FIGURE 8-14 The pattern of amino acid modifications within the
are divided into three types, based on their promoter elements. RNA polymerase II CTD changes as the polymerase transcribes
Type 1 and Type 2 have promoter elements downstream of through a gene. The modifications, including phosphorylation of
the transcription start site. Type 3 promoter elements include serines 2 and 5 by kinases and dephosphorylation by phosphatases,
a TATA box and are positioned upstream of the transcription create a code that directs the binding of factors at specific stages
start site. of transcription. Factors that bind the CTD include regulators of
transcription elongation and termination as well as RNA processing
events, including capping, splicing, and cleavage and polyadenylation.
RNA polymerase II transcription Transcription initiation is not a green light that liberates
elongation RNA polymerase II to completely transcribe a gene unimpeded.
Shortly after transcription initiation, phosphorylation of In fact, transcription continues to be regulated all along the
RNA polymerase II by a protein kinase in TFIIH helps length of a gene. For example, for a large fraction of human
coordinate the processing of mRNAs as they are being genes, transcription elongation is temporarily stopped (i.e.,
transcribed. The carboxy-terminal domain (CTD) of paused) about 50 base pairs downstream of the transcrip-
the largest subunit of RNA polymerase II contains the tion start site. Pausing of this type is caused by the protein
sequence YSPTSPS (tyrosine-serine-proline-threonine-serine- factors NELF (negative elongation factor) and DSIF (DRB
proline-serine) tandemly repeated 26 times in yeast and 52 sensitivity-inducing factor) and relieved by P-TEFb (positive
times in humans. Phosphorylation of the serine in position 5 transcription elongation factor b) (see Figure 8-12). To release
of the repeat (S5) by TFIIH serves as a signal for the binding paused RNA polymerase II into productive elongation, P-TEFb
of enzymes that cap the 5′ end of the mRNA (discussed in the phosphorylates NELF and DSIF. NELF dissociates from the
next section) (Figure 8-14). The CTD is located near the site elongation complex, and DSIF travels along with RNA poly-
where nascent RNA emerges from RNA polymerase II, so it merase II and functions as a positive elongation factor. P-TEFb
is in an ideal place to orchestrate the binding and release of also phosphorylates the RNA polymerase II CTD on serine 2
proteins needed to process the nascent transcript while RNA (S2) within the YSPTSPS repeats (see Figure 8-14), which serves
synthesis continues. Post-translational modification of S5 and as a signal for the binding of factors involved in processing of
other amino acids in the CTD change as RNA polymerase II the pre-mRNA and transcription termination.
transcribes through a gene, creating different binding sites for
other processing factors as well as factors that regulate tran- Transcription termination in eukaryotes
scription elongation and termination. Transcription termination for the three RNA polymerases
occurs by different mechanisms. Elongating RNA polymerase I
KEY CONCEPT During elongation, the CTD of RNA poly-
is stopped by protein factors bound at specific DNA sequences
merase II is chemically modified to serve as a binding site for
called terminator elements and is released from DNA by other
other proteins involved in transcription and RNA processing.
factors. In contrast, RNA polymerase III terminates elongation
and dissociates from DNA after the synthesis of a poly(U) cleavage, the pre-mRNA is cleaved, and the new 5′-mono-
stretch, similar to factor-independent termination in bacteria. phosphorylated end that is formed is a substrate for a
Two models have been proposed for transcription ter- 5′-to-3′ exonuclease called Xrn2, which digests the RNA
mination by RNA polymerase II—the torpedo model and one nucleotide at a time until eventually reaching RNA
the allosteric model. The models are conceptually similar polymerase II and causing it to dissociate from DNA. Xrn2
to Rho-dependent and factor-independent mechanisms, is positioned to act in termination through its association
respectively, in E. coli (see Figure 8-9), but different factors with the CTD phosphorylated on serine 2 (see Figure 8-14).
are involved. Both RNA polymerase II termination models In the allosteric termination model (Figure 8-15b), tran-
couple 3′-end formation to termination. As described below scription through the site of cleavage causes elongation factors
in the section on polyadenylation, the 3′ ends of mRNAs to dissociate, leading to a conformational change within the
are determined by cleavage of the pre-mRNA and addition active site of RNA polymerase II and its release from DNA. In
of a poly(A) tail to the new 3′ end. this model, it remains to be determined how RNA polymerase
In the torpedo termination model (Figure 8-15a), RNA II senses passage through the site of cleavage and how this leads
polymerase II continues to transcribe past the site of to dissociation of elongation factors.
RNA Pol II
DNA
CP
complex
Poly(A) site
5′ mRNA
Cleavage and
polyadenylation
CTD CTD
CP CP
complex AA complex AA
AA AA
Poly(A) site Poly(A) site
RNA Pol II RNA Pol II
Xrn2
mRNA mRNA
Dissociation of mRNA
and RNA polymerase II
AA
AAA A
AA l II
A Po
RN
FIGURE 8-15 Current data support two models for how transcription exonuclease Xrn2 that degrades the RNA to elicit termination. (b) In
of mRNAs by RNA polymerase II is terminated. (a) In the torpedo model, the allosteric model, upon encountering cleavage and polyadenylation
the piece of RNA that stays associated with RNA polymerase II and signals, RNA polymerase II undergoes a conformation change that
continues to be synthesized after cleavage is a substrate for the 5′-to-3′ commits it to termination.
KEY CONCEPT Transcription termination by RNA poly- (i.e., co-transcriptionally) and are coordinated with transcrip-
merases I, II, and III occurs by different mechanisms. Ter- tion initiation, elongation, or termination as well as with one
mination of mRNA transcription by RNA polymerase II may another. Some themes emerge upon comparison of the process-
occur by allosteric or torpedo mechanisms that are analogous ing events: (1) sequence elements within mRNAs often direct
to factor-independent and Rho-dependent mechanisms, where processing occurs, and (2) sequence elements are bound
respectively, in E. coli and are directed by 3′-end formation by proteins or ncRNAs that are themselves enzymes or that
sequences in the mRNA.
recruit enzymes to carry out the processing.
Capping
(a)
O CH3
HN 1 6 N+
5 7 O O O
2 4 8
5′ 5′
Base 1
3 9
H2N N N O H2C O P O P O P O CH2 O
O− O− O−
OH OH O OH
m7G −O P O
CH2
Base 2
O O
O OH
(b)
S5 CT S5 CT S CT
MT P D MT P D MT 5P D
GT TP GT TP GT TP
5′ GTP SAM
3′
+1 +25 +1 +25 +1 +25
FIGURE 8-16 (a) The first nucleotide in an mRNA is modified by All three enzymes are positioned to act on the mRNA shortly after it
addition of an m7G cap. (b) Capping is initiated by a triphosphatase (TP) emerges from the exit tunnel of RNA polymerase II because of their
enzyme that removes the terminal phosphate from the first nucleotide. association with the CTD that is phosphorylated on serine 5 (S5P) of
Then a guanylyltransferase (GT) adds a guanine monophosphate repeat sequences.
nucleotide, and a methyltransferase (MT) adds a methyl group (CH3). ANIMATED ART Transcription in eukaryotes
(Figure 8-16b): RNA triphosphatase removes the gamma determine the site of cleavage that is bound by Cleavage
phosphate from the first nucleotide in the RNA chain, Factors I and II (CFI and CFII). During its synthesis, the
guanylyltransferase uses GTP as a substrate and links GMP poly(A) tail is bound by poly(A) binding protein (PABP),
to the first nucleotide by an unusual 5′,5′-triphosphate link- which in the cytoplasm protects the mRNA from decay by
age (GpppN, where N is the first nucleotide in the mRNA), exonucleases and promotes translation by interacting with
and 7-methyltransferase adds a methyl group (CH 3) from the translation machinery.
S-adenosylmethionine (SAM) to the N-7 position of the
newly added guanine base. Through their interaction with KEY CONCEPT The 3′ end of mRNAs is modified by addition
the CTD of RNA polymerase II, these enzymes are in posi- of a long stretch of adenosine nucleotides, which protects the
tion to act on RNAs early in the transcription process (see mRNA from decay and supports translation. The poly(A) tail is
Figure 8-14). added by a special type of RNA polymerase following mRNA
cleavage at a site that is determined by protein factors that
Caps serve multiple functions. They protect RNAs from
bind sequence elements in the mRNA.
decay by exonucleases, which often require 5′-phosphates
to recognize their substrates. Caps on mRNAs also serve as
a binding site for proteins such as the cap binding complex The discovery of splicing
(CBC) that mediate subsequent events, including splicing,
Transcription copies the DNA sequence of protein-coding
polyadenylation, and nuclear export, by interacting with
genes into mRNA, yet sequence comparison of most pairs of
processing and export factors. The CBC is also critical for
human mRNAs and genes shows that they are different: large
the first round of translation, while another cap binding
stretches of DNA sequence are transcribed into RNA and later
protein, eIF4E, is required for subsequent rounds of trans-
removed from the RNA. In 1977, the laboratories of Philip
lation, as described in Chapter 9.
Sharp and Richard Roberts independently discovered this pro-
KEY CONCEPT The 5′ end of a eukaryotic mRNA is modified
cess of mRNA splicing, which removes segments of mRNA
to prevent decay and to serve as a binding site for factors that called introns and links together the remaining segments
mediate mRNA processing and export. Capping of mRNAs is called exons. The cutting out of introns and the joining of
programmed to occur early in transcription through the associ- exons is called splicing because it resembles the way in which
ation of capping enzymes with phosphorylated serine 5 on the movie film is cut and rejoined to delete a specific segment.
CTD of RNA polymerase II.
KEY CONCEPT The sequence of an mRNA is not always
identical to its gene sequence because, as pre-mRNAs are
Polyadenylation transcribed, introns are removed and the exons that remain are
joined together in the process of splicing.
Like the 5′ end, the 3′ end of an mRNA is also processed to
protect it from decay and to promote translation. Process-
The number and size of introns varies from gene to
ing at the 3′ end consists of two events: cleavage, which cuts
gene and from organism to organism. For example, only
the mRNA away from the transcribing RNA polymerase
about 5 percent of the genes in yeast (S. cerevisiae) have
II, and polyadenylation, which adds 50–250 adenosine (A)
introns. Intron-containing genes in yeast almost always
residues to the end of the cleaved mRNA. Sequence ele-
have a single intron that ranges in length from 50 to 1000
ments within the 3′ UTR determine where cleavage occurs.
nucleotides, with an average length of 250 nucleotides.
In humans, the highly conserved six-nucleotide (hexa
In contrast, 85 percent of human genes have at least one
nucleotide) sequence AAUAAA is located 10–30 nucle-
intron, and an average human gene has eight introns and
otides upstream of the cleavage site, also known as the
nine exons. An extreme example is the gene that is mutated
poly(A) site (Figure 8-17a). A less well conserved U-rich or
in Duchenne muscular dystrophy, which has 78 introns
GU-rich downstream sequence element (DSE) is located
and 79 exons spread across 2.3 million base pairs. Human
20–40 nucleotides downstream of the poly(A) site. Cleav-
introns vary in length from 50 to 300,000 nucleotides, with
age at the poly(A) site often occurs after a CA or UA
an average of 6000 nucleotides, whereas the average exon
(Figure 8-17b). The AAUAAA is important for both cleav-
length is 300 nucleotides. The 20-fold-larger average size
age and polyadenylation because it is bound by a protein
of introns relative to exons means that introns account for
complex called Cleavage and Polyadenylation Specificity
a much greater fraction of the human genome than exons.
Factor (CPSF), which contains the endonuclease enzyme
that executes the cleavage step. CPSF also recruits poly(A)
polymerase (PAP), the enzyme that uses ATP as a substrate The splicing mechanism
to add a string of A’s onto the 3′-OH of the mRNA, which After the discovery of exons and introns, researchers
is referred to as a poly(A) tail. PAP is an unusual RNA turned their attention to the mechanism of mRNA splicing.
polymerase because, unlike DNA polymerases and other Because splicing must occur with single nucleotide preci-
RNA polymerases, it does not copy a nucleic acid template. sion to maintain the information that directs translation,
In addition to CPSF, the cleavage step involves CstF (Cleav- the intron-containing mRNA precursor (pre-mRNA) must
age Stimulatory Factor), which binds the DSE and helps hold the information that points the splicing machine called
Polyadenylation
(a)
CPSF PAP
S2 CT
CF P D
AAU
AAA
CstF
Poly(A
) site
RNA Pol II
DSE
5′ 3′
AAUAAA Poly(A) site DSE
3′ 5′
15–30 nt 0–20 nt
Cleavage site
(b)
Cleavage by CPSF
Decay by Xrn2
Polyadenylation by PAP
FIGURE 8-17 The 3′ end of an mRNA is generated by consecutive ends of genes by association with the RNA polymerase II
cleavage and polyadenylation reactions. (a) In humans, the site of CTD phosphorylated on serine 2 (S2P) of repeat sequences. (b)
cleavage is directed by three sequences in the 3′ UTR of an mRNA (i.e., Cleavage at the poly(A) site by CPSF is followed by addition of a poly(A)
AAUAAA, poly(A) site, and DSE), each of which is bound by protein tail to the new 3′ end by PAP and by association of PABP with the
factors (i.e., CPSF, CF, and CstF, respectively) that are recruited to the poly(A) tail.
the spliceosome where to act. Researchers hypothesized Patients with a variety of autoimmune diseases, including
that the information would be provided by sequences at the systemic lupus erythematosus, produce antibodies against
boundaries between exons and introns. In fact, alignments their own proteins. In the course of analyzing blood sam-
of boundary sequences for many pre-mRNAs revealed ples from patients with lupus, Steitz and colleagues identi-
that almost all introns begin with GU and end with AG fied antibodies that bound nuclear RNA-protein complexes
( Figure 8-18a). In addition, there is high conservation of called small nuclear ribonucleoproteins (snRNPs), pro-
intron and exon nucleotides adjacent to the GU and AG. nounced “snurps,” that are comprised of a small nuclear
The GU and AG sequence elements define the 5′ splice site RNA (snRNA) 100–200 nucleotides in length that serves as a
and 3′ splice site, respectively, where cuts are made by the scaffold for binding several proteins. They observed that the
spliceosome to remove the intron. In addition, a third con- sequence at the 5′ end of the snRNA named U1 has extensive
served sequence called the branch point is located 15–45 complementarity to the sequence at 5′ splice sites, suggesting
nucleotides upstream of the 3′ splice site. An invariant ade- that the U1 snRNA identifies 5′ splice sites by base pairing
nosine within the branch point participates in the first cata- (Figure 8-18b). To test this hypothesis, the laboratory of Alan
lytic step of splicing. The existence of conserved nucleotide Weiner performed a mutational analysis. They found that
sequences at splice sites and the branch point suggested that splicing was dramatically reduced by mutations in a 5′ splice
components of the spliceosome are directed to act at spe- site sequence that partially disrupted base pairing with the
cific places in pre-mRNAs by binding to these sequences. U1 snRNA (Figure 8-18c). Moreover, they found that splic-
A serendipitous finding in the laboratory of Joan Steitz ing of the mutant pre-mRNA was recovered by mutations in
led to the discovery of components of the spliceosome. the 5′ end of the U1 snRNA that restored base pairing. This
G U C C A U U C A U A 5′ cap
U1 snRNA
A G G G U G A G G A G G G U G A A U A A G G U G A A U
G U C C A U U C A U A G U C C A U U C A U A G U C C A U U U A U A
FIGURE 8-18 (a) Conserved nucleotide sequences are present at the consensus 5′ splice site. (c) The efficiency of splicing is affected by the
junctions of exons and introns (i.e., 5′ and 3′ splice sites) as well as at strength of U1 snRNA base pairing at the 5′ splice site. Mutations in
the branch point, which is near the 3′ splice site. Invariant nucleotides the 5′ splice site that reduce the number of hydrogen bonds lead to a
(GU at the 5′ splice site, A at the branch point, and AG at the 3′ splice decrease in the efficiency of splicing (compare left and middle); however,
site) are shown in bold font, and N represents any base. (b) The U1 splicing efficiency can be restored by compensatory mutations in the U1
snRNA recognizes 5′ splice sites by base pairing. Sequences near snRNA (right).
the 5′ end of the U1 snRNA form eight consecutive base pairs with a
“compensatory mutation” analysis demonstrated that base Splicing, then, takes place by means of two transes-
pairing between snRNAs and the pre-mRNA is important for terification reactions (Figure 8-19b). The first step of the
the selection of splice sites. reaction involves nucleophilic attack by the 2′-OH of
the unpaired branch point adenosine at the phosphodi-
KEY CONCEPT snRNAs facilitate splicing by base pairing ester bond at the 5′ splice site, which cuts the pre-mRNA
with conserved sequences in the pre-mRNA. between the 5′ exon and the intron and produces an intron
with a loop structure called a lariat because it resembles the
shape of a cowboy’s lariat (lasso). The second step of the
In addition to the U1 snRNP, the spliceosome contains U2,
reaction involves nucleophilic attack by the 3′-OH of the 5′
U4, U5, and U6 snRNPs as well as many proteins that have
exon at the phosphodiester bond at the 3′ splice site, which
conserved functions in eukaryotes from yeast to humans.
covalently links together the 5′ and 3′ exons and frees the
The splicing reaction begins with stepwise recognition of pre-
intron as a lariat. Lastly, the U2, U5, and U6 snRNPs are
mRNA sequence elements (Figure 8-19a). First, U1 binds the 5′
released from the excised lariat and participate in another
splice site and U2 binds the branch point, with the U2 snRNA
cycle of splicing along with previously released U1 and
base pairing to nucleotides across the branch point, except for
U4 snRNPs. This process is repeated for each intron in a
the key adenosine. Spliceosome assembly is completed by entry
pre-mRNA.
of the U4, U5, and U6 snRNPs as a preassembled tri-snRNP
complex. At this point, the spliceosome undergoes several con-
formational changes to become catalytically active. The U1 and KEY CONCEPT Splicing is a two-step reaction. The first step
U4 snRNPs are released from the spliceosome, the U6 snRNP is cleavage at the 5′ splice, and the second step is cleavage
at the 3′ splice site, which results in removal of the intron and
base pairs to the 5′ splice site, and the U5 snRNP base pairs to
joining of the exons.
both exon sequences, placing the splice sites in close proximity.
Splicing mechanism
(a) (b)
5′ GU A AG 3′ Branch point adenosine
Exon 1 Intron Exon 2
5′ splice site 3′ splice site
U1 snRNP binds
the 5′ splice site O
2′
O H O
U1 U2 5′ exon 3′ exon
3′ 5′ 5′ 3′ 3′ 5′
U2 snRNP binds Exon 1 O P O GU O A O AG O P O Exon 2
the branch point
O− O−
GU A AG
U1 U2
Step 1 Transesterification
U4 U4, U5, and U6 snRNPs
U6 U5
join the spliceosome as
a preformed complex
U1
UG O
U4 5′
U6 U5 UG O P O− + Exon 1
3′
O−
2′
A AG O O
U2 5′ 3′ 3′ 5′
O A O AG O P O Exon 2
U1 and U4 snRNPs O−
U1 U4 are released from
the spliceosome
UG
Step 2 Transesterification
U6 U5
A AG
U2
The first catalytic O
5′
step of splicing UG O P O−
2′
O O
5′
O A O
3′
AG OH
3′
+ 3′
Exon 1 O P O Exon 2
5′
U6 U5
UG
O−
A AG
U2 Lariat intron Linked exons
The second catalytic
step of splicing
U6 U5
UG
A AG +
U2
Lariat intron Linked exons
FIGURE 8-19 (a) The spliceosome is formed by sequential assembly bond at the 5′ splice site. This produces two intermediates, a 5′ exon
of five snRNPs (U1, U2, U4, U5, and U6) onto an intron. Conformational and an intron–3′ exon in a lariat structure. In the second step of splicing,
changes that lead to release of the U1 and U4 snRNPs position the U2 the 3′-hydroxyl of the 5′ exon attacks the phosphodiester bond at the 3′
and U6 snRNAs and the branch point adenosine near the 5′ splice site splice site, producing linked exons and the released lariat intron.
for the first catalytic step of splicing. (b) In the first step of splicing, the
2′-hydroxyl of the branch point adenosine attacks the phosphodiester ANIMATED ART Mechanism of mRNA splicing
OH (GTP)
OH
G
5′ G 3′ 5′ A 3′
1 1
5′ G G 3′
5′ OH A 3′
2
5′ OH
5′ 3′ Linked exons
+ 2
5′ G G 3′
OH
3 5′ 3′ Linked exons
+
5′ G 3′ + G
Lariat intron Lariat intron
A
FIGURE 8-20 Similar to splicing by the spliceosome, RNA-catalyzed self-splicing of Group I and
Group II introns involves two transesterification reactions. In Group I splicing, the first reaction is
carried out by GTP, whereas in Group II splicing and spliceosomal splicing. The first reaction is
carried out by the branch point adenosine.
structure of Group II introns and the chemistry of splicing by ( Figure 8-21). The FGFR2 protein is made up of several
the spliceosome is the same as that carried out by Group II domains, including an extracellular ligand-binding domain.
introns (Figure 8-20), it has been hypothesized that spliceo- Alternative splicing results in two protein isoforms that
somal and Group II splicing mechanisms are evolutionarily differ in their extracellular domains. Because of this differ-
related and that the spliceosome is a ribozyme. ence, each isoform binds different growth factors. In addi-
tion, as illustrated by the α-tropomyosin gene (Figure 8-22),
KEY CONCEPT Intron removal and exon joining are cata- protein isoforms can be produced in particular cells by cell
lyzed by RNA molecules. In eukaryotes, the snRNAs of the type-specific alternative splicing, and they can also be pro-
spliceosome catalyze the removal of introns from pre-mRNA. duced at different stages of development. Thus, alternative
Some introns are self-splicing; in these cases, the intron cat-
splicing expands the proteome (the set of all proteins that
alyzes its own removal. RNAs capable of catalysis are called
can be expressed) of eukaryotic organisms.
ribozymes.
In humans, about 95 percent of intron-containing genes
undergo alternative splicing to encode two or more protein
isoforms. There are four general types of alternative splic-
Alternative splicing can expand the ing, the most common of which is exon skipping, where an
proteome exon is either included or excluded in the mature mRNA
A major rationale for having genes with introns is that (Figure 8-23). The other types of alternative splicing are alter-
introns provide a mechanism to encode different proteins native 3′ splice sites (i.e., one 5′ splice site and a choice of
(called protein isoforms) from a single gene. Through the two 3′ splice sites), alternative 5′ splice sites (i.e., one 3′ splice
process of alternative splicing, exons in a pre-mRNA can be site and a choice of two 5′ splice sites), and mutually exclu-
joined together in different combinations to produce differ- sive exons (i.e., only one of several exons is included in the
ent mature mRNAs that encode protein isoforms. An extreme mature mRNA, as illustrated by FGFR2 in Figure 8-21).
example is the Dscam gene in Drosophila that via alternative A key feature of alternative splicing mechanisms is that
splicing can produce 38,016 different Dscam proteins. 5′ and 3′ splice site sequences differ among exons. Alterna-
Alternative splicing can produce protein isoforms with tive exons tend to have weak splice site sequences that have
different functional domains. This is illustrated by FGFR2, lower affinity for spliceosome components than splice sites
a human gene that encodes a receptor that binds fibroblast associated with constitutive exons (i.e., exons that are always
growth factors and then transduces a signal inside the cell spliced into the mature mRNA). Weak splice sites are subject
Alternative splicing
mRNA
Ligands: Ligands:
FGF10 FGF2
FGF7 FGF9
FGF4
FGF8
Exterior FGF6
Cell membrane
Cytoplasm
FGFR2 FGFR2
(First isoform) (Second isoform)
Fibroblast
TM-3
Fibroblast
TM-5a
Fibroblast
TM-5b
Exon skipping
FIGURE 8-23 A single pre-mRNA that contains exons and introns can be spliced in different patterns
to produce mature mRNAs that encode different proteins. There are four major types of alternative
splicing. Dotted lines on the top and bottom of each pre-mRNA show how the 5′ and 3′ splice sites
can be alternatively joined to produce different mature mRNAs.
to alternative splicing because their use by the spliceosome is rather than adenosine (Figures 8-24b and 8-24c), changing
not optimal and can thus be enhanced or suppressed by regu- the amino acid sequence of a protein.
latory factors. Alternative splicing can also be affected by the A-to-I editing can also affect regulatory elements in RNAs
rate of transcription elongation. Central to this mechanism is that function by base pairing to another RNA or are bound
the fact that most splicing events take place during transcrip- by a protein. For example, conversion of a stable A-U base
tion. Spliceosomes assemble on introns as soon as they are pair into a less stable I-U base pair can alter splicing by affect-
transcribed. Slow elongation by RNA polymerase II provides ing base pairing between snRNAs and pre-mRNAs. High-
a longer window of opportunity for the use of an alternative throughput RNA sequencing methods have identified over two
weak splice site before a strong splice site is transcribed. million A-to-I edited sites in the human transcriptome (the set
of all RNAs that can be expressed in humans). The physiolog-
ical consequence of A-to-I editing at most of these sites is yet
KEY CONCEPT The joining of exons in different patterns via
alternative splicing greatly expands the number of proteins to be determined, but global effects on A-to-I editing caused
encoded in the human genome and other eukaryotic genomes. by mutation of ADARs leads to behavioral and locomotion
abnormalities in Drosophila and seizures and early death in
mice, highlighting the importance of A-to-I editing.
RNA editing
RNA sequences encoded in eukaryotic genomes are not RNA nucleotide modification
only changed by RNA processing events such as splicing The structure and function of RNAs can be altered by
but also by RNA editing. RNA editing is a general term post-transcriptional chemical modifications. Most RNA mod-
that describes molecular processes through which nucle- ifications consist of the addition of a methyl group (CH3) to
otide sequences in RNAs are changed after transcription. a nucleoside base such as N6-methyladenosine (m6A) (Figure
Editing events include insertion and deletion of nucleotides 8-24d), to a ribose sugar such as 2′-O-methyladenosine (Am)
as well as base substitution. Many types of RNA, including (Figure 8-24e), or to both such as N6, 2′-O-dimethyladenosine
tRNAs, rRNAs, mRNAs, and snRNAs, are edited. (m6Am). More than 100 different chemical modifications of
In animals, the most common type of editing is RNA have been identified. Each modification can have distinct
adenosine-to-inosine (A-to-I) editing, which converts ade- effects on RNA structure and interactions with other RNAs
nosine to inosine by deamination. A-to-I editing is catalyzed and proteins, which can affect all aspects of RNA metabolism,
by double-strand RNA-binding enzymes called adenosine including the processing, stability, and translation of mRNAs.
deaminase acting on RNAs (ADARs) (Figure 8-24a). Inosine m 6 A is the most common modification in human
is a non-canonical nucleoside that can base pair with cyti- mRNAs. Of the more than 20,000 m6A sites that have been
dine, so during translation inosine is read as a guanosine, identified in humans, 70 percent occur in the last exon of a
N N ADAR N N N N Methyltransferase N N
Ribose Ribose Ribose Ribose
Adenosine Inosine Adenosine N6-methyladenosine (m6A)
(b) H (e)
N N H O
NH2 NH2
N N H N N SAM SAH N
N N
Ribose N N
O N N Methyltransferase N N
Ribose
Adenosine Uridine
O O O O
(c) H H H H H
N O H N H H H H
O OH O OCH3
N N H N
Ribose N N Adenosine 2′-O-methyladenosine (Am)
O Ribose
Inosine Cytidine
FIGURE 8-24 The information contained in RNA can be changed by editing and modification.
(a) A-to-I editing is carried out by ADAR enzymes. (b) Adenosine base pairs with uridine, but (c) inosine
base pairs with cytidine. (d) The chemical structure of ribonucleoside bases and (e) ribose sugars can
be changed by methylation. These reactions are catalyzed by different methyltransferases, but both
reactions use SAM as the methyl-group donor.
transcript and nearly half occur in the 3′ UTR. m6A mod- proteins, and then returned to the nucleus. Export of mRNAs
ifications are added during transcription and prior to the and snRNAs from the nucleus occurs by different mecha-
completion of splicing. A major function of m6A is to desta- nisms, but both mechanisms involve adaptor proteins that
bilize mRNAs, as demonstrated by the finding that knock- bind the RNAs early in their biogenesis and escort them to
ing out the methyltransferase that writes m6A into mRNA the cytoplasm through channels in the nuclear membrane
results in longer half-lives of m6A-containing mRNAs. Newly called nuclear pores. In human cells, mRNAs are transported
developed technologies that map the sites of chemical mod- out of the nucleus by the TREX (transcription export) com-
ifications in RNAs are making it possible for researchers to plex, whereas snRNA are transported by PHAX (phosphory
identify the writers, readers, and erasers of modifications as lated adaptor for RNA export). Both TREX and PHAX
well as determine their molecular and biological functions. interact with their RNA cargo during transcription through
binding to the cap binding complex (CBC).
KEY CONCEPT RNAs are subject to editing and modifica-
tion. Editing can change the protein sequence encoded by KEY CONCEPT Mechanisms exist in eukaryotic organisms to
an mRNA, and both editing and modification can create new transport and localize RNAs to particular places in cells.
signals in mRNAs and ncRNAs that change their structure,
function, and stability.
8.5 DECAY OF mRNA IN
RNA export from the nucleus EUKARYOTES
Many eukaryotic RNAs that are transcribed in the nucleus LO 8.5 Describe how siRNAs regulate the abundance of
from the nuclear genome spend some part of their life in specific RNAs and play a role in maintaining genome
integrity in eukaryotes.
the cytoplasm. For example, mRNAs are exported from
the nucleus to the cytoplasm where they are translated into
proteins, and snRNAs involved in splicing are produced in As in bacteria, decay counterbalances transcription to regu-
the nucleus, exported to the cytoplasm for assembly with late the abundance of mRNAs in eukaryotes. The half-life of
A
the exosome and decapping by DcpS.
A
Dcp1/Dcp2
Exosome
m7Gpp p m7Gppp
Xrn1 DcpS
m7Gp pp
eukaryotic mRNAs can vary widely. For example, the β-globin The efficiency of decay pathways can be enhanced
mRNA, which encodes a subunit of hemoglobin in blood cells, or suppressed by RNAs and proteins that bind spe-
has a half-life of 20–24 hours. In contrast, the c-Myc mRNA, cific sequences within 3′ UTRs of mRNAs and affect the
which encodes a transcription factor that regulates the cell recruitment of decay factors. For example, base pairing of
cycle, has a half-life of 20–30 minutes. Mechanisms that con- miRNAs to sequences within the 3′ UTR not only inhib-
trol mRNA half-life help ensure that proteins are only present its translation (discussed in Chapter 9) but also enhances
in cells when they are needed. mRNAs with a long half-life, decay by recruiting deadenylase and decapping enzymes.
like β-globin, tend to encode proteins with structural or meta- Similarly, sequences in the 3′ UTR that are rich in adeno
bolic functions, and mRNAs with a short half-life, like c-Myc, sine and uridine known as AU-rich elements (AREs) serve
tend to encode proteins with regulatory functions. as binding sites for RNA-binding proteins that enhance or
The half-life of an mRNA is experimentally determined suppress decay by affecting recruitment of decay factors.
by turning off all RNA polymerase II transcription and then Related but distinct pathways detect and rapidly decay
measuring how long it takes for half of the existing mRNA particular types of abnormal mRNAs during their transla-
molecules to be degraded. The most common method tion and prevent the production of truncated or erroneous
involves treating cells with Actinomycin D, an inhibitor of proteins. Nonsense-mediated decay (NMD) detects mRNAs
RNA polymerase II elongation, and quantifying mRNA that have a premature translation stop site, non-stop decay
abundance by Northern blot or reverse transcription–PCR (NSD) detects mRNAs that lack a translation stop site, and
analysis (discussed in Chapter 10). no-go decay (NGD) detects mRNAs that contain sequences
or structures such as strong stem-loops that stall transla-
mRNA decay mechanisms tion elongation. Decay of abnormal mRNAs involves many
of the same enzymes as decay of normal mRNAs, including
mRNA decay typically occurs in the cytoplasm and requires
decapping enzymes, deadenylases, and exonucleases, but it
translation. Most mRNA decay occurs by two general path-
can also involve endonucleases.
ways that are both initiated via removal of the poly(A) tail by
a deadenylase, a special type of exonuclease that specifically
cleaves phosphodiester bonds between adenosine nucleotides KEY CONCEPT After removal of the poly(A) tail, mRNA decay
one at a time in the 3′-to-5′ direction (Figure 8-25). Deade- by specialized enzymes occurs in both the 5′-to-3′ and 3′-to-5′
nylation is sometimes followed by removal of the 5′ m7G cap direction.
by the decapping enzyme Dcp1/Dcp2. This enzyme cuts off
5′-m7Gpp (see Figure 8-16a), leaving behind an mRNA with
a 5′-monophosphate that serves as a substrate for complete The discovery of RNA interference (RNAi)
digestion by a 5′-to-3′ exonuclease Xrn1. Alternatively, dead- In 2002, one of the leading science journals, Science maga-
enylation is followed by digestion by a 3′-to-5′ exonuclease zine, named “Small RNA” as their Breakthrough of the Year.
called the exosome. Following decay of the mRNA body by The RNAs to which they were referring were not the pre-
the exosome, a different decapping enzyme called the scav- viously described small RNAs such as snRNAs or tRNAs,
enger decapping enzyme (DcpS) catalyzes the hydrolysis of which are considered to have housekeeping roles and,
the 5′ m7G cap, releasing 5′-m7Gp. The major difference as such, are synthesized all the time (i.e., constitutively).
between decay in bacteria and decay in eukaryotes is that Instead, these other small RNAs are synthesized in response
most decay in eukaryotes is initiated by an exonuclease, not to changes in a cell’s developmental state or its surroundings.
an endonuclease (compare Figures 8-10 and 8-25). We now know that small RNAs are critically important for
(a) Jorgensen: insertion of transgene (b) Fire/Mello: injection of dsRNA (c) Baulcombe: insertion of viral gene
1. Pigment transgene inserted into the 1. ssRNA and dsRNA synthesized 1. Viral gene inserted into tobacco plant.
petunia genome. in the lab.
unc-22 Antisense
dsRNA
Sense Sense
2. Transgenic plants have white sectors 2. ssRNA and dsRNA injected into 2. Plant exposed to virus but remains
in flowers. adult worms. healthy.
Conclusion: dsRNAs produced from Conclusion: injected dsRNAs silence Conclusion: dsRNAs produced from the
the transgene silence expression unc-22 expression. inserted viral gene silence expression
from the transgenic and endogenous of the gene from invading viruses.
pigment genes.
FIGURE 8-26 Three experiments reveal key features of gene silencing. (a) Jorgensen discovered
that a transgene can silence an endogenous petunia gene necessary for floral color. (b) Fire and
Mello demonstrated that dsRNA can selectively silence genes in C. elegans. (c) Baulcombe showed
that plants with a copy of a viral transgene were resistant to viral infection and produced siRNAs
complementary to the viral genome.
the regulation of gene expression and the maintenance of a In a totally unexpected outcome, Jorgensen found that the
stable genome. Studies that led to the discovery of one class transgene suppressed the expression of its own mRNA as
of small RNA called small interfering RNA (siRNA) were well as the mRNA produced from the endogenous pigment
preceded by several reports of unanticipated changes in gene gene (i.e., the one that is normally in the petunia genome).
expression when RNAs were injected into an organism or Therefore, he called the phenomenon co-suppression.
expressed from a transgene (a gene introduced by research- The mechanism underlying co-suppression remained
ers into the chromosomes of an organism). a mystery until 1998, when Andrew Fire and Craig Mello
One of the greatest joys of doing scientific research is used the roundworm C. elegans to test the hypothesis that
observing a completely unexpected result. In 1990, this is double-stranded RNA (dsRNA) was the agent that trig-
precisely what occurred for Richard Jorgensen in his stud- gered co-suppression. They injected C. elegans with RNAs
ies of plant coloration. To increase pigmentation in petunia that were identical in sequence to an endogenous gene that
plants that had pale pink flowers, Jorgensen inserted a petu- when mutated causes adult worms to twitch (Figure 8-26b).
nia gene that encodes an enzyme necessary for the synthesis If the injected RNAs triggered co-suppression, they expected
of purple-blue flower pigment (Figure 8-26a). As a control, he to see the twitcher phenotype even though the endogenous
inserted the same gene into plants with deep violet flowers. gene was intact. Indeed, they found that, relative to injection
He expected that relative to the parental plants, the flower of single-stranded RNA, injection of dsRNA caused a much
color of the pale pink transgenic plants would be darker, stronger twitcher phenotype, demonstrating that dsRNA
but the deep violet transgenic plants would be unchanged. mediates suppression of endogenous gene expression in a
However, none of the transgenic plants had darker flowers. process that is now called RNA interference (RNAi). The
In fact, all the transgenic plants produced flowers that were study by Fire and Mello also uncovered other remarkable
either pure white or a variety of white patterns (Figure 8-27). features of RNAi: (1) RNAi is very specific, and only RNAs
FIGURE 8-27 (a) The wild-type (no transgene) phenotype. (b and c) So-called co-suppression
phenotypes resulting from insertion of a transgene that controls pigmentation into the genome of a
wild-type petunia. In white regions of the petals, both the transgene and the endogenous chromosomal
copy of the same gene have been inactivated. [Richard Jorgensen, Department of Plant Biology, Carnegie
Institution for Science.]
with perfect complementarity to the dsRNA are affected; (2) involves three main components: (1) small interfering
RNAi is extremely potent, as only a few dsRNA molecules RNAs (siRNAs) that provide the specificity of RNAi by
are required per cell to inhibit expression of the targeted base pairing to target RNAs, (2) Dicer, an endonuclease
gene, indicating that the process is catalytic; (3) RNAi can that precisely cuts dsRNAs into siRNAs, and (3) Argonaute
affect cells and tissues that are far removed from the site (Ago), an RNA endonuclease that is programmed to cut
of introduction, indicating that there is an RNA transport RNAs that base pair to bound siRNAs.
mechanism; and (4) RNAi affects the progeny of injected ani- siRNAs are approximately 21-nucleotide dsRNAs.
mals, indicating that the targeting information is heritable. Each strand of an siRNA has a 5′-monophosphate, a
Many labs continue to study the mechanism underlying 3′-hydroxyl, and a two-nucleotide 3′ overhang beyond
RNAi, which is discussed in the next section. Nevertheless, the core base-paired region of 19 nucleotides (Figure 8-28).
even without a complete understanding of the mechanism, These features of siRNAs are important for their recogni-
RNAi has had a tremendous impact on almost all fields tion by proteins that carry out RNAi. Dicer uses its PAZ
of biology research through its use as a tool to perform domain to bind the 3′ overhang and generate an siRNA
loss-of-function experiments. Researchers have developed from a hairpin or long dsRNA via its two endonuclease
creative methods to introduce or express dsRNAs in cells domains (Figure 8-29). Once bound to the end of a dsRNA,
and whole organisms to reduce the expression of a specific the endonuclease domains are positioned to make cuts in
gene and determine its necessity for molecular, cellular, and the strands that are 21 nucleotides away and staggered
organismal processes. The terms knockdown and silencing by two nucleotides. Dicer can repeat this process, produc-
are used in conjunction with RNAi because the reduction ing multiple siRNAs from a single dsRNA. One of the 3′
in the abundance of targeted RNAs is rarely complete; overhangs of siRNAs is then bound by the PAZ domain
instead, it is knocked down or silenced. of Ago, which is part of a multi-protein complex called
RNAi technologies have also been developed to per-
form genome-wide screens for genes involved in cellular
processes. In this approach, libraries of dsRNAs are gener- siRNA structure
ated that target all of the protein-coding genes in an organ- 2-nt 2-nt
ism, and screens are performed to identify the few dsRNAs overhang 19 bp overhang
that produce a desired phenotype. Thus, RNAi has made it
Passenger strand (same sequence as target mRNA)
much easier to perform genetic studies in many organisms
5′-PO4 OH-3′
and cell culture systems for which there had been no simple
method to manipulate gene expression. 3′-OH PO4-5′
Guide strand (complementary sequence to target mRNA)
siRNA-mediated RNA decay and FIGURE 8-28 Small interfering RNAs (siRNAs) that are produced
transcriptional silencing from hairpin or long dsRNAs by Dicer have specific features that are
important for recognition by Ago. siRNAs are 19–21 base pairs in
RNAi silences gene expression by targeting RNAs for decay length and each strand has a 5′-phosphate and a 2-nucleotide
in the cytoplasm of cells. The RNAi decay mechanism 3′ overhang with a 3′-hydroxyl.
Cleavage by Dicer into siRNAs In some cases, siRNAs can enter the nucleus and block
and assembly into RISC
transcription of target genes by inducing heterochroma-
tin formation (discussed in Chapter 12). For example, in
RISC C. elegans, siRNAs are transported from the cytoplasm
Passenger strand PO4 AGO OH to the nucleus by an Ago protein that lacks endonucle-
Guide strand HO PO4 ase activity. The Ago-siRNA complex localizes to a spe-
Passenger strand is cific gene, presumably by base pairing of the siRNA to
cleaved and removed an mRNA during its synthesis. This localization leads to
recruitment of a methyltransferase enzyme that methylates
histone H3 on lysine 9 (H3K9me) to generate heterochro-
AGO matin and turn off transcription of the gene.
PO4
SUMMARY
RNAs serve numerous and varied purposes in bacteria signals into a nascent mRNA, and multiple mechanisms are
and eukaryotic cells. mRNA garners much of the attention involved in recognizing the termination signals and releas-
in this book and elsewhere because it is the template for ing the mRNA and RNA polymerase from DNA.
the synthesis of proteins, which perform the vast major- In eukaryotic organisms, RNAs undergo extensive pro-
ity of structural and enzymatic roles in cells. Nevertheless, cessing, often while they are being transcribed. For example,
ncRNAs are also important. For example, the translation mRNAs are modified with a cap at the 5′ end and a poly(A)
of mRNA into proteins cannot take place without tRNAs tail at the 3′ end, and introns are removed. Modifications
and rRNAs, and ncRNAs function at many steps in gene at the ends increase the stability of an mRNA and assist
expression. ncRNAs operate in three general ways: they translation. Sequences within a pre-mRNA together with
interact with other RNAs and DNA by base pairing, snRNA or protein factors that bind them define sequences
they serve as scaffolds for the assembly of proteins, and as intron or exon for splicing as well as dictate the site of
they carry out enzymatic reactions. The activities of cleavage and polyadenylation. Additional factors that bind
ncRNAs are integral to transcription and RNA processing other RNA sequence elements can enhance or suppress the
as well as other events in gene expression, including trans- use of particular sites for splicing. This leads to alternative
lation (Chapter 9). For example, in ribosomes (the pro- splicing, which can increase the types of proteins encoded
tein synthesis machines), sequences in tRNAs base pair to by a gene from one to, in some cases, thousands. In bacteria,
sequences in mRNA, rRNAs serve as scaffolds for assembly ribosomes associate with mRNAs as they are being tran-
of ribosomal proteins, and rRNAs catalyze peptide bond scribed, whereas in eukaryotes ribosome association and
formation between amino acids. The variety of RNA func- translation can take place only after mRNAs are exported
tions is possible because of RNA’s unique chemical features. from the nucleus to the cytoplasm.
In bacteria, all RNAs are synthesized by a single RNA The last step in the life cycle of an RNA is decay. mRNA
polymerase, whereas in eukaryotic cells there is a division of decay in both bacteria and eukaryotes occurs via defined
labor among three RNA polymerases (I, II, and III). Regard- pathways that begin with recruitment of specific enzymes.
less of the organism or the RNA polymerase, the sites of In bacteria an endonuclease is recruited by interacting with
transcription initiation are marked by conserved promoter a 5′-monophosphate, and in eukaryotes an exonuclease is
elements that are located near the transcription start site recruited by interacting with proteins that associate with
and are bound by proteins. In bacteria, a subunit of RNA the 3′ UTR. The initiating steps of decay generate recog-
polymerase binds promoter elements, but in eukaryotes, nition sites for further decay by other enzymes. Addition-
GTFs bind promoter elements and recruit a specific RNA ally, in some eukaryotes, very short RNAs such as siRNAs
polymerase. After RNA polymerase recruitment, the DNA base pair to mRNAs and bring along an endonuclease that
is locally unwound and RNA polymerase begins incorpo- initiates decay. One of the normal functions of siRNAs is
rating ribonucleotides in the 5′-to-3′ direction that are com- to silence the expression of repetitive genes in genomes
plementary to the template DNA strand. As transcription such as transposons. Researchers have taken advantage
transitions from the initiation to the elongation phase, the of this decay activity to perform targeted loss-of-function
factors associated with RNA polymerase change. This is experiments.
exemplified by sigma factors in bacteria and factors that The recent development of technologies that detect
associate with different modified forms of the RNA poly- low-abundance RNAs has led to the discovery of thou-
merase II CTD in eukaryotes. In the case of the CTD, asso- sands of RNAs of unknown function. However, researchers
ciated factors are involved in processing of the pre-mRNA have a head start in figuring out how these RNAs are tran-
and can alter the elongation rate of RNA polymerase II, scribed, processed, transported, and decayed as well as how
including causing RNA polymerase II to pause at particu- they function, because it is likely that aspects of the mech-
lar sites. In both bacteria and eukaryotes, RNA polymerase anisms are shared with mRNAs, tRNAs, rRNAs, snRNAs,
terminates transcription after transcribing termination or siRNAs.
KEY TERMS
3′ splice site (p. 285) alternative splicing (p. 288) deadenylase (p. 292)
3′ untranslated region (3′ UTR) (p. 275) branch point (p. 285) decapping enzyme (p. 292)
7-methylguanosine (m7G) (p. 283) cap (p. 283) decay (p. 276)
5′ splice site (p. 285) carboxy-terminal domain (CTD) deoxyribose (p. 270)
5′ untranslated region (5′ UTR) (p. 274) (p. 277) double-stranded RNA (dsRNA)
allosteric termination model (p. 282) consensus sequence (p. 274) (p. 293)
downstream (p. 274) phosphodiester bond (p. 273) sigma factor (σ) (p. 274)
elongation (p. 273) poly(A) polymerase (PAP) (p. 284) small interfering RNA (siRNA) (p. 271)
endonuclease (p. 276) poly(A) tail (p. 284) small nuclear ribonucleoprotein
exon (p. 284) polyadenylation (p. 284) (snRNP) (p. 285)
exonuclease (p. 276) precursor RNA (pre-RNA) (p. 277) small nuclear RNA (snRNA) (p. 271)
factor-independent termination (p. 275) preinitiation complex (PIC) (p. 280) spliceosome (p. 285)
general transcription factor (GTF) promoter (p. 274) splicing (p. 284)
(p. 277) proteome (p. 288) TATA box (p. 279)
half-life (p. 276) Rho-dependent termination (p. 275) template strand (noncoding strand)
helicase (p. 276) ribose (p. 270) (p. 272)
initiation (p. 273) ribosomal RNA (rRNA) (p. 271) termination (p. 273)
intron (p. 284) ribozyme (p. 287) torpedo termination model (p. 282)
isoform (p. 288) RNA editing (p. 290) transcript (p. 272)
long noncoding RNA (lncRNA) RNA interference (RNAi) (p. 293) transcription (p. 268)
(p. 271) RNA polymerase (p. 273) transcription bubble (p. 272)
messenger RNA (mRNA) (p. 270) RNA polymerase I (p. 277) transcription start site (p. 274)
microRNA (miRNA) (p. 271) RNA polymerase II (p. 277) transcriptome (p. 290)
noncoding RNA (ncRNA) (p. 270) RNA polymerase III (p. 277) transfer RNA (tRNA) (p. 271)
non-template strand (coding strand) RNA polymerase core enzyme (p. 274) transgene (p. 293)
(p. 272) RNA polymerase holoenzyme (p. 274) upstream (p. 274)
nucleolus (p. 278) RNA processing (p. 277) uracil (U) (p. 270)
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
WORKING WITH THE FIGURES 10. In Figure 8-9, write the sequence of the DNA tem-
(The first 30 questions require inspection of text figures.) plate and non-template strands that encode the factor-
independent mRNA stem-loop termination signal.
1. In Figures 8-1a and b, draw a generic mRNA at each
stage in the life cycle. 11. In Figure 8-10, why is pyrophosphohydrolase required
only for the first endonucleolytic cleavage by RNase E
2. In Figure 8-2, if the chase was continued for a longer
and not for subsequent cleavage events?
period of time, how would the distribution of radioac-
tive RNAs change, and why? 12. In Figure 8-11, why are RNA polymerase II or III not
recruited to rDNA promoters?
3. In Figure 8-3 , draw the ribonucleotide uridine-5′-
monophosphate (UMP) base paired to adenosine-5′- 13. In Figure 8-12, what GTF is most likely to recognize
monophosphate (AMP). promoters that lack a BRE, TATA box, Inr, and DPE?
Provide a rationale for your answer.
4. In Figure 8-4, what would have to happen in a cell to
convert uracil into uridine triphosphate? 14. In Figure 8-13, which GTF could be mutated to block
transcription of all three types of RNA polymerase III
5. In Figure 8-5a, place an arrow at the location of the
genes?
transcriptional promoter for each gene.
15. In Figure 8-14 , what happens to the CTD between
6. In Figure 8-5c, put a circle around 5′ ribose carbons
the third and fourth steps in the model? What type of
and a square around 3′ ribose carbons.
enzyme carries out this reaction?
7. In Figure 8-6, are adjacent rDNA genes transcribed in
16. In Figure 8-15, 3′-end formation is “coupled” to tran-
the same direction, or in different directions?
scription termination. Explain what this means in the
8. In Figure 8-7b, write the sequence of the first 10 nucle- context of the allosteric and torpedo models.
otides of the rrn D1 transcript.
17. In Figure 8-16a, does the identity of the first nucleo-
9. In Figure 8-8, what change occurs to RNA polymerase tide in the RNA chain (A, C, G, or U) affect capping?
as it initiates transcription? Why or why not?
18. In Figure 8-17, which of the sequence elements that 38. Write the sequence of the template and non-template
regulate cleavage and polyadenylation are retained in strands of DNA that encode the following fragment of
the mRNA after the reaction is complete? a bacterial mRNA:
19. In Figure 8-18, mutations of which five intron nucle- pppGUUCACUGGGACUAAAGCCCGGGAACUAGG
otides is most likely to block splicing? What data sup- 39. Write the sequence of the template and non-template
port your hypothesis? strands of DNA that encode the following eukaryotic
20. In Figure 8-19a, what snRNPs are in the spliceosome mRNA, where the underlined sequence is the poly(A)
when the catalytic steps of splicing occur? tail:
21. In Figure 8-20, what is the primary difference between m7GpppGUUCACUGGGACUGAAUAAAGGGAAC-
Group I and Group II self-splicing? UAGGA AAAAAAAAAAAA(n = 150)
22. In Figure 8-21, what other mRNAs could be produced 40. Draw the possible alternative splicing products of the
by alternative splicing of the blue and green exons? following pre-mRNA, where the white boxes are con-
23. In Figure 8-22, which exons are spliced by a mutually stitutive exons and the shaded boxes are alternative
exclusive mechanism? exons:
50. If you had the sequence of an mRNA and the genome 62. The following data represent the base compositions
of a new organism, how would you determine the of double-stranded DNA from two different bacterial
location in the genome of the transcription start site of species and their RNA products obtained in experi-
the gene that encodes the mRNA? ments conducted in vitro:
51. Describe two functions for ATP and GTP in the pro-
duction of RNA. Species (A + T) (A + U) (A + G)
52. Draw base-paired passenger and guide strands of an (G + C) (G + C) (U + C)
siRNA that could be used to knock down the follow- Bacillus subtilis 1.36 1.30 1.02
ing mRNA: E. coli 1.00 0.98 0.80
5′-AAGUCCGGCAAUGCGACCAAGUCGUAAGCU-
UUAGGCGUCUUGGCAAAGA-3′ a. From these data, determine whether the RNA of
53. In bacteria and eukaryotes, describe what else is hap- these species is copied from a single strand or from
pening to an mRNA while RNA polymerase is synthe- both strands of the DNA. Draw a diagram to show
sizing it from the DNA template. how you solve this problem.
54. Based on the experiment used to test the require- b. How can you tell if the RNA itself is single-
ment for base pairing between the U1 snRNA and the stranded or double-stranded?
5′ splice site (Figure 8-18c), how would you test the 63. Researchers performed a genetic screen for genes that
requirement of base pairing in the stem-loop structure increase the lifespan of C. elegans. They sequenced the
for Rho-dependent termination (Figure 8-9a)? complete genome of a mutant with a longer lifespan
55. In Figure 8-29, propose how the 5′ mRNA fragment and found a single A-to-T base change. List the possi-
produced by RNAi is decayed to nucleotides. ble ways in which the A-to-T change could alter gene
expression to produce the longer lifespan phenotype.
56. What makes poly(A) polymerase an unusual nucleic
For example, the A-to-T could change the amino acid
acid polymerase?
sequence of a protein.
57. A researcher sequenced an RNA and found that there
was a G in a position where there was an A in the
genome. What is likely to have happened to the RNA? GENETICS AND SOCIETY
58. Which of the types of RNA polymerase III genes is In 2018, the first therapy based on RNAi (RNA interfer-
most likely to also be transcribed by RNA polymerase ence) was approved by the US Food and Drug Adminis-
II, and why? tration (FDA). A pharmaceutical company used RNAi to
59. Why is single-stranded RNA less stable in a test tube treat hereditary transthyretin amyloidosis, a progressive
than single-stranded DNA? and often fatal disease caused by an autosomal dominant
mutation in the transthyretin gene that makes a toxic form
60. List four similarities and four differences between
of the transthyretin protein. The drug silences expression
eukaryotic mRNAs and ncRNAs.
of the transthyretin gene using an siRNA (small interfering
RNA) that targets the transthyretin mRNA for destruction.
CHALLENGING PROBLEMS
Based on your knowledge of how RNAi works, why do you
61. What information argues for and against the possi- think that this first RNAi drug has provided great optimism
bility that DNA is directly used as a template for that RNAi-based therapeutics will become a widespread
translation? approach to address genetic diseases?
9.4 TRANSLATION
LO 9.5 Outline the molecular events that take place during translation initiation,
elongation, and termination.
301
R
NAs play several important roles in translation. As you and tRNAs. First, they are much more stable than mRNAs,
learned in Chapter 8, RNAs are classified as either mes- so they remain intact much longer. Second, ribosomes are an
senger RNA (mRNA) or non-coding RNA (ncRNA). abundant component of cells. There are tens of thousands of
The majority of genes encode mRNAs, whose function is to ribosomes in bacterial cells, about 200,000 ribosomes in yeast
serve as an intermediate in the synthesis of proteins. In con- cells, and several million ribosomes in mammalian cells.
trast, ncRNAs are active as functional RNAs; they are never Components of the translational machinery and the
translated into proteins. ncRNAs involved in protein synthesis process of translation are very similar in bacteria and
include transfer RNAs and ribosomal RNAs. Transfer RNAs eukaryotes. In addition to ribosomes, tRNAs, and mRNAs,
(tRNAs) carry out the decoding work of translation, associat- each phase of translation involves a distinct set of protein
ing three-nucleotide sequences in an mRNA with their corre- regulatory factors; initiation factors (IFs) start translation
sponding amino acids. This decoding occurs inside ribosomes, at the beginning of the mRNA open reading frame (ORF),
which are composed of several types of ribosomal RNAs elongation factors (EFs) maintain translation through the
(rRNAs) and many different proteins (Figure 9-1). Ribosomes ORF, and termination factors, also called release factors
assemble on mRNAs and catalyze protein synthesis by chemi- (RFs), stop translation at the end of the ORF.
cally binding together the amino acids brought to the ribosome The major feature that distinguishes translation in bacte-
by tRNAs. Like tRNAs, ribosomes are general in function, in ria from that in eukaryotes is the location where transcription
the sense that they can translate any mRNA. and translation take place in the cell: the two processes take
Although most genes encode mRNAs, ncRNAs make up place in the same compartment in bacteria, whereas they are
the largest fraction of total cellular RNA. In a typical actively physically separated in eukaryotes by the nuclear membrane.
dividing eukaryotic cell, rRNA and tRNA account for almost After processing, eukaryotic mRNAs are exported from the
95 percent of the total RNA, whereas mRNA accounts for only nucleus for translation by ribosomes that reside in the cyto-
about 5 percent. Two factors explain the abundance of rRNAs plasm. In contrast, transcription and translation are coupled
Large
subunit
Decoding
center
Small
subunit
302
in bacteria: translation of an RNA begins at its 5′ end while composed of building blocks called amino acids. In other
the rest of the mRNA is still being transcribed. words, a protein is a chain of amino acids. Because amino
acids were once called peptides, a chain is sometimes referred
KEY CONCEPT Translation occurs within ribosomes and to as a polypeptide. Amino acids have the general formula
requires three types of RNAs: mRNAs carry the sequence
Amino group H Carboxyl group
information from DNA to ribosomes, tRNAs decode mRNA
nucleotide sequences into amino acids, and rRNAs are struc- H3N+ C COO−
tural and functional components of ribosomes.
R
9.1 PROTEIN STRUCTURE All amino acids have two functional groups (an amino
group and a carboxyl group, shown above) bonded to the same
LO 9.1 Explain how the interactions of amino acids carbon atom (called the α carbon). Also attached to the α car-
determine the structure of proteins. bon is a hydrogen (H) atom and a side chain, known as an
R (reactive) group. There are 20 common amino acids that can
Before considering how proteins are made, let’s start with a make up proteins, each amino acid having a different R group
discussion of the structure of proteins. Proteins are polymers that gives it unique properties (Figure 9-2). The side chains are
Nonpolar
H H
H H H3N+ C COO− H2N+ C COO−
H3N+ C COO− H3N+ C COO− CH H2C CH2
H CH3 CH3 CH3 CH2
Lysine Arginine
H H
(Lys/K) (Arg/R)
H H3N+ C COO− H H3N+ C COO−
H H3N+ C COO− CH2 H3N+ C COO− CH2
+ − CH2
H3N C COO CH2 CH2 CH2
+ C
CH2 C C HN
SH H2N O H2N O NH O− O
FIGURE 9-2 The chemical structures of the side chains (R groups) of charge. Each amino acid is labeled with its full name, abbreviation, and
the 20 common amino acids are grouped based on their polarity and single letter designation such as Glycine, Gly, and G, respectively.
categorized into four groups based on their chemical proper- The peptide bond
ties: nonpolar, polar/uncharged , polar/positively charged,
and polar/negatively charged. As described in this chapter, the (a)
R1 H
chemical properties of side chains play a role in determining the O O
folded structures of proteins. H3N+ C C + H3N+ C C
O− O−
In proteins, amino acids are linked together by cova- H R2
lent bonds called peptide bonds. A peptide bond is formed
by linkage of the amino group (NH3+ ) of one amino acid
H2O
with the carboxyl group (COO− ) of another amino acid
(Figure 9-3). One water molecule is removed during the reac-
tion. Because of the way in which peptide bonds form, a R1 H H
polypeptide chain always has an amino end (N-terminal Amino O Carboxyl
(N-terminal H3N+ C C N C C (C-terminal
end) and a carboxyl end (C-terminal end). end) O− end)
H O R2
KEY CONCEPT Each of the 20 common amino acids has
Peptide
an amino group and a carboxyl group as well as a side chain bond
(R group) whose different chemical and physical properties
determine the structure and function of proteins. In poly-
peptides, amino acids are linked together by a peptide bond (b)
H
between the carboxyl group of one amino acid and the amino
group of the next.
R1 H
N
C
Protein structures have four levels of organization, illus-
trated in Figure 9-4. The linear sequence of amino acids in a C C
Peptide
protein constitutes the primary structure. Local regions of the bond
H R2
protein fold into specific shapes, called secondary structures.
Each shape arises from bonding forces between amino acids,
O
including several types of non-covalent interactions, notably
electrostatic forces such as hydrogen bonds, van der Waals
forces (a type of electrostatic interaction involving dipoles), FIGURE 9-3 (a) Peptide bonds form between amino acids by the
removal of water. R1 and R2 represent R groups (side chains) of amino
and hydrophobic effects (i.e., the tendency of nonpolar mole-
acids. (b) Peptide bonds are rigid planar units with R groups projecting
cules to gather together to exclude water molecules). The most out from the carbon-nitrogen (C–N) backbone.
common secondary structures are the α-helix and the β-sheet ANIMATED ART Peptide bond formation
(Figure 9-4b). Proteins can contain neither, one, or both of these
structures. There are 3.6 amino acids per turn in an α-helix ,
which means that each amino acid occupies 100 degrees of
rotation (360 degrees/3.6). The α-helix structure is stabilized of side chains. Within the active site, side chains of certain
by hydrogen bonds between carbonyl oxygen atoms (C=O) amino acids are strategically positioned to interact with a
and amide groups (NH) four amino acids away. β-sheets con- substrate and catalyze a specific chemical reaction.
sist of pairs of β-strands (stretches of 3–10 amino acids in an Most proteins have a hydrophobic core containing non-
extended conformation) lying side by side that are kept together polar amino acid side chains. In contrast, the surface of pro-
by inter-strand hydrogen bonds, again between carbonyl oxy- teins, which is exposed to the aqueous environment, is made
gen atoms and amide groups. Antiparallel β-strands are ori- up of polar amino acids, including those that are positively or
ented (N-terminus to C-terminus) in opposite directions, as in negatively charged. The surface location of polar amino acids
Figure 9-4b, and parallel β-strands are oriented in the same such as serine, threonine, tyrosine, lysine, and arginine makes
direction. Lastly, turns composed of a few amino acids and them accessible to post-translation modification by enzymes, a
loops of longer stretches of amino acids often connect α-helices topic discussed later in this chapter. The nonpolar amino acid
to α-helices, β-strands to β-strands, and α-helices to β-strands. proline is unique among all amino acids because it incorporates
Tertiary structures are the overall three-dimensional the amino group into the side chain (see Figure 9-2). Prolines
shape of an entire polypeptide. In addition to noncovalent are infrequently found in the middle of α-helices and β-sheets
interactions, tertiary structures can be stabilized by cova- because they are unable to contribute to the hydrogen bonding
lent disulfide bridges between cysteine side chains. Cysteine pattern of the helices. Instead, prolines are often found in turns
is the only amino acid whose side chain can form a cova- and loops, at the ends of α-helices, and in the edge strands of
lent bond. Each enzyme has a pocket called the active site β-sheets.
into which its substrate or substrates fit. The active sites of The folding of polypeptides into their correct conforma-
enzymes are good illustrations of the precise interactions tion will be discussed at the end of this chapter. At present,
R2 R4 R6 R8 R10
β-strand C H H H H H
O C O C O C O C O C
C N C N C N C N C N C N C N C N C N C N
O C H C H C C C
O O H O H O H
R1 R3 R5 R7 R9
R2 R4 R6 R8 R10
β-strand H H H H
O O H O O O
C N C C N C C N C C N C C N C C
C N C N C N C N C N
C C C C C
O H O H O H O H O H
R1 R3 R5 R7 R9
β-sheet
Heme β β
Heme group
β polypeptide α
α
FIGURE 9-4 Proteins have four levels of structure. (a) Primary structure: the overall three-dimensional shape of a polypeptide. In
structure: the sequence of amino acids. (b) Secondary structure: the hemoglobin, heme is a non-protein molecule. (d) Quaternary structure:
three-dimensional form of regions of a protein. Examples are α-helices the arrangement of polypeptides in a protein complex. As an example,
and β-sheets. Antiparallel β-sheets have two polypeptide segments hemoglobin is composed of four subunits, two α subunits and two
arranged in opposite polarity, as indicated by the arrows. (c) Tertiary β subunits.
the rules by which primary structure is converted into 5′ to 3′ corresponds to the amino acid sequence in protein
secondary and tertiary structures are imperfectly under- going from N-terminus to C-terminus.
stood. However, from knowledge of the primary amino If genes are segments of DNA, and if a strand of DNA
acid sequence of a polypeptide, the functions of specific is just a string of nucleotides, the sequence of nucleotides
regions can be predicted. For example, some characteristic must somehow dictate the sequence of amino acids in pro-
polypeptide sequences are the contact points with mem- teins. How does DNA sequence dictate protein sequence?
brane phospholipids that position a protein in a membrane. Simple logic tells us that, if nucleotides are the “letters” in a
Other characteristic sequences act to bind DNA or RNA. code, a combination of letters can form “words” represent-
Amino acid sequences or protein folds that are associated ing different amino acids. However, in the 1960s, research-
with particular functions are called domains. A polypeptide ers were faced with many questions about how the code is
may contain one or more separate domains. read. How many letters make up a word, or codon, in the
Lastly, quaternary structure refers to how polypep- code? Are codons overlapping or nonoverlapping? Is the
tides interact with one another to form a multi-polypeptide code continuous or discontinuous? Which codon or codons
protein complex. Individual polypeptides in complexes are represent each amino acid? The cracking of the genetic
called subunits and are joined together by weak bonds. code is the story told in this section.
Quaternary associations can be between different types of
polypeptides (resulting in a heterodimer, if there are two A degenerate three-letter genetic code
subunits) or between identical polypeptides (making a specifies the 20 amino acids
homodimer). Hemoglobin is an example of a heterotetramer
If an mRNA is read from one end to the other, only one of
(tetramer meaning four subunits), composed of two copies
four different bases, A, C, G, or U, is found at each position.
each of two different polypeptides (Figure 9-4d).
Thus, if the words encoding amino acids were one letter
There are two general types of proteins, globular and
long, only four words would be possible. This vocabulary
fibrous. Globular proteins have a compact, round shape and
cannot be the genetic code because there must be a word
play functional roles. Enzymes, hemoglobin, and antibodies
for each of the 20 amino acids commonly found in proteins.
are examples of globular proteins. In contrast, fibrous proteins
If the words were two letters long, 4 × 4 = 16 words would
have a long, narrow shape and play structural roles. Collagen
be possible; for example, AU, CU, or CC. This vocabulary
and keratin are examples of fibrous proteins. Collagen is the
is still not large enough. But if the words were three letters
main structural protein found in connective tissues such as skin,
long, 4 × 4 × 4 = 64 words would be possible; for example,
and keratin is involved in the structure of hair and fingernails.
AUU, GCG, or UGC. This vocabulary provides more than
enough words to describe the 20 amino acids. Therefore,
KEY CONCEPT Proteins have four levels of structure. Primary
structure is the sequence of amino acids. Secondary structure codons must consist of at least three nucleotides. However,
is the shape of a region of amino acids, such as α-helices or if all three-nucleotide combinations specify an amino acid,
β-sheets . Tertiary structure is the three-dimensional shape of a the genetic code must be degenerate, meaning that some
whole polypeptide, and quaternary structure is the assembly of amino acids are specified by two or more different triplets.
multiple polypeptides into a protein complex.
The genetic code is nonoverlapping
and continuous
9.2 THE GENETIC CODE The genetic code could either be overlapping or non-
overlapping. Figure 9-5 illustrates these possibilities for a
LO 9.2 Outline the experimental evidence supporting the three-nucleotide, or triplet, code. For a nonoverlapping
rules of the genetic code. code, consecutive amino acids are specified by consecu-
LO 9.3 Describe features of the genetic code that minimize tive code words (codons), and a single nucleotide mutation
effects of point mutations on protein function. would only alter one codon and one amino acid. For an
overlapping code, consecutive amino acids are specified by
The one-gene–one-polypeptide hypothesis of Beadle and codons that have nucleotides in common; for example, the
Tatum (Chapter 5) was the source of the first exciting third nucleotide in one codon could be the second or first
insight into the functions of genes: genes were somehow nucleotide in adjacent codons. In this case, a single nucle-
responsible for the function of enzymes, and each gene otide mutation would alter three codons and three amino
apparently controlled one enzyme. This hypothesis became acids. By 1961, it was already clear that the genetic code
one of the great unifying principles in biology because it was nonoverlapping. Analyses of mutant proteins showed
provided a bridge between the concepts and research tech- that almost all of the time, only one amino acid changed,
niques of genetics and biochemistry. When the structure of which is predicted by a nonoverlapping code.
DNA was deduced in 1953, it seemed likely that there was The genetic code could be continuous or discontin-
a linear correspondence between the nucleotide sequence in uous. In a continuous code, codons are arranged side by
DNA and the amino acid sequence in a protein. It was soon side with no gaps, whereas in a discontinuous code, codons
deduced that the nucleotide sequence in mRNA going from are separated by one or more nucleotides that act to pause
Nonoverlapping versus overlapping genetic codes effects of another mutation so that the bacterium is more
like wild type.
Nonoverlapping code A U GGC U C AG C U U G AC How can these results be explained? If we assume that
the gene is read from one end only, the original insertion or
deletion induced by proflavin could interrupt the normal
aa1 aa2 aa3 reading mechanism that establishes the grouping of bases
to be read as words. For example, if each group of three
bases in an mRNA makes a word, the reading frame might
Overlapping code A U GG C U C AG C U U G AC be established by taking the first three bases from the end
AUG as the first word, the next three as the second word, and so
UGG
forth (Figure 9-6a). In that case, a proflavin-induced inser-
tion of a single base pair in the DNA would shift the reading
GGC frame on the mRNA from that point on, causing all fol-
lowing words to be misread (Figure 9-6b). Such frameshift
aa1 aa2 aa3 mutations could reduce most of the genetic message to gar-
bage. However, the proper reading frame could be restored
by a compensatory deletion somewhere else, limiting the
FIGURE 9-5 Nonoverlapping and overlapping genetic codes would garbage to the segment between the two mutations (Figure
translate into different amino acid sequences. The example uses codons
9-6d). We have assumed here that the original frameshift
with three nucleotides (a triplet code). (Top) In a nonoverlapping code, a
protein is translated by reading nucleotides sequentially in sets of three. mutation was an insertion, but the explanation works just
Each nucleotide is found in only one codon. In this example, the third as well if the original FCO mutation was a deletion and the
nucleotide (G) in the RNA is only in the first codon. (Bottom) In contrast, suppressor was an insertion (Figures 9-6c and d). The few
in an overlapping code, each nucleotide occupies positions in multiple wrong words in the suppressed genotype could account for
codons. In this case, the third nucleotide (G) is found in all three codons.
the fact that the revertants (suppressed phenotypes) did not
look exactly like true wild types. These data demonstrated
that the genetic code is continuous.
translation and restart anew at the next codon. If the code Crick and his colleagues also found that insertions or
was continuous, which turns out to be the case, insertion deletions of two bases produced the mutant phenotype
or deletion of a single nucleotide would cause a shift in the (Figures 9-6e and f); however, a third mutation of the same
reading frame starting at the site of the mutation and con- type restored a wild-type phenotype because it corrected
tinuing to the end of the open reading frame (Figures 9-6b the reading frame (Figures 9-6g and h). This observation
and c). In contrast, if the code was discontinuous, insertion provided the first experimental evidence that a word in the
or deletion of a single nucleotide would only affect one genetic code consists of three successive nucleotides, or a
codon, and this error would not be propagated through the triplet.
rest of the open reading frame.
Convincing proof that the genetic code is continuous KEY CONCEPT The genetic code has the following features:
came from genetic experiments first reported in 1961 by 1. The linear sequence of nucleotides in a gene determines
Francis Crick, Sidney Brenner, and their co-workers. These the linear sequence of amino acids in the encoded protein.
experiments used mutants in the rII locus of T4 phage. 2. A codon of three nucleotides specifies an amino acid.
The use of rII mutations in recombination analyses was 3. The genetic code is degenerate; more than one codon can
discussed in Chapter 6. Phage T4 is usually able to grow specify the same amino acid.
on two different E. coli strains, called B and K. However,
4. The genetic code is nonoverlapping; each nucleotide is
mutations in the rII gene change the host range of the part of only one codon.
phage: mutant phages can still grow on an E. coli B host,
5. The genetic code is continuous; it is read from a fixed
but they cannot grow on an E. coli K host. Mutations
starting point and continues uninterrupted to the end of the
causing the rII− phenotype were induced using a chemical open reading frame.
called proflavin, which causes insertion or deletion of single
base pairs in DNA. Starting with one particular proflavin-
induced mutation called FCO, Crick and his colleagues Cracking the code
found “reversions” (reversals of the mutation) that were Deciphering the genetic code—determining the amino acid
able to grow on E. coli strain K. Genetic analysis of these specified by each triplet—is one of the most exciting genetic
phages revealed that the “revertants” were not identical to breakthroughs to occur since elucidation of the structure of
true wild types. In fact, reversions were found to be due to the DNA. After the necessary experimental techniques became
presence of a second mutation at a different site from that available, the genetic code was cracked quickly.
of FCO, although in the same gene. This second mutation One technical breakthrough was the discovery that single-
“suppressed” mutant expression of the original FCO. Recall stranded RNA can be synthesized in vitro by the enzyme
from Chapter 5 that a suppressor mutation counteracts the polynucleotide phosphorylase. Unlike transcription, no
Mutation Phenotype
(a)
Wild type CAU CAU CAU CAU CAU CAU CAU CAU CAU None rII+
(b)
FCO mutant CAU GCA UCA UCA UCA UCA UCA UCA UCA U + rII−
(c) (C)
FCO mutant CAU CAU CAU AUC AUC AUC AUC AUC AU − rII−
(d) (C)
Suppression of FCO CAU GCA UCA UAU CAU CAU CAU CAU CAU +− rII+
(e)
Two base insertion CAU GCA UCA UGC AUC AUC AUC AUC AUC AU ++ rII−
Two base deletion CAU CAC AUA UCA UCA UCA UCA UCA U −− rII−
(g)
Three base insertion CAU GCA UGC AGU CAU CAU CAU CAU CAU CAU +++ rII+
Three base deletion CAU CAC AUA UCU CAU CAU CAU CAU CAU −−− rII+
FIGURE 9-6 Examination of single, double, and triple mutations by Crick and colleagues demonstrated
that the genetic code is continuous and read in triplets. A wild-type sequence (a) produces the wild-type rII+
phenotype. A single base insertion (b) or deletion (c) causes the rII− phenotype. The phenotype is suppressed by
combining these mutations (d), but not by a second insertion (e) or deletion (f). Suppression restores the normal
reading frame and indicates that the genetic code is continuous. Furthermore, a double insertion or deletion is
suppressed by a third insertion (g) or deletion (h), indicating that the genetic code is read in triplets.
DNA template is needed for this synthesis, so the nucle- assignment of 61 of the 64 codons was the development
otides are incorporated at random. The ability to enzy- by H. Gobind Khorana of methods to chemically synthe-
matically synthesize RNA offered the exciting prospect of size RNAs of defined sequences. In 1968, Nirenberg and
creating specific RNA sequences and then seeing which Khorana were awarded the Nobel Prize for deciphering the
amino acids they would specify. The first synthetic RNA genetic code.
was made using only uracil nucleotides, producing . . . Virtually all organisms use the same genetic code
UUUU . . . [poly(U)]. In 1961, Marshall Nirenberg and (Figure 9-8). There are just a few exceptions in which a small
Heinrich Matthaei mixed poly(U) with the protein- number of codons have different meanings—for example,
synthesizing machinery of E. coli in vitro and observed the in mitochondrial and nuclear genomes of ciliates and some
formation of a protein. The main excitement centered on other protozoans. How is it that genomes can have very
the question of the amino acid sequence of this protein. It different base compositions but use the same code? Part of
proved to be polyphenylalanine—a string of phenylalanine the answer is that degeneracy of the code permits the DNA
(Phe) amino acids (Figure 9-7a). Thus, the triplet UUU codes base composition of genomes to vary over a wide range
for phenylalanine. Expanding this approach to other single and still encode all 20 amino acids. For example, the G + C
nucleotides and combinations of nucleotides led to assign- content of the coding regions of bacterial genomes ranges
ment of about 40 codons to particular amino acids (Fig- from about 20 percent to 70 percent but encodes proteins
ures 9-7b and c). The technical breakthrough that led to with very similar sequences. Furthermore, not all codons
Cracking the genetic code using synthetic RNAs organism can be correctly translated in another organism.
For example, bacteria can be used to make large amounts
(a) UUU UUU UUU UUU UUU Reading frames 1, 2, and 3 of human proteins by translating human mRNA sequences.
Phe Phe Phe Phe Phe
Stop codons
(b) CCC CCC CCC CCC CCC Reading frames 1, 2, and 3 Three of the 64 codons are stop codons, or termination
Pro Pro Pro Pro Pro codons, that, instead of encoding an amino acid, stop
translation. They can be regarded as being similar to peri-
ods punctuating the message encoded in the DNA. One of
(c) GUU GUU GUU GUU GUU Reading frame 1 the first indications of the existence of stop codons came
Val Val Val Val Val in 1965 from Sydney Brenner’s work with the T4 phage.
or Brenner analyzed certain mutations (m1 − m6 ) in a single
gene that controls the head protein of the phage. He found
G UUG UUG UUG UUG UU Reading frame 2
that the head protein of each mutant was shorter than that
Leu Leu Leu Leu
of the wild type. Brenner examined the ends of the short-
or ened proteins and compared them with the wild-type pro-
GU UGU UGU UGU UGU U Reading frame 3 tein. For each mutant, he recorded the next amino acid that
Cys Cys Cys Cys would have been inserted to continue the wild-type chain.
The amino acids for the six mutations were glutamine,
FIGURE 9-7 The genetic code (shown in Figure 9-8) was established lysine, glutamic acid, tyrosine, tryptophan, and serine.
by discovering the amino acids coded by synthetic RNAs containing
repeats of a single nucleotide (a and b) or combinations of nucleotides
These results presented no immediately obvious pattern,
(c). Regardless of the reading frame, repeats of single nucleotides but Brenner deduced that each of these codons can mutate
code for the same amino acid, whereas repeats of different nucleotides to the codon UAG by a single nucleotide change (Figure 9-9).
can code for three different amino acids in the three different reading He therefore postulated that UAG is a stop (termination)
frames. codon—a signal to the translation mechanism that the pro-
tein is now complete. The two other stop codons are UGA
for an amino acid, termed synonymous codons, are used and UAA. Stop codons are often called nonsense codons
with equal frequency in an organism, and this frequency because they do not designate an amino acid.
can change dramatically among organisms. The topic of
synonymous codon usage bias is discussed in greater detail Degeneracy of the genetic code limits the
in Chapter 14. From a practical standpoint, the almost uni- effects of point mutations
versality of the genetic code means that mRNAs from one
The genetic code is not random, and degeneracy of the
code minimizes the potentially detrimental effects of point
The genetic code mutations (single nucleotide mutations). The genetic code
can be thought of as 16 blocks, each containing four entries
Second letter
defined by the first two nucleotides in a codon (Figure 9-10).
U C A G
Eight of the 16 blocks specify only one amino acid, mean-
UUU UCU UAU UGU U ing that mutations in the third position of a codon do not
Phe Tyr Cys
UUC UCC Ser UAC UGC C
U
Stop A
change the encoded amino acid. As an example, GUA, GUC,
UUA UCA UAA Stop UGA
Leu GUG, and GUU all code for valine. Furthermore, mutations
UUG UCG UAG Stop UGG Trp G
in the first position of a codon often lead to a conservative
CUU CCU CAU CGU U
CUC CCC CAC
His
CGC C
change in amino acid—that is, an amino acid with similar
C Leu Pro Arg chemical properties and size. For example, mutation of the
CUA CCA CAA CGA A
Third letter
First letter
Gln
CUG CCG CAG CGG G
AUU ACU AAU AGU U Single nucleotide mutations that produce a UAG stop codon
AUC Ile ACC AAC Asn AGC Ser C
A Thr U AG St o p
AUA ACA AAA AGA A
AUG Met ACG AAG Lys AGG Arg G
GUU GCU GAU GGU U AAG L y s U CG Se r UA A S t o p
GUC GCC GAC Asp GGC C CAG G l n U GG T r p UA C T yr
G Val Ala Gly
GUA GCA GAA GGA A G AG G l u U UG L e u UA U T yr
Glu
GUG GCG GAG GGG G
FIGURE 9-9 The UAG stop codon is produced by a single nucleotide
FIGURE 9-8 The genetic code designates the amino acids specified mutation in any of nine codons, eight that code for amino acids and
by each codon. Sixty-one codons specify amino acids, and three one that codes for another stop codon. Note that a single nucleotide
codons specify translation stop. This code is almost universal among mutation is not sufficient to convert some codons to any of the three
all organisms. stop codons.
Organization of the genetic code model, quickly dismissed as naive and unlikely, proposed
that mRNA codons could fold up and form 20 distinct cav-
Second letter
ities that directly bind specific amino acids in the correct
U C A G
order. Instead, in 1958, Crick recognized the following:
UUU UCU UAU UGU U
Phe Tyr Cys
U UUC UCC Ser UAC UGC C It is therefore a natural hypothesis that the amino
UUA UCA UAA Stop UGA Stop A acid is carried to the template by an adaptor
Leu
UUG UCG UAG Stop UGG Trp G
molecule, and that the adaptor is the part which
CUU CCU CAU
His
CGU U actually fits on to the RNA. In its simplest form
CUC CCC CAC CGC C [this hypothesis] would require twenty adaptors,
C Leu Pro Arg
CUA CCA CAA CGA A
Third letter
First letter
(a) 3′ (b)
OH Amino acid
Amino acid attachment site
attachment site A
(alanine) C 5′ 5′
TψC loop
C P
3′
Acceptor stem
Acceptor stem
TψC loop
D
D
C
ψ T
D loop
Variable loop
Additional residues
Anticodon stem in some tRNAs
Anticodon stem Anticodon loop
Anticodon loop 3′
3′ C G U 5′
Anticodon
5′ Anticodon
mRNA 5′ G C A 3′
Codon for alanine
Unmodified residues in all tRNAs
Modified residues in some or all tRNAs
FIGURE 9-11 (a) The structure of yeast alanine tRNA. The labeled features of the structure
are important for tRNA function and are discussed in detail in the text. (b) Diagram of the three-
dimensional structure of yeast alanine tRNA. [Based on PDB ID 3WQY.]
2. Numerous nucleotides in all tRNAs undergo post- KEY CONCEPT tRNAs have four important structural fea-
transcriptional modification; that is, tRNAs are tran- tures: (1) the sequence CCA at the 3′ end; (2) modified nucleo-
scribed with the standard four nucleotides, which are tides such as dihydrouridine, pseudouridine, and inosine; (3) an
then altered by enzymes. In yeast, each tRNA contains overall inverted L shape; and (4) an anticodon.
between 7 and 17 modified nucleotides. For example,
the D loop contains the nucleotide dihydrouridine (D),
which has two extra hydrogen atoms on the uracil base; Amino acids are attached to tRNAs by enzymes
the TψC loop contains pseudouridine (ψ), which has called aminoacyl-tRNA synthetases. The tRNA with an
uracil attached to the ribose sugar at a carbon rather attached amino acid is said to be charged tRNA. There are
than a nitrogen; and the anticodon can contain the base 20 synthetases, one for each of the 20 amino acids. Because
inosine (I), which is structurally similar to guanosine the code is degenerate, some synthetases act on multiple
(G). Nucleotide modifications in the anticodon affect tRNAs. Charging by aminoacyl-tRNA synthetases occurs
base pairing with the codon (see the discussion about in two steps (Figure 9-12). In the first step, the carboxyl
wobble in the next section), and nucleotide modifica- group of the amino acid reacts with the α-phosphate of ATP
tions at other sites affect tRNA recognition, folding, to form 5′-aminoacyl-AMP and release pyrophosphate (PPi ).
and stability. However, the precise function of many 5′-aminoacyl-AMP is referred to as an activated amino acid.
modifications is yet to be determined. In the second step of charging, the amino acid is transferred
3. A tRNA normally exists as an inverted L-shaped to the adenosine (A) of the invariant CCA sequence at the
three-dimensional structure, as shown in Figure 3′ end of the tRNA, and AMP is released as a by-product.
9-11b, rather than the flattened cloverleaf shown in What would happen if the wrong amino acid was
Figure 9-11a. Although tRNAs differ in their primary covalently attached to a tRNA? A convincing experiment
nucleotide sequence, all tRNAs fold into virtually the answered this question. The experiment used tRNACys, the
same L-shaped conformation, indicating that the shape tRNA specific for cysteine. This tRNA was charged with
of a tRNA is important for its function. cysteine, meaning that cysteine was attached to the tRNA.
CH3 CH3
Step 1 Step 2
Aminoacyl-tRNA Aminoacyl-tRNA
C G U synthetase P P C G U synthetase AMP C G U
P
FIGURE 9-12 In a two-step reaction, aminoacyl-tRNA synthetases charge tRNAs with the correct A N I M ATED A RT
amino acid. In the first step, the amino acid is linked to AMP to form aminoacyl-AMP, with PP i as a
tRNA charging
by-product. In the second step, the tRNA replaces AMP to form aminoacyl-tRNA, with AMP as a
by-product.
The charged tRNA was treated with nickel hydride, which discrimination step, Val-tRNA Ile , but not Ile-tRNA Ile, fits
converted the cysteine (while still bound to tRNACys) into into a separate active site of the synthetase and is hydro-
another amino acid, alanine, without affecting the tRNA: lyzed to valine and tRNA Ile. Because of this proofreading
nickel hydride
cysteine-tRNA Cys → alanine-tRNA Cys mechanism, the error rate of protein synthesis is very low,
in the range of 1 in 104−105 amino acids incorporated.
Proteins synthesized with alanine-tRNA Cys had alanine
wherever cysteine was coded. The experiment demon- KEY CONCEPT tRNAs are charged by tRNA synthetases in a
strated that amino acids are “illiterate”; they are inserted at two-step reaction requiring ATP. There is a different synthetase
the proper position because the tRNA adaptors recognize for each amino acid. Two proofreading steps ensure that syn-
the mRNA codons. Thus, attachment of the correct amino thetases charge tRNAs with the correct amino acid.
acid to its corresponding tRNA (also called its cognate
tRNA) by an aminoacyl-tRNA synthetase is the critical step
in ensuring that the mRNA code is translated correctly. If
the wrong amino acid is attached, there is no way to pre-
Wobble base pairing allows tRNAs
vent it from being incorporated into a growing polypeptide to recognize more than one codon
chain. If perfect Watson-Crick base pairing between tRNA anti-
Correct charging of tRNAs depends on selection of codons and mRNA codons was required to recognize all
appropriately paired tRNAs and amino acids by aminoacyl- of the codons, there would need to be 61 different tRNAs.
tRNA synthetases. These enzymes are good at recogniz- However, this is not the case; some tRNAs can recognize
ing the correct tRNA because tRNAs have numerous multiple codons through a different kind of base pairing
distinguishing structural features, including nucleotide at the third position of a codon, termed the wobble posi-
sequence and nucleotide modifications. However, the only tion. For example, the charged tRNA Ser can form either a
distinguishing feature of amino acids is their side chain, normal Watson-Crick G–C base pair or an unusual G–U
which can be very similar. Therefore, to prevent mistakes, wobble base pair with serine codons (Figure 9-13), so one
aminoacyl-tRNA synthetases have a two-step mechanism tRNA can be used for both serine codons. In addition,
that discriminates between chemically similar amino acids inosine (I), a rare modified base in tRNA, can base pair to
such as valine (Val) and isoleucine (Ile), which differ by only C, U, and A (Table 9-1). Therefore, because of wobble base
a single CH 2 group (see Figure 9-2). The first discrimina- pairing, cells require fewer than 61 tRNAs to read all of
tion step occurs in the activation site of the enzyme where the codons.
the amino acid is bound and activated to form aminoacyl-
AMP. This step rejects amino acids that do not fit into the KEY CONCEPT The genetic code is called degenerate
activation site because they are too large. So, Val-tRNA because, in many cases, more than one codon is assigned to
a single amino acid; in addition, wobble base pairing allows
synthetase will reject Ile because it is too large. In con-
the anticodon of some tRNAs to pair with more than one
trast, Ile-tRNA synthetase will sometimes charge tRNA Ile
codon.
with valine to produce Val-tRNA Ile . But, in the second
U C C U C U
5′ Codon 3′ mRNA 5′ Codon 3′ mRNA
tRNA
Guanine N N Guanine N N
O N O N
N N
H N H H
O N H
H H
H N H N H
N
Uracil O
Cytosine O
N
N
mRNA
28S rRNA
23S rRNA 4718 nt
2904 nt
+
5.8S rRNA
+ 16S rRNA 158 nt 18S rRNA
5S rRNA 120 nt + 1874 nt
1541 nt 5S rRNA 120 nt
+ + + +
31 proteins 21 proteins 49 proteins 33 proteins
FIGURE 9-14 Ribosomes contain a large and a small subunit. Each subunit contains one large
RNA and a set of proteins. In addition, the large subunit of bacterial ribosomes contains one small
RNA, 5S rRNA, whereas (b) the large subunit of eukaryotic ribosomes contains two small rRNAs,
5S rRNA and 5.8S rRNA.
N N N N
N N N N
N N
N N N N
N NNN NN N N N N
N N N N N N
N NN N N N N N
N N NNNNNNNNNNN N
N NN N N N
N N N N
N N N N NN
NNNNNNNNNNN N N N
N N N N N
N N N N N
N N N N
N N N N N N
N NN N
5′ N N N N N
N N A N N
3′ N N C NN N N
N N C Base pairs to the mRNA
N N U
N N C Shine–Dalgarno sequence
N N C
N N U
N N N
N N N
N N
N N 3′
N N
FIGURE 9-15 The folded structure of bacterial 16S rRNA. The magnified area shows the details of
the complex secondary structure of 16S rRNA and the sequence at the 3′ end of 16S that binds an
mRNA Shine–Dalgarno sequence.
A site of the 30S subunit. Proceeding in the 5′ direction on of a peptide bond between their amino acids, disconnecting
the mRNA, the next codon interacts with the anticodon of the P site amino acid from its tRNA. The growing peptide
the tRNA in the P site (for peptidyl site) of the 30S sub- chain fits into a tunnel-like structure in the 50S subunit.
unit. The P and A sites are situated to facilitate formation The E site (for exit site) contains a deacylated tRNA (it no
Key sites of interaction in the ribosome Two additional regions in the ribosome are critical for
protein synthesis. The decoding center in the 30S subunit
(a) Computer model
Polypeptide ensures that only tRNAs carrying anticodons that match
chain the codon will be accepted into the A site. The peptidyl-
transferase center in the 50S subunit is the site where
peptide bond formation is catalyzed. The Nobel Prize
Exit tunnel
50S in Chemistry was awarded in 2009 to Thomas Steitz,
Venkatraman Ramakrishnan, and Ada Yonath for their lab-
oratories’ work using X-ray crystallography to determine
the structure of the ribosome at the atomic level. The results
of their elegant studies clearly show that both the decoding
and peptidyltransferase centers are composed entirely of
E
5′ P A sites rRNA. Therefore, peptide bond formation is thought to be
catalyzed by an active site in the rRNA and assisted only
mRNA
by ribosomal proteins. In other words, the large ribosomal
subunit functions as a ribozyme to catalyze peptide bond
30S formation.
tRNAs
Mo KEY CONCEPT In all organisms, ribosomes have large
ve
me
nt and small subunits, each containing rRNA and proteins. Two
of
rib
os
key sites in ribosomes, the peptidyltransferase center in the
om large subunit that carries out peptide bond formation and the
e
decoding center in the small subunit that accepts the correct
3′ tRNA, are built of rRNA.
+ Growing
(b) Schematic model NH 3 polypeptide
chain
9.4 TRANSLATION
Deacylated
tRNA released LO 9.5 Outline the molecular events that take place during
Peptidyl-
from E site transferase
translation initiation, elongation, and termination.
center
Translation is carried out by ribosomes moving along mRNA
in the 5′-to-3′ direction. tRNAs bring amino acids to the ribo-
some, and their anticodons base pair to mRNA codons. An
CC
C incoming amino acid becomes bonded to the amino end of
the growing polypeptide chain in the ribosome. The pro-
A A AUCG cess of translation can be divided into three phases: initia-
5′ GG G U U U A G C 3′ tion, elongation, and termination. Aside from the ribosome,
mRNA E P A
mRNA, and tRNAs, other proteins (factors) are required for
Decoding center each phase (Table 9-2). Because certain steps in initiation differ
Movement of ribosome significantly in bacteria and eukaryotes, initiation is described
separately for the two groups. The elongation and termina-
FIGURE 9-16 Ribosome interactions during the elongation phase of tion phases are described largely as they take place in bacteria.
translation. (a) A computer model of the three-dimensional structure of
the ribosome with mRNA, tRNAs, and the nascent polypeptide chain
as it emerges from the large ribosomal subunit. (b) A schematic model Translation initiation
of the ribosome during translation elongation. [Part (a) drawn using PDB
IDs 1VSA, 2OW8, and 1GIX.] The main task of initiation is to place the first aminoacyl-
tRNA in the P site of the ribosome and, in this way, establish
the correct reading frame of the mRNA. In most bacteria
longer carries an amino acid) that is ready to be released and all eukaryotes, the first amino acid in any newly syn-
from the ribosome. Whether codon–anticodon interactions thesized polypeptide is methionine (Met), specified by the
also take place between the tRNA and the mRNA in the E initiation codon AUG. In bacteria, there are two tRNAs for
site is not yet clear. methionine. tRNA Met is used at AUGs in internal positions
in mRNA and an initiator tRNA, tRNA fMet , is used at AUG
KEY CONCEPT mRNA base pairs with tRNA in the small initiation codons. tRNA fMet is charged with methionine to
subunit of the ribosome, while tRNAs fit into sites that span form Met-tRNA fMet and then a formyl (f) group is added
both subunits. tRNAs begin in the A site, peptide bond forma-
to methionine to generate N-formylmethionyl-tRNA fMet
tion occurs in the P site, and tRNAs exit from the E site.
(fMet-tRNA fMet ). The formyl group on fMet-tRNA fMet is
removed during or shortly after synthesis of the polypep- The Shine–Dalgarno sequence positions the mRNA in the
tide. Eukaryotes also use distinct methionine tRNAs for ribosome
internal and initiation AUG codons, called tRNA Met and
30S ribosomal subunit
tRNA i Met , respectively. The use of Met-tRNA i Met rather
Shine–Dalgarno Start
than Met-tRNA Met for initiation is specified by interactions sequence codon
with translation initiation factors. mRNA
How does the translation machinery know where to begin? C A U C C U A GG A GGU U UG A U C C U A UGCG
In other words, how is the initiation AUG codon selected from 5′ 3′
among the many AUG codons in an mRNA? Recall that, in U U C C U C C A
16S rRNA A
both bacteria and eukaryotes, mRNA has a 5′ untranslated 3′ 5′
region (UTR) consisting of the sequence between the transcrip-
FIGURE 9-17 In bacteria, base pairing between the 3′ end of the 16S
tion start site and the translational start site (see Figure 8-7).
rRNA of the small ribosomal subunit (30S) and the Shine–Dalgarno
As you will see below, the nucleotide sequence of the 5′ UTR sequence of the mRNA positions the ribosome to correctly initiate
adjacent to the AUG initiation codon is critical for ribosome translation at the downstream AUG initiation codon.
binding in bacteria, and the 5′ cap is critical for ribosome bind-
ing and scanning for the AUG initiation codon in eukaryotes.
first AUG and initiate at the second, third, or subsequent
In bacteria, AUG initiation codons in mRNA are preceded
AUG. Marilyn Kozak found that bypass occurs because
by a special sequence called the Shine–Dalgarno sequence,
the sequence surrounding the AUG codon affects the effi-
also known as the ribosome-binding site (RBS), that base pairs
ciency of its use in initiation. Kozak found that the sequence
with the 3′ end of 16S rRNA in the 30S ribosomal subunit
CC(A−3 /G)CCA+1 UGG+4, called the Kozak sequence, com-
(Figure 9-17). This base pairing correctly positions the AUG in
monly surrounds initiating AUGs, and mutagenesis studies
the P site where the initiator tRNA will bind. The mRNA can
showed that the A at −3 and the G at +4 are particularly
interact only with a 30S subunit that is dissociated from the
important for specifying the AUG initiation codon.
rest of the ribosome. Note again that rRNA performs the key
function in ensuring that the ribosome is at the right place to KEY CONCEPT Translation initiation begins when a charged
start translation. Once the initiator tRNA is bound, the 50S initiator tRNA anticodon and an mRNA AUG initiation codon
ribosomal subunit binds to form the 70S initiation complex. assemble in the P site of a ribosomal small subunit. In bacte-
In eukaryotes, translation initiation involves binding of ria, the Shine–Dalgarno sequence in the 5′ UTR of the mRNA
the 40S ribosomal subunit to the capped 5′ end of an mRNA, base pairs with the 16S rRNA to position the start codon in the
followed by scanning of the 5′ UTR for an AUG initiation P site. In eukaryotes, the 5′ cap of the mRNA is bound by an
codon. The 5′ cap (m 7 G) that is added to mRNAs during initiation factor that recruits the ribosomal small subunit. Once
transcription (described in Chapter 8) is directly bound by bound, the ribosome scans the mRNA for the initiation codon
a translation initiation factor, which in turn binds other ini- within the Kozak sequence.
tiation factors to recruit the small ribosomal subunit to the
mRNA. The ribosome subsequently scans the mRNA in the In bacteria, three proteins—IF1, IF2, and IF3 (for ini-
5′-to-3′ direction until it encounters the first AUG codon. tiation factor)—are required to assemble an active 70S
In 5 to 10 percent of cases, the ribosome will bypass the ribosome (Figure 9-18a). Assembly begins by positioning the
Translation initiation
(a) Bacteria (b) Eukaryotes
eIF5
30S 40S
mRNA
Met-tRNAiMet
eIF2-GTP
Shine–Dalgarno
IF3 IF1
P Met
5′ AUG 3′
5′ UTR 3′ 5′ eIF5 eIF2 GTP
Next
16S rRNA
codon eIF1 eIF1A
Start eIF3 UAC
codon
fMet-tRNAfMet
IF2-GTP
Anticodon
mRNA
fMet eIF4F
IF2 GTP
FIGURE 9-18 Initiation factors assist assembly of the ribosome at the translation start site and A N I M ATED A RT
then dissociate before translation elongation. (a) In bacteria, three initiation factors (IF1, IF2, and IF3)
position fMet-tRNA fMet in the P site of the 70S ribosome. (b) In eukaryotes, many eIFs are required Translation
to position Met-tRNA iMet in the P site of the 80S ribosome. Factors that serve similar functions in
bacteria and eukaryotes are colored the same.
mRNA AUG initiation codon in the 30S ribosomal subunit and is transported along with the mRNA to the cytoplasm
P site through base pairing of the Shine–Dalgarno sequence (Chapter 8). Another eukaryotic-specific factor is eIF1,
with 16S rRNA and by base pairing of fMet-tRNA fMet with which associates with eIF4G and promotes scanning for
the AUG initiation codon. During these early assembly steps, the AUG initiation codon. Finally, as in bacteria, eukary-
IF2 bound to GTP promotes binding of fMet-tRNA fMet to otic initiation factors dissociate from the ribosome before
the P site. In addition, binding of IF1 in the A site blocks the elongation phase of translation begins. The exception
tRNA binding to the second codon, and binding of IF3, is eIF4F, which remains associated with PABP, setting the
with help from IF1, blocks association of the 50S subunit. stage for additional rounds of translation.
Once formation of the 30S initiation complex is complete,
IF3 and IF1 are released, which enables association of the KEY CONCEPT In both bacteria and eukaryotes, initiation
50S subunit, and hydrolysis of GTP to GDP leads to release factors bring the initiator tRNA to the ribosomal small subunit
of IF2 to generate a functional 70S ribosome. Hydrolysis of and prevent premature binding of the large subunit. Some
GTP and sometimes ATP provides energy for conformation initiation factors hydrolyze GTP to proceed through the steps
of initiation. In eukaryotes, additional initiation factors facilitate
changes that are required to progress through several steps
mRNA scanning and circularization.
of translation.
Like bacteria, eukaryotes use a set of initiation fac-
tors (IFs) to assemble the 80S translation initiation com-
plex containing the small and large ribosomal subunits, an Translation elongation
mRNA bound to the small subunit, and a charged initiator During translation elongation, the ribosome functions as
tRNA (Met-tRNA i Met ) in the P site that is base paired to the a factory, repeating the same steps over and over again.
AUG initiation codon (Figure 9-18b). The names of eukary- The mRNA acts as a blueprint specifying the delivery of
otic factors all begin with an “e” to distinguish them from tRNAs, each carrying as cargo an amino acid. Each amino
bacterial factors. Four eIFs have functions similar to IFs in acid is added to the growing polypeptide chain, while the
bacteria. Like bacterial IF1, eIF1A blocks tRNA binding deacylated tRNA is recycled by being charged with another
in the A site of the small subunit (40S) as well as prema- amino acid. Figure 9-19 details the steps in elongation. In
ture association of the large subunit (60S) with the small bacteria, two protein elongation factors (EFs) called elon-
subunit. Eukaryotes have two factors, eIF2 and eIF5B, that gation factor Tu (EF-Tu) and elongation factor G (EF-G)
have functions similar to bacterial IF2. A ternary complex assist the elongation process.
composed of eIF2, GTP, and Met-tRNA i Met associates with As described earlier in this chapter, an aminoacyl-tRNA
the small ribosomal subunit, while eIF5B promotes asso- is formed by the covalent attachment of an amino acid to the
ciation of the small and large subunits. In addition, both 3′ end of a tRNA that contains the correct anticodon. Before
eIF2 and eIF5B use hydrolysis of GTP to GDP to carry out aminoacyl-tRNAs can be used in protein synthesis, they
their jobs. Lastly, like bacterial IF3, eIF3 binds the small associate with the protein factor EF-Tu to form a ternary
ribosomal subunit and blocks association of the large complex composed of EF-Tu, GTP, and aminoacyl-tRNA.
subunit. The elongation cycle commences with fMet-tRNA fMet in the
Because of differences between eukaryotic and bacterial P site and with the A site ready to accept a ternary complex
mRNAs, there are other eIFs that do not have counterparts (Figure 9-19). Codon–anticodon recognition in the decoding
in bacteria. Three eIFs—eIF4A, eIF4E, and eIF4G—interact center of the small subunit determines which of the different
with one another in a complex called eIF4F that carries out ternary complexes to accept (see Figure 9-16b). When the
activities unique to eukaryotes. eIF4E binds the cap struc- correct match has been made, the ribosome changes shape,
ture at the 5′ end of an mRNA. m 7 G caps are present only EF-Tu hydrolyzes GTP to GDP and leaves the ternary com-
on mRNAs, so the requirement for cap binding by eIF4E plex, and the two amino acids are juxtaposed in the peptidyl
ensures that only mRNAs are translated. eIF4A has RNA transferase center of the large subunit (see F igure 9-16b).
helicase activity that unwinds regions in the 5′ UTR that are There, a peptide bond is formed with transfer of fMet in the
double-stranded RNA due to intramolecular base pairing. P site to the amino acid in the A site.
This allow ribosomes to scan along single-stranded RNA At this point, the second elongation factor, EF-G, plays
in search of the AUG initiation codon. Hydrolysis of ATP its part. EF-G is structurally similar to a ternary complex
to ADP by eIF4A is required for its activity. Lastly, eIF4G and fits into the A site, displacing the peptidyl-tRNA.
binds poly(A) binding proteins (PABPs) that associate with Hydrolysis of GTP to GDP by EF-G changes its structure
the mRNA poly(A) tail, thereby bringing together the 5′ as well as that of the ribosome and shifts the tRNAs in the
and 3′ ends of the mRNA. Circularization of the mRNA A and P sites to the P and E sites, respectively. When EF-G
is thought to enhance the rate of translation of capped and leaves the ribosome, the A site is open to accept the next
polyadenylated mRNAs by coordinating the initiation of ternary complex. As elongation progresses, the number of
ribosomes that have recently terminated translation. amino acids on the peptidyl-tRNA (at the P site) increases.
In the first round of translation, the function of eIF4F Eventually, the amino-terminal end of the growing poly-
is carried out by the cap binding complex (CBC), which peptide emerges from the tunnel in the 50S subunit and
associates with newly synthesized mRNAs in the nucleus protrudes from the ribosome.
Translation termination
EF-Tu GTP The elongation cycle continues until the codon in the A
site is one of the three stop codons: UGA, UAA, or UAG.
tRNAs do not recognize these codons. Instead, pro-
E P A
teins called release factors (RFs) recognize stop codons
( Figure 9-20). Just like the structure of EF-G mimics the
GTP hydrolysis EF-Tu GDP
structure of a ternary complex, RF1 and RF2 mimic the
AA2
structure of a tRNA. In bacteria, RF1 recognizes UAA or
UAG, whereas RF2 recognizes UAA or UGA. The interac-
tion between RF1 or RF2 and the A site differs from that of
the ternary complex in two important ways. First, the stop
codons are recognized by tripeptides in the RF proteins, not
E P A by an anticodon. Second, RFs fit into the A site of the 30S
subunit but do not participate in peptide bond formation.
Peptide bond forms Instead, a water molecule gets into the peptidyltransferase
center, and its presence leads to release of the polypeptide
from the tRNA in the P site. Following release of the pep-
tide chain, RF3 promotes release of RF1 or RF2 from the
ribosome. Hydrolysis of GTP to GDP is involved in releas-
ing RF3 from the ribosome. In eukaryotes, translation
termination is very similar; eRF1 recognizes stop codons,
E P A and eRF3 stimulates peptide chain release by eRF1. How-
ever, unlike bacteria, eRF1 recognizes all three stop codons
EF-G GTP (UAA, UAG, and UGA).
Translocation
To prepare for a new round of translation, the ribosome
Pi recycling factor (RRF) disassembles the post-termination
complex (Figure 9-20). With the help of EF-G and IF-3,
RRF binds in the A site, translocates to the P site, releases
deacylated tRNAs from the E and P sites, and dissociates
EF-G GDP the small and large ribosome subunits from each other and
from the mRNA. IF-3 bound to the small subunit is now
ready to initiate translation (see Figure 9-18a).
E P A
P
u GT FIGURE 9-19 In bacteria, two elongation factors, EF-Tu and EF-G,
EF-T
perform repetitive functions for each amino acid that is added to the
growing polypeptide. EF-Tu escorts charged tRNAs to the A site and
positions them for peptide bond formation with the peptide attached
E P A to the tRNA in the P site. The protein factor EF-G then drives the
repositioning of tRNAs from the P and A sites into the E and P sites,
respectively.
RRF
EF-G GTP
9.5 TRANSLATIONAL AND POST-
IF3 TRANSLATIONAL REGULATION
LO 9.6 Describe how protein synthesis and function are
regulated.
(a) Wild type: no mutations. The tyrosine (b) A mutation introduces a UAG stop (c) A further mutation changes the tyrosine
tRNA binds to the codon UAC. codon. Translation stops. UAG stop codon. Translation continues.
Tyr Tyr
Gly Gly Gly
Gln Gln Gln
RF1
AU RF1 AU
G C
GUC GUC GUC
5′ G G G C A G UAC A AG 3′ 5′ G G G C A G UAG A AG 3′ 5′ G G G C A G UAG A AG 3′
Stop Stop
FIGURE 9-21 A suppressor allows translation to continue past a nonsense mutation. (a) In a wild-
type mRNA, a Tyr-tRNA Tyr reads the codon UAC, and translation elongation continues. (b) Mutation A N I M ATED A RT
of UAC to the stop codon UAG terminates translation through the RF1 mechanism. (c) A mutation Nonsense suppression at the molec-
in the anticodon of Tyr-tRNA Tyr allows the tRNA to read the UAG codon, add a tyrosine to the ular level
polypeptide chain, and permit translation elongation to continue.
Stress → decrease in general transcription modified, and proteins need to be transported to their sites of
action within or outside the cell. Some protein folding, mod-
increase in transcription of ification, and targeting takes place co-translationally (while
stress-response proteins the protein is being synthesized), and the rest takes place
Growth stimuli → permissive to general transcription post-translationally (after synthesis is complete).
increase in transcription of cell
KEY CONCEPT Developmental and environmental signals
growth proteins
regulate the translation of many mRNAs at the same time
In eukaryotes, signals that inhibit the translation of by altering the function of translation initiation factors and of
many mRNAs often act by directly altering the ability of selected mRNAs via RNA-binding factors (proteins and RNAs)
eIF4E to bind mRNA 5′ caps and assemble eIF4F to initi- that act on initiation and elongation factors.
ate translation (see Figure 9-18). On the other hand, sig-
nals that inhibit the translation of particular mRNAs act
through microRNAs (miRNAs) or RNA-binding proteins Protein folding
(RNA-BPs) that bind specific sequences or structural motifs Protein folding is the process by which proteins attain their
in an mRNA’s 5′ or 3′ UTR ( Figure 9-22). Bound factors functional tertiary structure. A protein that is folded cor-
in turn block translation through multiple mechanisms, rectly is said to be in its native conformation (in contrast
including inhibiting ribosome assembly and elongation, with an unfolded or misfolded protein that is nonnative).
inhibiting mRNA circularization, and promoting cleavage Folding involves the stepwise formation of secondary struc-
of the polypeptide chain. tures such as α-helices and β-sheets that are stabilized by
miRNAs are small RNAs (~21 nucleotides) that bind non-covalent hydrogen bonds. Folded regions then guide
with imperfect complementarity to their target mRNAs. and stabilize subsequent folding to progressively build the
Humans express over 2,500 miRNAs, and about 50% tertiary structure. The distinct three-dimensional structures
of human mRNAs are subject to regulation by miRNAs. of proteins are essential for their enzymatic activity, for
So, most biological processes, including cell differentia- their ability to bind to other molecules, and for their struc-
tion, growth, and proliferation, are regulated by miRNAs. tural roles in the cell. Although it has been known since
Single-stranded miRNAs associate with Argonaute (Ago) the 1950s that the amino acid sequence of a protein deter-
proteins and other proteins to form an RNA-induced silenc- mines its three-dimensional structure, it is also known that
ing complex (RISC). Guided to specific mRNAs by sequence the aqueous environment inside the cell does not favor the
complementary between the miRNA and the target mRNA, correct folding of most proteins. Given that proteins do in
RISC interacts with other proteins to inhibit translation as fact fold correctly in the cell, a long-standing question has
well as promote mRNA decay. been, how is correct folding accomplished?
Most newly synthesized proteins are unable to function The answer seems to be that folding often begins
until regulatory mechanisms alter their structure and cellular co-translationally and is helped by chaperones—a class of
location. To become functional, proteins need to be folded cor- proteins found in all organisms from bacteria to plants to
rectly, the amino acids of some proteins need to be chemically humans. Chaperones typically bind hydrophobic regions of
Circularization Elongation
Cap binding
Subunit joining Proteolysis N Protein
60S C
N
Met-tRNAiMet
RNA-BP
Cap miRNA PABP
eEFs eRFs
Circularization
FIGURE 9-22 In eukaryotes, translation is regulated by developmental protein (PABP) (blue arrows). Signal-dependent inhibition of translation
and environmental signals (red arrow) that affect the function of initiation through these mechanisms is sometimes mediated by microRNAs
and elongation factors (blue and purple arrows). Signals elicit their (miRNAs) and RNA-binding proteins (RNA-BPs) that bind sequences
effects on translation by inhibiting or promoting mRNA cap binding by in the mRNA 3′ UTR (purple arrows). AUG is the translation start site,
eIF4E, joining of the small and large ribosomal subunits, elongation by UAA is the translation stop site, AAAAAA is the poly(A) tail, UTRs are
the ribosome, proteolysis of the nascent polypeptide, and circularization untranslated regions, and ribosomal subunits are indicated in gray.
of the mRNA through the interaction of eIF4G and poly(A)-binding
incorrectly or incompletely folded proteins to arrest fold- five broad categories: the addition of chemical groups, com-
ing or promote unfolding and then release them to undergo plex molecules, or polypeptides to amino acids; the modi-
spontaneous refolding. Multiple rounds of binding and fication of amino acids; and the cleavage of peptide bonds
release that are driven by ATP hydrolysis occur until the between amino acids to convert inactive precursor proteins
protein is properly folded. Chaperones exist in all cell com- into smaller, active proteins (Figure 9-23).
partments. Some chaperones are expressed all the time in
cells, while others are upregulated by heat shock and other KEY CONCEPT Proteins can undergo five types of
stresses that increase protein misfolding. The latter class of post-translational modification:
chaperones is classified as stress or heat shock proteins. 1. Addition of chemical groups to amino acids (e.g., phosphory-
lation, methylation, acetylation, hydroxylation)
KEY CONCEPT The folding of newly synthesized proteins 2. Addition of complex molecules to amino acids (e.g., glyco-
into precise three-dimensional structures is determined by the sylation, lipidation)
primary amino acid sequence and assisted by a class of pro-
3. Addition of polypeptides to amino acids (e.g., ubiquitina-
teins called chaperones.
tion, sumoylation)
4. Modification of amino acids (e.g., deamidation, disulfide
Post-translational modification bond formation)
of amino acid side chains 5. Cleavage of peptide bonds to convert inactive precursor
proteins into smaller, active proteins (proteolysis)
Chemical modifications of amino acids greatly increase the
functionality of proteins. More than 200 different types of
amino acid modifications have been identified, many of Three types of proteins are involved in chemical modi-
which occur post-translationally. Amino acids that undergo fications: enzymes called writers that add chemical mod-
post-translational modification often have a functional ifications, enzymes called erasers that remove chemical
group that acts as a nucleophile in the enzymatic modifica- modifications, and structural proteins called readers that
tion reaction. Examples of nucleophiles include the hydroxyl bind chemical modifications. Sometimes, a single protein or a
group of serine, threonine, and tyrosine; the amine group of protein complex can be both a writer or eraser and a reader.
lysine, arginine, and histidine, the thiol group of cysteine; Addition and removal of chemical modifications
and the carboxylate group of aspartic acid and glutamic serves as a reversible switch to control the traits of pro-
acid (see Figure 9-2). The modifications themselves fall into teins, including increasing or decreasing their biological
(f) es
e p ti d
lyp
po Ubiquitination
of Adds ubiquitin
Phosphorylation
Ad
on (a protein of 76
Adds a phosphate
group to serine,
di
ti
amino acids)
di
t io
threonine, tyrosine,
Ad
to lysine
n
Sumoylation or others
Methylation
of
Adds SUMO (small Adds a methyl
ch
ubiquitin-like modifier) group to lysine
em
to lysine or arginine
ic al
o m plex m olec ule s
groups
Glycosylation Acetylation
Adds a sugar Adds an acetyl
to nitrogen or Target group to lysine or
oxygen protein the N-terminus
Lipidation Hydroxylation
Adds a hydroxyl
fc
Adds a lipid
such as a group
no
fatty acid
io
Deamidation
d it
Proteolysis
Removes an
Ad
Cleaves into
amide group
smaller poly- Disulfide bond
from asparagine
peptides or formation
or glutamine
Cl amino acids Covalently links
ea sulfur atoms of n
va
ge two cysteines a tio
d ific Reversible
mo
A m ino acid Irreversible
FIGURE 9-23 Proteins can be reversibly and irreversibly modified in numerous ways. (a–e)
Examples of the five types of post-translational modifications, shown in red. (f) A larger list of
examples of the five types of post-translational modifications.
activity, promoting or inhibiting their transport between Signal-dependent regulation of gene expression
cell compartments, increasing or decreasing their stabil-
ity, and enhancing or suppressing their interactions with Ligand (signal)
DNA, RNA, and other proteins. Two of the more com-
monly encountered post-translational modifications—
phosphorylation and ubiquitination—are considered next.
SUMMARY
This chapter has described translation, the process by which each containing rRNAs and proteins. rRNAs occupy
the nucleotide sequence of an mRNA is converted into important sites in ribosomes such as the decoding cen-
the amino acid sequence of a protein. Translation is the ter, which has E, A, and P binding sites for tRNAs, and
last step in the transfer of information from DNA to pro- the peptidyltransferase center, where peptide bonds are
teins. It occurs in three phases—initiation, elongation, and formed between amino acids. The accuracy of translation
termination—that are regulated by distinct factors—IFs, depends on the enzymatic linkage of an amino acid with
EFs, and RFs, respectively. In addition, initiation, elon- its cognate tRNA, generating a charged tRNA molecule.
gation, and termination involve conformational changes As adaptors, tRNAs decode the information in mRNA
driven by the hydrolysis of GTP, and elongation and ter- by base pairing between their anticodons and an mRNA
mination involve protein factors that mimic the function of codon, a triplet of nucleotides. Some tRNAs recognize
tRNAs. Some parts of the initiation mechanism are similar more than one codon by forming both Watson-Crick and
between bacteria and eukaryotes, and others are different wobble base pairs. RNA base pairing also occurs between
because bacterial translation occurs co-transcriptionally, the mRNA Shine–Dalgarno sequence and the 3′ end of
whereas eukaryotic transcription and translation are physi- 16S rRNA to position bacterial ribosomes at translation
cally separated into the nucleus and cytoplasm, respectively. initiation codons. Translation begins at a fixed starting
In addition, eukaryotic mRNAs have 5′ caps and 3′ poly(A) point (an AUG initiation codon that codes for methionine)
tails that play critical roles in translation. On the other and continues uninterrupted to the end of the open read-
hand, translation elongation and termination mechanisms ing frame (an in-frame stop codon that does not code for
are largely conserved between bacteria and eukaryotes. an amino acid).
Ribosomes, tRNAs charged with amino acids, and Just because an mRNA is present in a cell does not mean
mRNAs are central players in translation. Ribosomes are that it is translated. Translation is regulated, and this regu-
complex structures made up of small and large subunits, lation largely occurs at the initiation and elongation phases.
Some signals globally affect translation by altering the multiprotein complexes called quaternary structures. Pro-
activity of initiation factors that act on most mRNAs, while teins are the enzymes responsible for cell metabolism,
other signals affect the translation of particular mRNAs via including DNA and RNA synthesis; they are regulatory
factors that bind specific sequences or structures in mRNA factors required for expression of the genetic program, and
untranslated regions and influence multiple steps in transla- they are structural factors that confer stiffness to otherwise
tion initiation and elongation. flexible biological machines. Lastly, proteins contain short
The genetic code that deciphers mRNA sequence into sequence elements that act as zip codes to target them to
amino acids is almost universal among organisms. It con- distinct places in cells such as inside organelles, organelle or
tains 64 triplet codons, 61 that code for amino acids and 3 cell membranes, or the exterior of a cell.
that code for stop codons. The code is degenerate, nonover- An immense variety of amino acid modifications affect
lapping, continuous, and organized in a non-random pat- the structure, function, and localization of proteins. Many
tern. The organization minimizes effects of nonsynonymous modifications are reversible, so they serve as on/off switches
mutations on the functionality of proteins by having codons that control molecular processes such as transcription and
that differ by one nucleotide code for amino acids with sim- translation in response to signals. Modifications fall into
ilar polarity, charge, and size. five broad categories: cleavage of the peptide backbone,
Proteins are polymers of amino acids linked together modifications of amino acids, or the addition of complex
by peptide bonds. The 20 common amino acids all have an molecules, polypeptides, or chemical groups to amino acids.
amino group, a carboxyl group, and a unique side chain. The process of post-translational modifications involves
Differences in the polarity, charge, and size of amino acid enzymes that add modifications (writer), enzymes that
side chains affect the structure and function of proteins. remove modifications (erasers), and proteins that interpret
With the assistance of chaperones, proteins fold into sec- modifications (readers). In summary, the versatility of pro-
ondary structures such as α-helices and β-sheets as well teins as biological molecules is manifested in the diversity of
as tertiary structures. Proteins also assemble into stable shapes that they can assume.
KEY TERMS
active site (p. 304) initiation factor (IF) (p. 302) release factor (RF) (p. 302)
amino acid (p. 303) initiator tRNA (p. 315) ribosomal RNA (rRNA) (p. 302)
aminoacyl tRNA-binding site (A site) kinase (p. 324) ribosome (p. 302)
(p. 313) Kozak sequence (p. 316) ribosome-binding site (RBS) (p. 316)
aminoacyl-tRNA synthetase (p. 311) large ribosomal subunit (p. 310) secondary structure (of a protein)
amino group (p. 304) nonsense codon (p. 309) (p. 304)
anticodon (p. 310) nuclear localization sequence (NLS) Shine–Dalgarno sequence (p. 316)
carboxyl group (p. 304) (p. 325) signal sequence (p. 325)
chaperone (p. 321) peptide bond (p. 304) small ribosomal subunit (p. 310)
charged tRNA (p. 311) peptidyl site (P site) (p. 314) stop codon (termination codon)
codon (p. 306) peptidyltransferase center (p. 315) (p. 309)
decoding center (p. 315) phosphatase (p. 324) subunit (p. 306)
degenerate code (p. 306) phosphorylation (p. 324) synonymous codon (p. 309)
domain (p. 306) polypeptide (p. 303) termination factor (p. 302)
elongation factor (EF) (p. 302) primary structure (of a protein) (p. 304) tertiary structure (of a protein)
exit site (E site) (p. 314) proteasome (p. 324) (p. 304)
fibrous protein (p. 306) quaternary structure (of a protein) transfer RNA (tRNA) (p. 302)
frameshift mutation (p. 307) (p. 306) triplet (p. 306)
genetic code (p. 306) reactive group (R group), side chain ubiquitin (p. 324)
globular protein (p. 306) (p. 303) ubiquitination (p. 324)
initiation codon (p. 315) reading frame (p. 307) wobble (p. 312)
SOLVED PROBLEMS
SOLVED PROBLEM 1 – His – Thr – Glu – Asp – Trp – Leu – His – Gln – Asp
Using Figure 9-8, show the consequences on subsequent U U
translation of the addition of an adenine nucleotide to the U A U U A U
–CA –ACC–GA –GA –UGG–CUC–CA –CA –GA
beginning of the following coding sequence: C A G C C G C
A
G G
A
UUA
G
–CGA–UCG–GAA–CCA–CGU–GAU–AAG–CAU–
Because the protein-sequence change given to us at the
– Arg – Ser – Glu – Pro – Arg – Asp – Lys – His –
beginning of the problem begins after the first amino acid
SOLUTION (His) owing to a single nucleotide addition, we can deduce
that a Thr codon must change to an Asp codon. This change
With the addition of A at the beginning of the coding
must result from the addition of a G directly before the Thr
sequence, the reading frame shifts, and a different set of
codon (indicated by a box), which shifts the reading frame,
amino acids is specified by the sequence, as shown here
as shown here:
(note that a set of nonsense codons is encountered, which
results in chain termination):
–A
–ACG–AUC–GGA–ACC–ACG–UGA–UAA–GCA– U A G U
–CA – G AC–UGA – GA– U UG–G C U–UCA– U CA↑–GA –
– Thr – Ile – Gly – Thr – Thr – stop C G C
C C U C C
A A
SOLVED PROBLEM 2 G G
A single nucleotide addition followed by a single nucleo- – His – Asp – Arg – Gly – Leu – Ala – Thr – Ser – Asp –
tide deletion approximately 20 bp apart in DNA causes a
Additionally, because a deletion of a nucleotide must
change in the protein sequence from
restore the final Asp codon to the correct reading frame, an
–His–Thr–Glu–Asp–Trp–Leu–His–Gln–Asp– A or G must have been deleted from the end of the origi-
nal next-to-last codon, as shown by the arrow. The origi-
to
nal protein sequence permits us to draw the mRNA with
–His–Asp–Arg–Gly–Leu–Ala–Thr–Ser–Asp– a number of ambiguities. However, the protein sequence
resulting from the frameshift allows us to determine which
Which nucleotide has been added, and which nucleotide
nucleotide was in the original mRNA at most of these
has been deleted? What are the original and the new mRNA
points of ambiguity. Nucleotides that could have appeared
sequences? (Hint: Consult Figure 9-8.)
in the original sequence are circled. In only a few cases does
SOLUTION the ambiguity remain.
We can draw the mRNA sequence for the original protein
sequence (with the inherent ambiguities at this stage):
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the www
WORKING WITH THE FIGURES 3. In Figure 9-3a, draw the reaction that takes place to
(The first 25 questions require inspection of text figures.) add a third amino acid to the chain.
1. In Figure 9-1, circle a protein α-helix , an RNA stem, 4. In Figure 9-4c, where are hydrophobic and hydrophilic
and an RNA loop. amino acids most likely located?
2. In Figure 9-2 , for which amino acids does the 5. In Figure 9-5, what is the sequence of the fourth codon
single-letter abbreviation not match the first letter of in the nonoverlapping mechanism?
the amino acid name?
6. In Figure 9-6, what phenotype (rII+ or rII− ) would be to right and that the columns represent transcrip-
expected for two insertions and one deletion? tional and translational alignments.
7. In Figure 9-7 , what amino acids are encoded by a
repeat of the sequence CCA? T G A
DNA double helix
8. In Figure 9-8, list the amino acids that are coded for by C
1, 2, 3, 4, and 6 codons. C A U mRNA transcribed
9. In Figure 9-9, create an analogous figure for the UAA G C A Appropriate tRNA
stop codon. anticodon
10. In Figure 9-10, in general, what effect does a mutation Amino acids
of a purine to a purine or a pyrimidine to a pyrimidine Trp incorporated into
in the first nucleotide of a codon have on the polarity protein
and charge of an encoded amino acid?
11. In Figure 9-11 , draw the secondary structure of b. Label the 5′ and 3′ ends of DNA and RNA, as well as
tRNA Trp, include the sequence of the anticodon and the amino and carboxyl ends of the protein. www
label the 5′ and 3′ ends as well as the location of amino 27. Consider the following segment of DNA: www
16. In Figure 9-16, which subunit of the ribosome binds d. Label its amino and carboxyl ends.
the mRNA, and which subunit carries out peptide Repeat parts a through d, assuming that the top strand
bond formation? is the template strand.
17. Using Figure 9-17, circle the Shine–Dalgarno sequence 28. A mutational event inserts an extra base pair into DNA.
in Figure 9-15. Which of the following outcomes do you expect? (1)
18. In Figure 9-18 , describe three mechanistic differ- No protein at all; (2) a protein in which one amino acid
ences between translation initiation in bacteria and is changed; (3) a protein in which three amino acids are
eukaryotes. changed; (4) a protein in which two amino acids are
changed; (5) a protein in which most amino acids after
19. In Figure 9-19, draw the next step in elongation. the site of the insertion are changed.
20. In Figure 9-20, is RF1 a tRNA or a protein? 29. a. In how many cases in the genetic code would you
21. In Figure 9-21, what mutation in tRNA Tyr would sup- fail to know the amino acid specified by a codon
press a UAA nonsense mutation? if you knew only the first two nucleotides of the
22. In Figure 9-22, why do miRNAs and RNA-BPs affect codon?
only the translation of specific mRNAs? b. In how many cases would you fail to know the first
23. In Figure 9-23, draw phosphotyrosine. two nucleotides of the codon if you knew which
amino acid is specified by it?
24. In Figure 9-24, how might this pathway get turned
off? 30. If a polyribonucleotide contains equal amounts of
randomly positioned adenine and uracil bases, what
25. In Figure 9-25, how does this diagram provide insight proportion of its triplets will encode (a) phenylalanine,
into the picture of ribosomes shown on the first page (b) isoleucine, (c) leucine, (d) tyrosine?
of the chapter?
31. In the fungus Neurospora, some mutants were
BASIC PROBLEMS obtained that lacked activity for a certain enzyme. The
mutations were found, by mapping, to be in either of
26. a. Use the codon table in Figure 9-8 to complete the two unlinked genes. Provide a possible explanation in
following table. Assume that reading is from left reference to quaternary protein structure.
32. What is meant by the statement “The genetic code is Biotechnology companies must produce large quanti-
universal”? What is the significance of this finding? ties of these immune proteins for human testing and
33. A mutant has no activity for the enzyme isocitrate eventual sale to the public. To this end, their scientists
lyase. Does this result prove that the mutation is in the engineer bacterial or human cell cultures to express
gene encoding isocitrate lyase? Why? these immune proteins. Explain why proteins isolated
from bacterial cultures are often inactive, whereas the
34. A certain nonsense suppressor corrects a non-growing same proteins isolated from human cell cultures are
mutant to a state that is near, but not exactly, wild type active (functional).
(it has abnormal growth). Suggest a possible reason
why the reversion is not a full correction. 45. Would you expect to find nuclear localization
sequences (NLSs) in the proteins that make up bac-
35. In bacterial genes, as soon as a partial mRNA tran- terial and eukaryotic DNA and RNA polymerases?
script is produced by RNA polymerase, the ribosome Explain why or why not.
assembles on it and starts translating. Draw a diagram
of this process, identifying 5′ and 3′ ends of mRNA, CHALLENGING PROBLEMS
the amino and carboxyl ends of the protein, the RNA
polymerase, and at least one ribosome. Why couldn’t 46. Draw the structure and hydrogen bonding of a parallel
this system work in eukaryotes? β-sheet .
36. Researchers have found that aspartic acid and glu- 47. How were synthetic RNAs such as poly(U) that lacked
tamic acid can sometimes mimic the function of phos- a Shine–Dalgarno sequence translated in an E. coli
phoserine and phosphothreonine. Why might this be? extract?
37. Why might a mutation in the untranslated region of a 48. A single nucleotide addition and a single nucleotide
bacterial mRNA affect translation? How about for a deletion approximately 15 bases apart in the DNA
eukaryotic mRNA? cause a protein change in sequence from
38. In vitro translation systems have been developed in Phe–Ser–Pro– Arg–Leu– Asn– Ala– Val–Lys
which a specific mRNA can be added to a test tube to
containing a bacterial cell extract that includes all
Phe– Val–His– Ala–Leu–Met– Ala– Val–Lys
the components needed for translation (ribosomes,
tRNAs, and amino acids). If a radioactively labeled a. What are the old and new mRNA nucleotide
amino acid is included, any protein translated from sequences? Use the codon table in Figure 9-8.
that mRNA can be detected on a gel. If a eukaryotic b. Which nucleotide has been added? Which has been
mRNA is added to the in vitro system, would radioac- deleted?
tive protein be produced? Explain why or why not.
49. You are studying an E. coli gene that specifies a pro-
39. An in vitro translation system contains a eukaryotic tein. A part of its sequence is
cell extract that includes all the components needed
– Ala–Pro– Trp–Ser–Glu–Lys–Cys–His–
for translation (ribosomes, tRNAs, and amino acids).
If bacterial RNA is added to the test tube, would a You recover a series of mutants for this gene that show
protein be produced? Explain why or why not. no enzymatic activity. By isolating the mutant enzyme
products, you find the following sequences:
40. Would a chimeric translation system containing the
large ribosomal subunit from E. coli and the small Mutant 1:
ribosomal subunit from yeast (a unicellular eukaryote) – Ala–Pro– Trp– Arg–Glu–Lys–Cys–His–
be able to function in protein synthesis? Explain why
Mutant 2:
or why not.
– Ala–Pro–
41. Mutations that change a single amino acid in the active
site of an enzyme can result in the synthesis of wild- Mutant 3:
type amounts of an inactive enzyme. In what other – Ala–Pro–Gly – Val–Lys– Asn–Cys–His–
regions of a protein might a single amino acid change
Mutant 4:
have the same result?
– Ala–Pro– Trp–Phe–Phe– Thr–Cys–His–
42. What evidence supports the view that ribosomes are
ribozymes? What is the molecular basis for each mutation? What
is the DNA sequence that specifies this part of the
43. Explain why antibiotics, such as erythromycin and
protein?
Zithromax, that bind the large ribosomal subunit do
not harm us. 50. What structural features are shared by spliceosomes
(see Figure 8-19) and ribosomes? Why are both struc-
44. Our immune system makes many different proteins
tures used to support the RNA World theory?
that protect us from viral and bacterial infection.
51. A double-stranded DNA molecule with the sequence (UUC)n (Ser–Ser) and (Leu–Leu) and (Phe–Phe)
shown here produces, in vivo, a polypeptide that is five
amino acids long. (UUG)n (Leu–Leu) and (Val–Val) and (Cys–Cys)
a. Which strand of DNA is the template strand, and (UAC)n (Thr–Thr) and (Leu–Leu) and (Tyr–Tyr)
in which direction is it transcribed? (AUC)n (Ile–Ile) and (Ser–Ser) and (His–His)
b. Label the 5′ and 3′ end of each strand. (GUA)n (Ser–Ser) and (Val–Val)
c. If an inversion occurs between the second and the (GAU)n (Asp–Asp) and (Met–Met)
third triplets from the left and right ends, respec- (UAUC)n (Tyr–Leu–Ser–Ile)
tively, and the same strand of DNA is transcribed,
how long will the resultant polypeptide be? (UUAC)n (Leu–Leu–Thr–Tyr)
331
I
n this chapter, we describe experimental techniques used to Most toolboxes that we are familiar with are filled with
isolate and manipulate genes and their products, RNAs and tools like hammers, screwdrivers, and wrenches that are
proteins. There are many good reasons for including this designed by people and manufactured in factories. In con-
information in an introductory genetics textbook. To develop trast, the tools of the genetic engineer are molecules iso-
these techniques, researchers relied on knowledge of the chem- lated from cells. Most of these tools were the product of
ical and functional properties of DNA, RNAs, and proteins scientific discovery—where the objective was to answer a
as well as the mechanisms underlying fundamental molecu- biological question. Only later did some scientists appre-
lar genetic processes such as DNA replication, transcription, ciate the potential practical value of these molecules and
and translation. Thus, an understanding of experimental tech- invent ways to put them to use with the goal of isolating
niques will reinforce principles that are presented in other and amplifying DNA fragments. As an example, one way
chapters. It will also aid in the comprehension and evalua- to separate our gene of interest from the rest of the genome
tion of the primary research literature, as well as the design is to cut the genome with “molecular scissors” and isolate
of experiments to address yet-to-be-solved genetic problems. the small fragment containing the gene. Werner Arber dis-
Lastly, whether or not a problem can be solved is often deter- covered these molecular scissors, and for this discovery he
mined by the techniques that are available. Consequently, was awarded the Nobel Prize in Physiology or Medicine
some major advances in genetics have become possible only in 1978. However, Arber was not looking for a tool to cut
because of the development of a new technique. Standouts DNA precisely. Rather, he was trying to understand why
include techniques for isolating and manipulating fragments some bacteria are resistant to infection by bacterial viruses.
of DNA (DNA cloning), amplifying DNA (PCR), sequenc- By answering this biological question, he discovered that
ing DNA (dideoxy sequencing), and introducing DNA into resistant bacteria possess a previously unknown type of
an organism (transgenesis). This chapter ends with recently enzyme—a restriction endonuclease—that cuts DNA at
developed techniques such as CRISPR-Cas9 for the precise specific sequences. As we will see in this chapter, restriction
engineering of genomes. New genome engineering techniques enzymes are one of the cornerstones of genetic engineering
have made possible reverse genetic studies that aim to under- and a common tool found in the genetic engineer’s toolbox.
stand the function of a gene by analyzing the phenotypic con- As another example, it is unlikely that anyone would
sequences of altering the gene sequence or its expression. have predicted that DNA polymerase, discovered by Arthur
Genes are the central focus of genetics, and so, clearly, it Kornberg, could be fashioned into two powerful tools for
is desirable to isolate a gene of interest (or any DNA region) DNA isolation and analysis. To this day, many of the tech-
from the genome to study it. Isolating individual genes and niques used to determine the nucleotide sequence of DNA
producing enough copies of them to analyze can be a daunt- rely on synthesizing it with DNA polymerase. Similarly, most
ing task because a single gene is a tiny fraction of an entire of the protocols used to isolate and amplify specific regions
genome. For example, the haploid human genome contains of DNA from sources as disparate as a crime scene or a fossil
over 3 billion base pairs, whereas the coding region of an embedded in amber rely on the activity of DNA polymerase.
average gene contains only a few thousand base pairs. How DNA technologies are the collective techniques for
do scientists find the proverbial needle in the haystack—the obtaining, amplifying, and manipulating specific DNA
gene—and then produce suitable quantities of it for analysis? fragments. Since the mid-1970s, the development of DNA
Many investigations in genetics begin with the desire to technologies has revolutionized the study of biology, open-
study a trait or a disease. In Chapter 2, we described forward ing many areas of research to molecular investigation.
genetic approaches to search for mutants that exhibit an Genetic engineering, the application of DNA technologies
altered phenotype and crosses or pedigree analysis to deter- to specific biological, medical, or agricultural problems, is
mine whether that phenotype is determined by a single gene. now a well-established branch of technology. Genomics
In Chapter 4, we discussed how mapping by recombination is the ultimate extension of the technologies to the global
helps locate the gene at the DNA level. In this chapter, we analysis of the nucleic acids present in a nucleus, a cell, an
continue by presenting molecular methods for identifying a organism, or a group of related species (see Chapter 14).
gene of interest and studying its molecular function. In this chapter, we will illustrate gene isolation and
The first step in studying gene function is to isolate its manipulation techniques through their application to the
DNA and reproduce it in quantities suitable for study. Just insulin gene. The insulin gene encodes a protein that func-
like a construction worker, a genetic engineer needs tools. tions to maintain normal blood glucose levels by promoting
332
the uptake of glucose from blood into cells and by regulat- pancreases of animals processed in meatpacking plants and
ing the metabolism of carbohydrates, lipids, and proteins. purified on a large scale to eliminate the majority of pro-
In mammals, including humans and mice, the insulin gene teins and other contaminants in pancreas extracts. It took
is present in the genome of all cells, but it is expressed (i.e., about 8,000 pounds of pancreas from 23,500 animals to
transcribed and translated) only in beta (β) cells in the pan- purify one pound of insulin. Then, in 1982, the first recom-
creas. Diabetes is a human disease in which blood glucose binant human insulin came on the market. Because it could
levels are abnormally high, either because β cells do not pro- be produced on an industrial scale in bacteria by recom-
duce enough insulin (type I diabetes) or because cells are binant DNA techniques using the human gene sequence,
unable to respond to insulin (type II diabetes). Mild forms of insulin could be made in a purer form and at a lower cost
type I diabetes can be treated by dietary restrictions, but for than the previous method. We will use the generation of
many patients, daily insulin injections are necessary. recombinant human insulin as an example of the general
Until about 35 years ago, cows were the major steps necessary for making any recombinant DNA mole-
source of insulin protein. The protein was harvested from cule. These steps are summarized in Figure 10-1. The uses
Genomic DNA
Genomic DNA Plasmid DNA
Nucleus
Insulin mRNA
Other mRNAs
Extract and purify mRNA from cells
Isolate plasmid DNA
Convert mRNA to cDNA
Cut plasmid with restriction enzymes
Amplify the insulin cDNA by PCR
Cut insulin cDNA with restriction enzymes
Recombinant bacterium
Insulin protein
FIGURE 10-1 Recombinant human insulin is currently being produced in large quantities. As the recombinant bacteria grow, they transcribe
in bacteria. The first step in this process is the construction of an insulin mRNA and translate the mRNA into insulin protein. Finally, the
expression plasmid that contains a human insulin cDNA insert. The bacteria are harvested, and the insulin protein is extracted and purified
plasmid is then transformed into bacteria, and the bacteria are grown for therapeutic use in humans.
of recombinant DNA technologies are quite broad, rang- consider how the methods presented could be used to iso-
ing from gene isolation for basic biological research to gene late and manipulate insulin genes in other animals such as
therapy to treat human disease to the production of herbi- the fruit fly Drosophila melanogaster that lack a pancreas
cides and pesticides for crop plants. and β cells.
(a)
0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500
Base pairs/nucleotides
(b)
E X X X X E
Human Ins locus
Probe 1
13 kb
5 kb
P B B P P
Mouse Ins1 locus
Probe 2
9 kb 7.5 kb
3 kb
N E N E
Mouse Ins2 locus
Probe 3
15 kb
7.5 kb
FIGURE 10-3 (a) Structures of insulin genes and mRNAs in humans and mice. Boxes represent
exons. The mRNAs are spliced, capped at the 5′ end with m7G, and polyadenylated at the 3′ end
(Chapter 8). (b) Structures of insulin gene loci in humans and mice. Red lines indicate regions used
as probes for Southern and Northern blot analyses. Results for probe 1 are shown in Figure 10-6,
and questions related to probes 2 and 3 are in the Working with the Figures section at the end of
the chapter. Restriction enzymes used for Southern blot analysis are indicated by single letters:
B = BamHI, E = EcoRI, N = NsiI, P = PvuII, and X = XhoI (restriction sites for these enzymes are shown
in Table 10-1). Lines with arrows indicate the sizes of some restriction fragments.
Southern and Northern blot procedure Techniques analogous to Southern, Northern, and West-
ern blotting are used to detect DNA, RNAs, and proteins
DNA or RNA in vivo in cultured cells or in whole organisms. Detection
of DNA and RNA in vivo is carried out by hybridization
32P-labeled with radioactive single-stranded nucleic acid probes, fol-
size markers lowed by autoradiography. This technique is called in situ
hybridization (ISH). A modified form of ISH that uses flu-
orescently labeled probes rather than radioactive probes
Electrophoresis is called fluorescence in situ hybridization (FISH) (for
– example, see Figure 13-6 ). Proteins are detected in vivo
Migration
by immunofluorescence , which uses antibody probes (for
example, see Figure 13-15). A fluorescence microscope is
+
used to reveal the location and abundance of the fluores-
cence signal in FISH and immunofluorescence.
Movement of the
solution to the paper
towels transfers the KEY CONCEPT Specific DNA fragments or RNAs are
DNA or RNA from the detected in vitro and in vivo by hybridization with nucleic acid
Paper towels gel to the membrane probes, and specific proteins are detected by interaction with
antibody probes.
Gel
Sponge Membrane
The expected results of Northern and Western blot
Salt solution
analyses for insulin are shown in Figures 10-6a and
10-6b , respectively. Northern blot analysis of the thou-
sands of different mRNAs expressed in human cells with
Gel a probe complementary to the human Ins mRNA should
detect one band, that is, a single type of insulin mRNA in
Membrane β cells but not in kidney cells (Figure 10-6a, lanes 2 and 3).
Because the nucleotide sequence of the insulin mRNA is
DNA or RNA very similar between humans and mice, the human probe
transfered to
Hybridize with unique membrane
should also detect the two insulin mRNAs (Ins1 and Ins2)
nucleic acid probe in mouse β cells (Figure 10-6a, lane 5). Standardly, blots
are probed not only for a molecule of interest, but also for
other molecules that serve to confirm that the experiment
Incubate worked as expected or that serve as the basis for compar-
membrane
and probe in ison within or between samples, which is referred to as a
a sealed bag loading, normalization, or specificity control. In this case,
analysis of ribosomal protein S7 (RpS7) mRNA, a tran-
script that is present in all cells at similar levels, shows that
Wash away Probe hybridized the failure to detect insulin mRNA in the kidney cell sam-
unbound probe to complementary ples (Figure 10-6a, lanes 2 and 4) was not due to a prob-
nucleic acid lem with the experiment. Furthermore, the intensity of the
bands provides information about the abundance of the
mRNAs. The data show that, in human β cells, Ins mRNA
is less abundant than RpS7 mRNA, and, in mouse β cells,
Ins2 is more abundant than Ins1. Western blot analysis of
Expose X-ray film
to membrane
Northern, Western, and Southern blots for insulin proteins from the same types of cells with an anti-
body to human insulin shows that cell type-specific
(a) Northern blot expression of the insulin protein is similar to that of
Size Human Mouse the insulin mRNA (Figure 10-6b). The expected size
markers kidney Human kidney Mouse of a protein on a Western blot can be estimated based
(nt) cells β cells cells β cells
on the average molecular weight of an amino acid,
1000 110 Daltons (Da). One Da is one gram per mole, and
900 RpS7 one kilodalton (kDa) is 1000 grams per mole. Thus,
800 (loading the 110-amino-acid insulin protein is about 12 kDa
700 control)
(110 amino acids × 110 Da/amino acid). Note that only
600 Ins1 one insulin protein band should be detected in mouse
Ins β cells because the Ins1 and Ins2 proteins are the same
500
Ins2 size and, thus, should migrate to the same position
400 upon gel electrophoresis.
Sometimes, the starting material for Southern
300 blot analysis is chromosome-sized DNA molecules of
genomic DNA. Such large molecules can be analyzed
more precisely when cut into fragments of much smaller
200
size. Cutting is carried out by bacterial restriction
1 2 3 4 5 enzymes. These enzymes are endonucleases that cleave
phosphodiester bonds between nucleotides at specific
(b) Western blot DNA sequences, called restriction sites, that are usually
Size Human Mouse 4 to 8 base pairs long. Table 10-1 shows several exam-
markers kidney Human kidney Mouse ples of restriction enzymes and their restriction sites.
(kDa) cells β cells cells β cells
Restriction sites are palindromic, which means that
50 both strands have the same nucleotide sequence but in
antiparallel orientation (the same sequence reads 5′ to
40 3′ on each strand). For example, the restriction enzyme
↓
30 −GAATTC−3′
EcoRI recognizes the sequence 35′′−CTTAAG−5′
and cleaves
↑
5000
FIGURE 10-6 (a) Northern, (b) Western, (c) and Southern blot
analyses for insulin in different human and mouse tissues. The
4000
Southern and Northern blots were analyzed with probe 1 to exon
2 of the human Ins gene (see Figure 10-3b). The Northern blot
was also analyzed with a probe to the RpS7 mRNA. The Western
3000 blot was probed with antibodies to human insulin and RpS7
1 2 3 4 5 proteins.
ANIMATED ART Western blot analysis
The expected results of Southern blot analysis for the starting with only one copy. This is called DNA amplification.
insulin gene are shown in Figure 10-6c. Each sample on the To develop PCR, Mullis brought together several pieces of
blot contains genomic DNA that was digested with restric- information obtained through basic research. First, he knew
tion enzymes. If the sequence of the 3 × 109 base pair human from studies of DNA replication that DNA polymerases copy
genome was completely random, the 6-base pair restric- a single-stranded DNA template by extending off the 3′ end
tion site for EcoRI should occur every 4096 base pairs of an annealed primer. Second, he knew from techniques such
(46 , the number of possible base pairs at each position in as Southern blotting that, in solution, a short piece of single-
the restriction sitethe number of base pairs in the restriction site ), which stranded DNA called an oligonucleotide will specifically
means that digestion of the human genome would produce anneal to DNA sequences that are perfectly complementary.
about 730,000 EcoRI fragments (3 × 109/4096). Based on Third, he knew that high temperatures disrupt hydrogen
restriction maps of genomic DNA, as shown in Figure 10-3b, bonds between bases in double-stranded DNA to produce
the probe complementary to the insulin gene should detect single-stranded DNA. Lastly, he knew that about 20 years
a single EcoRI band out of the estimated 730,000 EcoRI earlier, the microbiologist Thomas Brock was able to culture
bands (Figure 10-6c, lane 3). The same probe should detect a a bacterium Thermus aquaticus that grows at high tempera-
smaller band when the DNA is digested with the restriction tures in the hot springs of Yellowstone National Park. The
enzyme XhoI or with both EcoRI and XhoI (Figure 10-6c, DNA polymerase from this bacterium, called Taq polymerase,
lanes 2 and 4), because the XhoI fragment is located within not only is active at high temperatures but also remains active
the EcoRI fragment. Southern blot analysis of DNA from β over many cycles of heating and cooling. By putting these
cells and kidney cells produces the same result (Figure 10-6c, pieces of information together, Mullis constructed a simple
lanes 4 and 5) because all cells in an organism have the same method to amplify any DNA sequence in vitro.
genomic DNA. Whereas Northern and Western blot analysis The basic strategy of PCR is outlined in Figure 10-7.
of β cells and kidney cells produces different results (Figures The process uses a pair of chemically synthesized oligo-
10-6a and 10-6b), because insulin is expressed only in β cells. nucleotide DNA primers that are each about 20 nucleo-
tides long. Each primer is designed to base pair to one end
of the target gene or region to be amplified, such that the
Detecting and amplifying DNA by the primers base pair to opposite DNA strands with their 3′
polymerase chain reaction ends pointing toward each other. The primers are added
In 1985, the ability of researchers to analyze and manipulate to a solution containing the DNA template (e.g., genomic
DNA was transformed by the invention of the polymerase DNA), the four deoxyribonucleoside triphosphates (dATP,
chain reaction (PCR) by Kary Mullis. PCR makes it possi- dCTP, dGTP, and dTTP) required for DNA synthesis
ble to produce billions of copies of a specific DNA sequence (Figure 7-5), and the heat-stable Taq DNA polymerase.
Polymerase chain reaction The DNA template is denatured by heat (95°C), resulting in
single-stranded DNA molecules. Upon cooling the reaction
Region of target DNA
to be amplified to between 50°C and 65°C, the primers anneal to their com-
(a) plementary sequences in the single-stranded DNA template.
5′ After the temperature is raised to 72°C, Taq polymerase
3′
replicates the single-stranded DNA segments by extending
Add DNA primers, dNTPs,
and Taq polymerase from the annealed primers. Complementary new strands
are synthesized, as in DNA replication in cells, forming
DENATURE: Heat to separate
target DNA strands (95°C) two double-stranded DNA molecules identical to the single
parental double-stranded DNA molecule. Thus, one cycle of
ANNEAL: Cool to allow primers to
(b) base pair to target DNA (~60°C) PCR consists of three main steps (denaturing, annealing, and
5′ extending) and results in doubling the starting amount of the
5′
5′ target sequence. Subsequent cycles of denaturing, anneal-
3′ ing, and extending also double the amount of the target
EXTEND: Heat to allow
sequence, resulting in an exponential increase (2number of cycles )
DNA synthesis (72°C) in the number of copies of DNA. Thus, a typical PCR with
(c) 30 cycles of five minutes each will amplify the target DNA
5′ about one billion-fold (230 ) in 2.5 hours.
5′ PCR is a powerful technique that is routinely used to
5′
3′ isolate specific regions of DNA when there is prior knowl-
edge of the sequence to be amplified. What makes PCR so
DENATURE, ANNEAL powerful is that only small amounts of starting material are
(d) needed, which makes it possible to work with DNA samples
5′ that are difficult to obtain, such as from a small number of
tumor cells. There are applications for PCR in genotyping,
sequencing, cloning, paternity testing, forensics, molecu-
lar archeology, detection of infectious diseases, and many
other clinical and basic research efforts that involve DNA.
3′ In recognition of the importance of PCR, Kary Mullis was
awarded the Nobel Prize in Chemistry in 1993.
EXTEND
(e)
KEY CONCEPT The polymerase chain reaction uses spe-
5′ cially designed primers to amplify specific regions of DNA in a
test tube.
Quantifying DNA by real-time PCR Since the amount of compare the amount of viral DNA in blood samples that
DNA produced by PCR doubles with each cycle, it is possi- are collected at different times.
ble to calculate the amount of DNA in a sample based on
the amount of DNA produced after a given number of PCR KEY CONCEPT Quantitative PCR (qPCR) is a method that
cycles. Quantification of DNA by PCR, called quantitative uses a real-time PCR instrument to determine the amount of a
PCR (qPCR), is automated by real-time PCR instruments that specific DNA molecule in a sample.
measure the amount of DNA product in “real-time” during
each PCR cycle. To be more exact, these instruments measure
the intensity of a fluorescent signal generated by a dye called Detecting and quantifying mRNA by reverse-
SYBR green, which, like ethidium bromide, intercalates into transcription PCR PCR can also be used to detect, amplify,
double-stranded DNA. As the amount of DNA increases with and quantify mRNA; however, single-stranded RNA must first
each PCR cycle, the fluorescent signal increases (Figure 10-8). be converted into double-stranded DNA. Complementary
The number of PCR cycles that it takes for the fluorescent DNA (cDNA) is a double-stranded DNA version of an mRNA
signal to be detected above background is called the C T (cycle molecule. cDNA is made from mRNA in vitro by a special
threshold) value. If two samples have C T values of 16 and enzyme called reverse transcriptase, originally isolated from
24, it would mean that the second sample had 256-fold less retroviruses (see Chapter 16). Reverse transcriptase is a type of
DNA than the first sample because eight more cycles were DNA polymerase that synthesizes a DNA strand complemen-
required to attain a fluorescent signal above background. In tary to an RNA template. Retroviruses such as human immuno-
other words, the fold difference in the amount of DNA in the deficiency virus (HIV) use reverse transcriptase to convert their
two samples is equal to 2−∆CT . So, in this case, RNA genomes into DNA as part of their replication cycle.
2−(16−24) = 28 = 256 cDNA synthesis begins with the purification of mRNA
from a tissue or specific cells such as the pancreas or β
qPCR is often used to compare the relative amounts of cells. Purification is necessary because mRNA accounts for
different DNA molecules in a single sample. For example, only about 5 percent of the total amount of cellular RNA,
to diagnose cancer due a somatic mutation, qPCR is used with rRNA accounting for about 80 percent and tRNA for
to determine the fraction of cells in a tumor sample that 15 percent. mRNAs from eukaryotic cells are commonly
contain the mutant gene versus the wild-type gene. qPCR is purified using affinity methods that target the unique fea-
also used to compare the relative amount of the same DNA tures of mRNA relative to other types of RNA, that is, the
molecule in different samples. For example, to determine 5′-m 7G cap and the 3′ poly(A) tail. Next, purified mRNA
the rate of progression of a viral infection, qPCR is used to is incubated with reverse transcriptase, the four dNTPs,
and an oligo-dT primer (an oligonucle-
Real-time PCR otide of about 20 T residues) (Figure 10-9).
The oligo-dT primer anneals to the poly(A)
3 tail of the mRNA, and, using the mRNA
as a template, reverse transcriptase synthe-
sizes single-stranded DNA starting from the
Sample A
oligo-dT primer and ending at the m 7 G cap.
2 This is called first-strand cDNA synthesis.
Fluorescence
Sample B
The RNA strand of the RNA-DNA hybrid
is then removed by alkaline hydrolysis (as
described in Chapter 8, RNA is susceptible
1 to base-catalyzed hydrolysis because of the
Threshold hydroxyl group at the ribose sugar 2′ posi-
tion) or by RNase H, an enzyme that cuts
the RNA strand of an RNA-DNA hybrid.
0 Second-strand cDNA synthesis is carried out by
0 10 20 30 40
Cycle number CT = 16 CT = 24 E. coli DNA polymerase I and primed by liga-
tion of an oligonucleotide of known sequence
FIGURE 10-8 The amount of a specific DNA molecule in a sample can be quantified by to the 3′ end of the first-strand cDNA.
real-time PCR. A real-time PCR instrument measures the fluorescence signal at each PCR
cycle. In this example, the same DNA molecule was analyzed in two samples, Sample KEY CONCEPT Reverse transcriptase syn-
A and Sample B. The cycle threshold (C T ) is the PCR cycle at which the fluorescence thesizes DNA using an RNA template and can
reached a threshold. The C T values for the two samples are used in the formula 2−∆CT to be used to create cDNA, a double-stranded
calculate the relative amount of the specific DNA molecule in the samples.
DNA copy of an mRNA molecule.
A N I M AT E D ART Real-time qPCR
cDNA synthesis by reverse transcription KEY CONCEPT Conversion of mRNA into cDNA makes it
possible to use PCR approaches to amplify and quantify spe-
mRNA cific mRNAs.
5′ m7G A A A A A A A A 3′
C C TC
C TT T
GA
G A AT T C G A AT T C G A AT T C CTCGAG
G
A
GAA
C T TAA G C T TAA G C T TAA G GAGCTC
G
Cleavage Cleavage Cleavage by Cleavage by
by EcoRI by EcoRI EcoRI and XhoI EcoRI and XhoI
A AT T C G A AT T C C
G C T TAA G GAGCT
Anneal Anneal
Nick Nick
Nick Nick
Nick Nick
Nick Nick
C G C C
T
G C A T G G AT
A T A
C
TA T
A AA
CT A T
CT A
G TC
G
T
T
TA
C
GA
TC
AG
GA
G
Recombinant Recombinant
DNA DNA
molecule molecule
FIGURE 10-10 To form a recombinant DNA molecule, restriction enzymes are used to cut a vector A N I M ATED A RT
and an insert. Because of sequence complementarity at the ends of the vector and insert, the vector
and insert anneal. DNA ligase then permanently links the vector and insert together. Depending on the Plasmid cloning
restriction enzymes used, the insert can hybridize with the vector in (a) two orientations or (b) only one
orientation.
ends, the insert can hybridize in two orientations relative to Choice of cloning vector Numerous cloning vectors
the vector (Figure 10-10a); whereas cloning with two dif- that meet a wide range of experimental needs are in cur-
ferent restriction enzymes (at least one of which produces a rent use. Vectors are used mainly to express a specific RNA
staggered cut) limits hybridization of the insert and vector or protein or to increase the amount of a specific DNA
to one orientation (Figure 10-10b). molecule so that the DNA can be sequenced or further
cloned. As described earlier, all vectors must have conve-
KEY CONCEPT Insert and vector DNAs with the same sticky
nient restriction sites at which the DNA to be cloned can
ends or with blunt ends can be joined efficiently and ligated.
be inserted. Other important features of vectors are ways to
quickly identify the desired recombinant vector as well as protein. pET vectors contain an epitope tag called a His-
to express and purify the recombinant protein. Some gen- tag that consists of six histidine amino acids (6X-His-tag).
eral classes of cloning vectors follow. Since the tag is small, it typically does not affect the struc-
ture or function of the recombinant protein. Purification
Plasmid vectors Bacterial plasmids, which we first encoun- of His-tagged proteins is based on the affinity of histidine
tered in Chapter 6, are small circular DNA molecules that for metal ions such as nickel (Ni2+ ). As shown in Figure
are replicated independently of the bacterial chromosome 10-11b, His-tagged recombinant proteins are purified by
because they have an origin of replication (ori). Plasmids Ni2+ affinity chromatography from an E. coli extract, a
routinely used as vectors carry a gene for drug resistance solution of E. coli cells that are broken open to release the
and a gene to distinguish plasmids with and without DNA recombinant protein as well as E. coli proteins. The E. coli
inserts. Genes that confer resistance to antibiotics such as extract is mixed with inert beads that have Ni2+ immobi-
ampicillin (ampR gene), tetracycline (tet R gene), and chlor- lized on their surface. The His-tagged recombinant protein
amphenicol (camR gene) provide a convenient way to select binds tightly to the beads. The beads are washed several
for bacterial cells transformed by plasmids: those cells still times to remove non-specifically bound E. coli proteins,
alive after exposure to the drug must carry the plasmid vec- leaving only His-tagged recombinant proteins bound to the
tor. However, because not all plasmids in transformed cells beads. The bound proteins are then released (i.e., eluted)
will contain DNA inserts, some plasmid vectors also have from the beads in a pure form by adding a chemical called
a system that allows researchers to identify bacterial colo- imidazole that competes with the His-tag for binding to
nies with plasmids containing DNA inserts. Such a feature the beads. Bacterial expression plasmids and methods of
is part of the pUC18 plasmid vector shown in Figure 10-11a; this kind are used to synthesize and purify human insulin
DNA inserts disrupt a gene (lacZ) in the plasmid that protein.
encodes an enzyme (β-galactosidase) capable of cleaving a
compound added to the bacterial culture plate (X-gal) so KEY CONCEPT The essential features of plasmids for clon-
that it produces a blue pigment. Thus, colonies that con- ing are an origin of replication so that the plasmid is replicated
tain plasmids with an insert will be white rather than blue when bacteria divide, a drug resistance gene so that bacteria
containing the plasmid can be identified, and a polylinker so
(i.e., they cannot cleave X-gal because they cannot produce
that DNA can be inserted with restriction enzymes.
β-galactosidase).
Plasmids called expression plasmids contain sequence
that control the transcription and translation of the inserted KEY CONCEPT Non-essential but useful features of plas-
DNA, often a gene in the form of a cDNA. Expression plas- mids include sequence elements for the identification of
mids can drive transcription of the inserted gene constitu- plasmids that contain inserts, the constitutive or inducible
tively (i.e., all the time) or inducibly (i.e., only in response expression of inserted genes, and the addition of an epitope
to a signal). For example, some pET plasmids are used for tag onto a recombinant protein.
constitutive expression of recombinant proteins in E. coli
(Figure 10-11b). These plasmids contain the bacteriophage Bacteriophage vectors A bacteriophage vector harbors
T7 promoter that drives transcription of the inserted gene DNA as an insert packaged inside the phage particle. Dif-
in E. coli that express bacteriophage T7 RNA polymerase. ferent classes of bacteriophage vectors can carry different
Another type of the pET plasmid uses components of the sizes of insert DNA. Bacteriophage λ (lambda; discussed
lac operon (described in Chapter 11) to inducibly express in Chapters 6 and 11) is an effective cloning vector for
the inserted gene in E. coli. This type of plasmid contains double-stranded DNA inserts as long as 15 kb. The central
the lac operator site near the T7 promoter and also con- part of the phage genome is not required for replication or
tains the lacI gene that encodes the Lac repressor protein. packaging of λ DNA molecules in E. coli and so can be cut
In uninduced cells, the Lac repressor binds the lac opera- out by restriction enzymes and discarded. The deleted cen-
tor and represses transcription of the inserted gene by T7 tral part is then replaced by insert DNA.
polymerase. However, when the compound IPTG (isopro-
pyl β-D-1 thiogalactopyranoside) is added to the growth Vectors for larger DNA inserts The standard plasmid and
media, the Lac repressor is inactivated and T7 polymerase phage λ vectors just described can accept inserts as large as
can transcribe the inserted gene. Therefore, the recombinant 15 kb. However, many experiments require inserts well in
protein is expressed only in the presence of IPTG. Inducible excess of this size. To meet these needs, special vectors that
expression is helpful in cases where constitutive expression require more sophisticated methods for transferring DNA
of the recombinant protein produces a large amount of pro- into the host cell have been engineered. In each case, the
tein that is toxic to E. coli, or it makes the protein insoluble. recombinant DNAs replicate as large plasmids after they
A final feature of bacterial expression plasmids, includ- have been delivered into the bacterium.
ing pET vectors, is a sequence that encodes an epitope Fosmids are vectors that can carry 35- to 45-kb inserts
tag that can be used to purify recombinant proteins. Epi- (Figure 10-12). They are engineered hybrids of λ phage DNA
tope tags are short protein sequences that are translated and bacterial F plasmid DNA (Chapter 6). Because of their
in-frame, often at the N- or C-terminus of a recombinant cos sites from λ phage, fosmids are packaged into λ phage
(a) (b)
BamHI
Hindlll
EcoRl
Smal
Sphl
Xbal
Kpnl
Sacl
Pst l
Sal l
Polylinker
pUC18 vector pET vector lacO
Polylinker
PT7
amp R
lacZ'
lacI
amp R 2.7 kb lac promoter 5.4 kb
ori ori
Cut foreign DNA and Insert a cDNA into the pET plasmid
vector with restriction enzyme using restriction enzymes and
transform the plasmid into bacteria
FIGURE 10-11 (a) The pUC18 plasmid has been designed for use protein is controlled by three pET plasmid elements, the lac operator
as a vector for DNA cloning. The polylinker has multiple restriction sites (lacO), the lacI gene that encodes the Lac repressor protein, and the
into which donor DNA can be inserted. Insertion of DNA into pUC18 is T7 polymerase promoter. The plasmid also contains a sequence that
detected by inactivation of the β-galactosidase function of lacZ, resulting encodes a His-tag that is translated in-frame with the recombinant
in an inability to convert the artificial substrate X-gal into a blue dye. protein. IPTG induces expression of the His-tagged recombinant protein,
(b) The pET plasmid has been designed for expression and purification which is then purified based on the affinity of the His-tag for Ni2+-coated
of recombinant proteins. Inducible expression of a recombinant beads. [(a) Dr. James M. Burnette III and Dr. Leslie Bañuelos.]
ampR
COS
COS Human
site
site insert
comprises
Fosmid Poly- 35–45 kb ~80% of Single ~75,000
linker fosmid copy
F-factor camR
origin of replication
camR
Human
insert
Poly- comprises
BAC 100–200 kb ~90% of Single 15,000–30,000
linker
BAC copy
F-factor ampR
origin of replication
FIGURE 10-12 Features of some large-insert cloning vectors. The number of clones needed to
cover the human genome once (1 ×) is based on a genome size of 3000 Mb (3 billion base pairs).
particles that introduce these big pieces of recombinant recombinant DNA molecules into bacterial cells: transfor-
DNA into recipient E. coli cells. Cos (an abbreviation for mation, transduction, and infection (Figure 10-13; see also
cohesive) sites are 12 base pair overlapping sticky ends that Sections 6.3, 6.4, and 6.5). In transformation, bacteria are
circularize the linear phage DNA through complementary incubated in a solution containing the recombinant DNA
base pairing. Once introduced into the bacterium, fosmids molecule. Because bacterial cells used in research do not nat-
form circular molecules that replicate extrachromosomally urally take up plasmids, they must be made competent (that
in a manner similar to plasmids. However, because of the is, able to take up the DNA from the surrounding media)
presence of an F plasmid origin of replication that couples by either incubation in a calcium solution (calcium chloride
plasmid replication to host cell chromosome duplication, transformation) or exposure to a high-voltage electrical pulse
very few copies accumulate in a cell. (electroporation). After entering a competent cell through
Bacterial artificial chromosomes (BACs) are another membrane pores, the recombinant molecule becomes a plas-
type of vector for carrying large inserts. Derived from the mid chromosome (Figure 10-13a). Electroporation is the
F plasmid, BACs can carry inserts ranging from 100 to method of choice for introducing especially large DNAs such
200 kb, although the vector itself is only about 7 kb (see as BACs into bacterial cells. In t ransduction, the recombi-
Figure 10-12). The DNA to be cloned is inserted into nant DNA is combined with phage proteins to produce a
the plasmid, and this large circular recombinant DNA is virus that contains largely non-viral DNA. These engineered
introduced into the bacterium. BACs were the workhorse phages inject their DNA into the bacterial cells, but new
vectors for the extensive cloning required by large-scale phages cannot form because they do not carry the viral genes
genome-sequencing projects, including the public project to necessary for phage replication. Fosmids are introduced into
sequence the human genome (Chapter 14). cells by transduction (Figure 10-13b). In contrast to trans-
duction, which produces plasmids and bacterial colonies but
KEY CONCEPT Cloning vectors accept inserts of small sizes not new viruses, infection of bacteria produces recombinant
for plasmids, to medium sizes for bacteriophage, to large sizes phage particles (Figure 10-13c). Through repeated rounds of
for fosmids and BACs. infection, a plaque full of λ phage particles forms from each
initial bacterium that was infected. Each phage particle in
E n t r y o f re c o m b i n a n t D N A m o l e c u l e s i n t o a plaque contains not only the recombinant DNA, but also
b acterial cells Three methods are used to introduce viral genes needed to create new infective phage particles.
Transformation
+ Bacterial colony Bacterial colony
(b) Fosmids
Transduction
+ Bacterial colony
Lysis
FIGURE 10-13 Recombinant DNA can be delivered into bacterial (c) Bacteriophage vectors such as phage λ infect and lyse bacteria,
cells by transformation, transduction, or infection. (a) Plasmid and forming a plaque that contains progeny phages all carrying the identical
BAC vectors are delivered by transformation of purified DNA. recombinant DNA molecule within the phage genome. Examples of
(b) Certain vectors such as fosmids are delivered within bacteriophage bacterial colonies and phage plaques are shown on the right. In the
heads (transduction); however, after having been injected into the bottom picture, uninfected bacteria are found in regions surrounding
bacterium, they form circles and replicate as large plasmids. plaques. [(a) JelenaMV/Medical Images; (b) Branko Velebit/Medical Images.]
Producing cDNA molecules with sticky ends Finding a clone of interest by screening a library
5′ AAAAA 3′
cDNA Colonies of Each colony is a
3′ T T T T T 5′
bacterial cells clone of cells
containing a containing a
Ligate linkers fosmid library fosmid with the
containing EcoRI site same insert
needed to direct the normal expression pattern of Ins1 in Producing PCR products with sticky ends
mice. This includes transcriptional regulatory sequences as
well as RNA processing sequences for splicing of mRNA
introns and polyadenylation of mRNA 3′ ends. However,
the genomic clone of Ins1 cannot be expressed in bacteria Initial steps of PCR
because bacterial proteins do not recognize eukaryotic tran- (heat to 95°C, then cool to
scriptional regulatory sequences and bacteria do not carry anneal and synthesize DNA)
out splicing. In contrast, a cDNA clone of Ins1 cannot be (a)
5′ 3′
expressed in mice because it lacks regulatory sequences for CTT
AAG
transcription and polyadenylation, since they are not tran- 5′ G 5′
AAT
TC
scribed into mRNA. Nevertheless, it is possible to express
the Ins1 cDNA in mice using vectors that contain these 3′ 5′
regulatory sequences from either the Ins1 gene or another
Second round of PCR
mouse gene. Likewise, the Ins1 cDNA cannot be expressed
in bacteria because it lacks bacterial transcriptional regu- (b)
latory sequences; but these sequences can be provided by 5′ GAATTC 3′
CTT
the vector, as described in Figure 10-11b. Furthermore, the AAG
5′
lack of splicing in bacteria is not an issue because splicing 5′ G
AAT
has already taken place. The same considerations apply TC
to genomic and cDNA clones of the human insulin gene, 3′ CTTAAG 5′
which is why a cDNA clone is used to produce recombi-
nant human insulin in bacteria (see Figure 10-1). Further rounds of PCR
(c)
KEY CONCEPT Genomic and cDNA clones of a gene are GAATTC GAATTC
GTTAAG CTTAAG
not functionally interchangeable. Both can be used for gene
expression, but under different conditions.
Digest with EcoRI
(d)
Cloning by PCR 5′ AATTC G
G CTTAA 5′
Since the widespread use of PCR in the 1990s, PCR, rather
than screening a library, is routinely used to construct a par- FIGURE 10-16 Adding EcoRI sites to the ends of PCR products.
ticular genomic or cDNA clone. For example, to clone the (a) A pair of PCR primers is designed so that their 3′ ends anneal to
human insulin cDNA, oligonucleotide primers are designed the target sequence, while their 5′ ends contain sequences encoding
that are complementary to the 5′ and 3′ ends of the insu- a restriction site (EcoRI in this case). The target DNA is denatured,
and 5′ ends with the restriction sites remain single stranded while the
lin cDNA, and PCR is carried out using cDNA generated
rest of each primer anneals and is extended by Taq DNA polymerase.
from pancreas mRNA as the template. To enable cloning of (b) In the second round of PCR—only the newly synthesized strands
a PCR product, a common approach is to use PCR primers are shown—the DNA primers anneal again, and this time DNA
that have restriction sites at their 5′ end (Figure 10-16). Thus, synthesis produces double-stranded DNA molecules with restriction
after digestion with a restriction enzyme, the PCR product sites at one end. (c) The products of all subsequent rounds have EcoRI
sites at both ends. (d) Sticky ends are produced when these PCR
can be ligated into a vector that is linearized with the same
products are cut with EcoRI.
restriction enzyme. A problem with this approach is that
the length of PCR products is limited to about 2 kb. To cir-
cumvent this problem, cDNAs and genes larger than 2 kb between fragments, (2) a DNA polymerase to fill in the gaps
can be cloned by stitching together multiple PCR products between annealed fragments, and (3) a DNA ligase to form
with restriction sites at their ends that direct the order in the final phosphodiester bonds between the fragments. The
which they are assembled (Figure 10-17a). main advantage of assembly over standard restriction enzyme-
As an alternative to restriction enzyme cloning, research- based cloning is that assembly allows the joining of any two
ers have developed other DNA assembly methods that can DNA fragments at any position, whereas restriction enzyme-
be used to construct large genomic regions and cDNAs and based cloning is limited to positions of natural or engineered
even whole chromosomes and genomes. As an example, Gib- restriction sites. In addition, assembly is faster than standard
son assembly can piece together multiple linear DNA frag- restriction enzyme-based cloning because a greater number of
ments that have 15- to 40-bp regions of sequence similarity, DNA fragments can be efficiently joined in a single reaction.
often referred to as homology regions, at their ends (Figure Assembly methods can also be used to put together
10-17b). The fragments can be produced by PCR or by chem- parts from different genes. For example, researchers com-
ical synthesis, which can generate oligonucleotides of up to monly combine the transcriptional regulatory region of
200 nucleotides in length. Assembly is achieved by incubat- one gene with the cDNA of another gene. Returning to the
ing the fragments with three enzymes: (1) an exonuclease insulin gene, this approach could be used to identify the β
that chews back the 5′ ends of each fragment, producing cells in a mouse pancreas. A researcher could put together
long single-stranded regions with sequence complementarity the transcriptional regulatory region of a mouse insulin
(a)
EcoRI A XhoI XhoI B BamHI EcoRI BamHI
GAAT TC CTCGAG CTCGAG GGATCC GAAT TC GGATCC
C T TAAG GAGCTC GAGCTC CCTAGG C T TAAG CCTAGG
V
Digest with restriction enzymes
A B
AAT TC C TCGAG G G GATCC
G GAGCT C CCTAG C T TAA G
V
Anneal DNA fragments
Seal nicks with DNA ligase
A B
GAAT TC CTCGAG GGATCC
C T TAAG GAGCTC CCTAGG
(b)
A B V
5′ 3′ 5′ 3′ 5′ 3′
3′ 5′ 3′ 5′ 3′ 5′
A B V
5′ 3′ 5′ 3′ 5′ 3′
3′ 5′ 3′ 5′ 3′ 5′
A B
FIGURE 10-17 Recombinant DNA molecules with large inserts can be Gibson assembly, (b) regions of sequence similarity (homology regions) at
constructed by assembly methods involving multiple DNA fragments (A and their ends. Both restriction sites and regions of sequence similarity can be
B) and a vector (V) with (a) compatible restriction sites at their ends or, as in built into the DNA fragments by the appropriate design of PCR primers.
gene, which is only expressed in β cells, and a cDNA for a Since the late 1970s, researchers have put consider-
reporter gene that encodes a protein that is easy to detect. able effort into developing techniques to sequence DNA.
The green fluorescent protein (GFP) gene from a jellyfish Currently, the most commonly used technique for small
is commonly used as a reporter because it encodes a small scale sequencing is called dideoxy sequencing or Sanger
protein (238 amino acids) that exhibits bright green flu- sequencing, after its inventor Fred Sanger. However, as
orescence when exposed to ultraviolet light of a particu- we will see in Chapter 14, other sequencing technologies
lar wavelength. In mice that have this engineered gene have largely supplanted this technique when the goal is
inserted into their genome (using techniques described to determine the sequence of an entire genome. The term
later in this chapter), β cells in the pancreas can be iden- dideoxy comes from a modified nucleotide, called a dide-
tified because they will be the only cells that express GFP oxynucleoside triphosphate (ddNTP). This modified nucle-
and glow green. Alternatively, a researcher could construct otide is key to the Sanger technique because of its ability
a gene that expresses a single protein, called a fusion pro- to be added to a growing DNA chain but to block con-
tein, that is comprised of both insulin and GFP amino acid tinued DNA synthesis. A dideoxynucleotide lacks the
sequences. This could be done by assembling an insulin ribose sugar 3′-hydroxyl group as well as the 2′-hydroxyl
gene along with the GFP cDNA such that the insulin and group that is absent in a regular deoxynucleotide
GFP protein-coding regions are translationally in-frame. ( Figure 10-18a ). For DNA synthesis to take place, DNA
In this scenario, the insulin protein is said to be “tagged” polymerase must catalyze phosphodiester bond formation
with GFP. Another way to tag a protein is to append to it between the 3′-hydroxyl group of the last nucleotide in the
a few amino acids that can be recognized by an antibody growing chain and the α-phosphate of the nucleotide to be
in vitro by Western blot analysis and in vivo by immunofluo- added. Because a dideoxynucleotide lacks the 3′-hydroxyl
rescence microscopy. Like His-tags that are used for protein group, this reaction cannot take place, and DNA synthesis
purification, such tags are called epitope tags. Commonly terminates.
used epitope tags for this purpose include the 7-amino-acid
FLAG tag (DYKDDDK), the 9-amino-acid HA tag (YPYD- KEY CONCEPT Dideoxynucleotides (ddNTPs) cause chain
VPDYA), and the 10-amino-acid Myc tag (NNKLISEEDL). termination because they lack the 3′-hydroxyl group on
The advantage of epitope tags over GFP is that, because of the ribose sugar that is essential for phosphodiester bond
formation.
their small size, epitope tags are less likely to alter the struc-
ture and function of the protein to which they are fused.
DNA sequencing requires four separate reactions,
KEY CONCEPT Assembly methods make it relatively easy each containing the DNA segment (e.g., a cloned plasmid
to modify the sequence of genes to create tools that are insert or a PCR product), a radioactive DNA primer that
useful for research, including reporter genes that reveal the will hybridize to exactly one location on the DNA seg-
expression pattern of transcriptional regulatory elements and ment, DNA polymerase, and the four deoxynucleoside
epitope-tagged genes that enable the detection and purifica-
triphosphates (dNTPs: dATP, dCTP, dGTP, and dTTP). In
tion of recombinant fusion proteins.
addition, each reaction receives a small amount of a dif-
ferent dideoxynucleoside triphosphate (ddNTP: ddATP,
ddCTP, ddGTP, or ddTTP). As in DNA replication, the
DNA polymerase will add deoxynucleotides to the 3′ end
10.3 SEQUENCING DNA of the primer, with the identity of the added deoxynucle-
otide being determined by base pairing complementarity to
LO 10.4 Diagram the steps of dideoxy DNA sequencing. the template strand (Figure 10-18b). Since the dNTPs and
each ddNTP are present in a ratio of about 300:1, most of
As described in many chapters in this text, the regulatory the time, DNA polymerase will add a dNTP and continue
and coding information in DNA is determined by its nucle- synthesis. However, every once in a while, DNA polymerase
otide sequence. To reveal this information, public and pri- will incorporate a ddNTP into the new strand, which will
vate institutions have invested heavily in the sequencing of terminate synthesis of that DNA strand. Hence, in the tube
DNA genomes, including the human genome, which was that contains ddATP, new DNA being synthesized will ter-
completed in 2001. Furthermore, since mRNA can be con- minate when a ddATP is added to the strand, thus mark-
verted to cDNA (see Figure 10-9), the same technologies ing locations of T nucleotides in the DNA segment being
used to sequence genomes have been extensively applied sequenced. When complete, the reaction in the ddATP tube
to sequence cDNAs and to discover the information that results in a collection of radiolabeled single-stranded DNA
flows from DNA to RNA by transcription in cells, tissues, fragments of different length, each ending with an A resi-
and organisms. Lastly, sequencing of DNA is a common due. This process is repeated for the reactions with ddCTP,
activity in individual laboratories for purposes such as ddGTP, and ddTTP. The DNA fragments in the four reac-
identifying specific DNA lesions in mutant alleles and con- tions are separated by polyacrylamide gel electrophoresis
firming the sequence of recombinant DNA molecules and and visualized by autoradiography (Figure 10-18c). Since
PCR products. polyacrylamide gels can resolve fragments of DNA that
O O O O O O
−
O P O P O P O O Base −
O P O P O P O O Base
O− O− O− O − O− O−
3′ 2′ 3′ 2′
H H OH H
5′ 3′
G C T C A G C G G T T C T A G T C C A G G C G T T A C G A A T G G C A
T C C G C A A T G C T T A C C G T
3′ H
5′
H
T C A G G T C C G C A A T G C T T A C C G T
3′ H
5′
H
T C G C C A A G A T C A G G T C C G C A A T G C T T A C C G T
3′ H
5′
H
(c) DNA template, labeled primer, DNA polymerase I, and dATP, dCTP, dGTP, and dTTP
A N I M ATED A RT
Dideoxy sequencing of DNA
5′ 3′
Sequence read from the gel Sequence of the DNA template
vary by only one nucleotide in length, the fragments in and red=T. The sequence also contains a single nucleotide
the gel are separated and ordered by size, with the lengths polymorphism (SNP) at position 144. This nucleotide is
increasing by one base at a time. Shorter DNA fragments read as both a T and a C, which means that the diploid
migrate fastest in the gel, so bands at the bottom of the gel individual has one allele with a T–A base pair and the other
represent the sequence closest to the primer. Therefore, the with a C–G base pair. This automated technology was used
sequence is read in the 5′-to-3′ direction from the bottom to to sequence the human genome as well as the genomes of
the top of the gel and is complementary to the DNA strand many other organisms, but more cost-effective and faster
being sequenced. technologies are currently in use for large sequencing proj-
ects (discussed in Chapter 14).
KEY CONCEPT DNA sequencing by the dideoxy (Sanger)
method uses dideoxynucleotides to terminate synthesis by KEY CONCEPT Automated sequencing is superior to the
DNA polymerase from a DNA template, producing DNA frag- original sequencing method that uses a radioactive primer
ments of different lengths that end at each nucleotide position because it involves only one reaction, not four, and it produces
in the template. longer sequencing reads.
FIGURE 10-19 Printout from an automatic sequencer that uses fluorescent dyes. Each of the four A N I M ATED A RT
colors represents a different base (A-green, C-blue, G-black, and T-red). The letter N represents a
base that cannot be assigned because there are overlapping peaks for two bases. This is diagnostic Dideoxy sequencing using fluorescent
of a single nucleotide polymorphism (SNP). Note that, if this was a sequencing gel, as in Figure nucleotides
10-18c, each of the peaks would correspond to one of the dark bands on the gel; in other words, the
colored peaks represent a different readout of the same data produced on a sequencing gel.
organism’s genotype and phenotype is termed genetic engi- Methods of introducing a transgene
neering, and its application for practical purposes is called
biotechnology. (a) Lipid vesicle
The techniques of genetic engineering described in the
first part of this chapter were originally developed in bacte-
ria. These techniques have been extended to model eukary- DNA
otes, which constitute a large proportion of organisms used Lipid
for research. Eukaryotic genes are still typically cloned into bilayer
bacterial vectors, but eventually they are introduced into a
eukaryote, either the original donor species or a completely
different one. The gene transferred is called a transgene,
and the engineered product is called a transgenic organism.
Transgenes are introduced into eukaryotic cells by
chemical, physical, and biological methods (Figure 10-20).
Chemical methods are based on the principle that DNA (b) Electroporation
co-precipitated with minerals such as calcium phosphate or
packaged inside tiny phospholipid vesicles can be taken up
into cells by endocytosis, a natural process by which cells
Electrical
take in molecules from the environment by engulfing them. pulse
Physical methods include electroporation, biolistic parti-
cle delivery, and microinjection. Electroporation involves
applying an electrical field to cells for a short period of
time to create microscopic holes in the plasma membrane
through which DNA can enter. Biolistic particle delivery
systems, also known as gene guns, bombard cells with
DNA-coated metal particles that are small enough to enter
cells but not destroy them. The last physical method, micro-
(c) Biolistic delivery
injection, directly delivers DNA into cells through a fine-
point needle. Biological methods use bacteria or viruses to
transfer DNA into cells. For example, as described shortly,
the bacterium Agrobacterium tumefaciens can transfer to
a plant genome part of its own genome that carries a gene
of interest, and viruses can transfer into animal cells their DNA-coated
genome that is engineered to include a gene of interest. particles
When a transgene enters a cell, it travels to the nucleus,
where it becomes a stable part of the genome by either
inserting into a chromosome or (in a few species) replicat-
ing as part of a plasmid. If insertion occurs, the transgene (d) Microinjection (e) Virus infection
can either replace the resident gene by homologous recom-
bination or insert ectopically—that is, at other locations in
the genome. Transgenes from other species typically insert
ectopically. We now turn to some examples in fungi, plants,
and animals.
(a) (b)
Marker Marker
Plasmid Plasmid
Gene X1 Gene X 1
1 2 1 2
Chromosome Chromosome
Gene X 2 Gene X 2
FIGURE 10-21 A plasmid bearing a functional allele (Gene X 1) inserts at position 2 also are possible but are not shown. The bacterial origin
into a recipient yeast strain bearing a defective allele (Gene X 2 ) by of replication is different from eukaryotic origins, so bacterial plasmids
homologous recombination. The result can be (a) replacement of the X 2 do not replicate in yeast. Therefore, the only way in which such vectors
by X 1 or (b) retention of X 2 and concurrent addition of X 1. The mutant can generate a stable modified genotype is if they are integrated into the
site of Gene X 2 is represented as a vertical black bar. Single crossovers yeast chromosome.
gene, by either a single or a double crossover (Figure 10-21). bacterium causes what is known as crown gall disease, in
As a result, either the entire plasmid is inserted, or the tar- which the infected plant produces uncontrolled growths
geted allele is replaced by the allele on the plasmid. The called tumors or galls. The key to tumor production is a
latter is an example of gene replacement—in this case, the large (200 kb) circular DNA plasmid—the Ti (tumor-
substitution of an engineered gene for the gene originally inducing) plasmid ( Figure 10-22 ) . When the bacterium
in the yeast cell. Gene replacement can be used to delete infects a plant cell, a part of the Ti plasmid is transferred
a gene or substitute a mutant allele for its wild-type coun- and inserted, apparently more or less at random, into the
terpart or, conversely, to substitute a wild-type allele for a genome of the host plant. The region of the Ti plasmid
mutant. that inserts into the host plant is called T-DNA, for trans-
fer DNA. The genes whose products catalyze this T-DNA
KEY CONCEPT Transgenic yeast cells are generated by
homologous recombination between a yeast chromosome and
a plasmid that is transformed into yeast and carries a gene of
interest. The Ti plasmid vector
transfer reside in a region of the Ti plasmid separate from Generation of a transgenic plant
the T-DNA region itself.
The natural behavior of the Ti plasmid makes it well
suited to the role of a vector for plant genetic engineering.
In particular, any DNA that is inserted between the left and
right T-DNA border sequences (24 base-pair ends) can be
mobilized by other functions provided by the Ti plasmid Make leaf
and inserted into plant chromosomes. Thus, scientists are discs
able to eliminate all of the T-DNA sequence between the
borders (including the tumor-causing genes) and replace it
with a gene of interest and a selectable marker (for exam-
Briefly culture leaf
ple, kanamycin resistance). One method of introducing the fragments in genetically
T-DNA into the plant genome is shown in Figure 10-23 . modified Agrobacter
Bacteria containing an engineered Ti plasmid are used to
infect cut segments of plant tissue, such as punched-out leaf Transfer to filter paper
discs. If the leaf disks are placed on a medium containing over nurse cells
kanamycin, only the plant cells that have acquired the kanR
gene engineered into the T-DNA will undergo cell division. Culture 2–3 days
Transformed cells grow into a clump, or callus, that can Nurse cells produce
be induced to form shoots and roots. These calli are trans- growth factors
ferred to soil, where they develop into transgenic plants.
Transfer to shoot
Typically, only a single copy of the T-DNA region inserts stimulation medium
into a given plant genome, where it segregates at meiosis
like a regular Mendelian allele. Therefore, one-quarter of
the progeny from crossing of the original transgenic plants
will get two copies of the T-DNA. The presence of the insert Plantlet
can be verified by Southern blot analysis of purified DNA
with a T-DNA probe or by PCR using primers specific for
the T-DNA.
Transgenic plants carrying any one of a variety of for-
eign genes are in current use, including crop plants carrying
genes that confer resistance to certain bacterial or fungal
pests, and many more are in development. Not only are the
qualities of plants themselves being manipulated, but, like
Transgenic
microorganisms, plants are also being used as convenient tobacco plant
“factories” to produce proteins encoded by foreign genes.
(a)
Syncytial Gonad
Micropipette with region
transgene DNA Nuclei
C. elegans Egg
Integrated array
Chromosome
FIGURE 10-24 Transgenic C. elegans are created by injecting transgene DNA directly into a gonad.
(a) The method of injection. (b) The two main types of transgenic results: extrachromosomal arrays
and arrays integrated in ectopic chromosomal locations.
which the DNA is injected becomes continuous with the for DNA sequences at the targeted gene. For knockout
nucleoplasm). Typically, the transgenic DNA forms multi- mice, part or all of the targeted gene is deleted, or a DNA
copy extrachromosomal arrays (Figure 10-24b) that exist sequence is inserted into the targeted gene to disrupt its
as independent units outside the chromosomes. The arrays expression, thereby creating a loss-of-function mutation.
are stably inherited, but not with the same efficiency as
chromosomes. More rarely, the transgenes will become Ectopic insertions To insert transgenes in random loca-
integrated into an ectopic position in a chromosome, still as tions, a solution of bacterially cloned DNA is injected into
a multicopy array. either the male or female pronucleus of a fertilized egg
(Figure 10-25). Several injected eggs are inserted into the ovi-
KEY CONCEPT Transgenic worms are generated by injection duct of a recipient mouse. Progeny are analyzed for integra-
of a plasmid containing a gene of interest into the gonad. The tion of the transgene. Typically, DNA extracted from a piece
plasmid is typically stably inherited as a multicopy extrachro- of the tail is used for Southern blot analysis or PCR analysis
mosomal array. for the transgene. Occasionally, mice are mosaic; that is, not
every cell contains the transgene because DNA integration
occurred at a two-cell or later stage of embryogenesis. Pos-
Transgenesis in M. musculus Mice are a very important
itive mice are subsequently mated, their offspring are ana-
model for mammalian genetics because they are relatively
lyzed for transgene expression, and positive mice are used
easy to breed and genetically manipulate. Furthermore,
to establish transgenic mouse lines with stable integration
many of the technologies developed in mice and biological
and expression of the transgene. The technique gives rise to
insights gained from studies of mice are potentially applica-
some problems: (1) the expression pattern of the randomly
ble to humans. There are two strategies for transgenesis in
inserted genes may be abnormal due to position effects
mice, each having its advantages and disadvantages:
from the local chromatin environment (see Chapter 12 for
• Ectopic insertions. Transgenes are inserted randomly in more on position effects), and (2) DNA rearrangements can
the genome, usually as multicopy arrays. Mice generated occur inside the multicopy arrays (in essence, mutating the
with an ectopic insertion are called transgenic mice. sequences). Nonetheless, this technique is much more effi-
• Gene targeting. Transgenes are inserted into a location cient and less laborious than gene targeting. Because of the
occupied by a homologous sequence in the genome. ease of generating mice with ectopic insertions, it has been
That is, the transgene replaces its normal homologous used to produce human antibodies for use as therapeutics.
counterpart. Mice generated by gene targeting are called
knock-in or knockout mice. For knock-in mice, new Gene targeting Gene targeting enables researchers to
DNA either is added to the targeted gene or is substituted eliminate a gene or modify its function. In one application,
Generation of transgenic mice Figure 10-26a , the gene was inactivated by insertion of
the neomycin-resistance gene (neoR ) into a protein-coding
region (exon 2) of the gene. The mutant gene was then
cloned into a vector containing the herpes virus thymidine
Pipette to
Microinjection kinase ( tk ) gene. In later steps, the neoR gene will serve
hold egg
needle as a marker to indicate that the transgene inserted in a
Fertilized mouse egg containing chromosome, and loss of the tk gene will ensure that the
transgene
DNA transgene is inserted at the homologous locus rather than
randomly in a chromosome (Figure 10-26b). These mark-
ers are standard, but others could be used instead. The
cloned DNA is microinjected into the nucleus of cultured
Pronuclei ES cells. The defective gene inserts far more frequently
into nonhomologous (ectopic) sites than into homologous
sites, so the next step is to select the rare cells in which the
defective gene has replaced the functioning gene as desired
(Figure 10-26c). To isolate cells carrying a targeted muta-
tion, the cells are cultured in medium containing drugs—
here, a neomycin analog (G418) and ganciclovir. G418 is
FIGURE 10-25 Transgenic mice are created by injection of lethal to cells unless they carry a functional neoR gene,
cloned DNA into fertilized eggs and subsequent insertion in ectopic
and so it eliminates cells in which no integration of vector
chromosomal locations. [RAPHO AGENCE/Science Source.]
DNA has taken place (yellow cells). Meanwhile, ganciclo-
ANIMATED ART vir kills any cells that harbor the tk gene, thereby elim-
Creating a transgenic mouse inating cells bearing a randomly integrated vector (red
cells). Consequently, the only cells that survive and pro-
liferate are those harboring the targeted insertion (green
called gene replacement, a mutant allele can be repaired cells).
by substituting a wild-type allele in its normal chromo- Stage 2: ES cells that contain one copy of the disrupted
somal location. Gene replacement avoids both position gene of interest, that is, a gene knockout, are injected into
effects and DNA rearrangements associated with ectopic a blastocyst-stage embryo, which is then implanted in a
insertion, because a single copy of the gene is inserted in surrogate mother ( Figure 10-27a ). Some of the ES cells
its normal chromosomal environment. Conversely, a gene may become incorporated into the host embryo, and if
may be inactivated by substituting an inactive gene for the that happens, the mouse that develops will be chimeric —
normal gene. Such a targeted inactivation is called a gene that is, it will contain cells from two different mouse
knockout. strains.
Gene targeting in mice is carried out in cultured embry- Stage 3: When the chimeric mouse reaches adulthood, it
onic stem cells (ES cells). In general, a stem cell is an is mated with a normal mouse. If the chimeric mouse con-
undifferentiated cell in a given tissue or organ that divides tained germ-line cells that were derived from the ES cells
asymmetrically to produce a progeny stem cell and a cell (with the knockout gene), some of the resulting offspring
that will differentiate into a terminal cell type. ES cells are will inherit the gene knockout in all their cells. Sibling mice
special stem cells called pluripotent stem cells that can dif- that are identified by Southern blot or PCR analysis as
ferentiate to form any cell type in the body—including, being heterozygous for the knockout version of the gene of
most important, the germ line. interest are then mated to produce mice that are homozy-
To illustrate the process of gene targeting, we look at gous for the knockout allele. If the gene is essential, homo-
how it achieves one of its typical outcomes—namely, the zygotes will be lethal, and none will be obtained from this
substitution of an inactive gene for the normal gene, or cross (Figure 10-27b).
gene knockout. The process requires three stages:
KEY CONCEPT Knock-in and knockout mice are gener-
1. An inactive gene is targeted to replace the functioning ated by homologous recombination in ES cells between a
gene in a culture of ES cells, producing ES cells containing mouse chromosome and a plasmid. Gene-targeted ES are
a gene knockout (Figure 10-26). then injected into embryos to generate chimeric mice that are
2. ES cells containing the inactive gene are transferred to crossed to assess heritable germ-line transmission.
mice embryos (Figure 10-27a).
3. Knockout mice are identified and bred to produce mice KEY CONCEPT Germ-line transgenic techniques have been
of known genotype (Figure 10-27b). developed for all well-studied eukaryotic species. These tech-
niques depend on an understanding of the reproductive biol-
Stage 1: A recombinant DNA molecule is generated
ogy of the recipient species.
that disrupts a gene of interest. In the example shown in
Gene of
interest
neoR
Targeting Targeting
Targeting vector vector
vector
+ + ×
Nonhomologous Nonhomologous Unchanged
region in region in chromosome
Homologous Chromosome chromosome Chromosome chromosome
gene in with targeted with random
chromosome insertion insertion
Neomycin
analog
(G418) Ganciclovir Cell with no
insertion
Add to medium
FIGURE 10-26 Producing cells that contain a mutation in one specific between but excluding the marker at the tip, take the place of the original
gene, known as a targeted mutation or a gene knockout. (a) The gene of gene. This event is important because the vector sequences serve as a
interest (cloned gene) is inactivated by insertion of the neoR gene in exon 2 useful tag for detecting the presence of this mutant gene. In many cells,
and cloned into a targeting vector containing the tk gene. Copies of a though, the full vector (complete with the extra marker at the tip) inserts
cloned gene are altered in vitro to produce the targeting vector. The vector ectopically (middle) or does not become integrated at all (right). (c) To
is then injected into ES cells. (b) When homologous recombination occurs isolate cells carrying a targeted mutation, all of the cells are cultured in
(left), the homologous regions on the vector, together with any DNA in media containing drugs to select for cells containing the targeted insertion.
(b)
a / a ; M/ M
plus
a/a;M/M a/a;m/m
A/ A; M/ m
a / a ; M/ M A/a;M/m A / A or a ; M / M or m
A/a;M/M a / a ; M / M or m
FIGURE 10-27 A knockout mouse is produced by inserting embryonic different strains of mice. Solid black coloring, in contrast, indicates that
stem (ES) cells carrying the targeted mutation into an embryo. (a) ES the ES cells have perished, and these mice are excluded. A represents
cells are isolated from an agouti (brown) mouse strain (A/A) and altered agouti; a, black; m is the targeted mutation; and M is its wild-type allele.
to carry a targeted mutation (m) in one chromosome. The ES cells are (b) Chimeric males are mated with black (nonagouti) females. Progeny
then inserted into young embryos, one of which is shown. The coat are screened for evidence of the targeted mutation (green in inset) in
color of the future newborns is a guide to whether the ES cells have the gene of interest. Direct examination of the genes in the agouti mice
survived in the embryo. Hence, ES cells are typically put into embryos reveals which of those animals (boxed) inherited the targeted mutation.
that, in the absence of the ES cells, would acquire a totally black coat. Males and females carrying the mutation are mated with one another
Such embryos are obtained from a black strain that lacks the dominant to produce mice whose cells carry the chosen mutation in both copies
agouti allele (a /a) . Embryos containing the ES cells grow to term in of the target gene (inset) and thus lack a functional gene. Such animals
surrogate mothers. Agouti shading intermixed with black indicates those (boxed) are identified definitively by direct analysis of their DNA. The
newborns in which the ES cells have survived and proliferated. Such knockout in this case results in a curly-tail phenotype.
mice are called chimeras because they contain cells derived from two
CRISPR-Cas9 genome engineering to occur at a particular place in the genome, repair by NHEJ
An alternative approach to transgenesis for engineering would create mutations that are likely to result in inactivation
genomes takes advantage of the natural ability of cells to repair of the targeted gene, while repair by HR using a homologous
DNA double-strand breaks (DSBs), a topic covered in detail in donor DNA with an altered sequence would create mutations
Chapter 15. In brief, DSBs in eukaryotes are usually repaired in the gene. To date, three technologies have been developed
by nonhomologous end joining (NHEJ), a mechanism that that create site-specific DSBs: zinc-finger nucleases (ZFNs),
reattaches the two chromosomal pieces but in a sloppy fash- transcription activator-like effector nucleases (TALENs), and
ion, causing nucleotides to be inserted or deleted at the site CRISPR RNA-guided Cas nucleases (CRISPR-Cas). ZFNs and
of the DSB. Alternatively, DSBs are repaired by homologous TALENs are proteins that contain two functional domains: a
recombination (HR), which fixes the break without errors domain with DNA-binding activity that is designed to bind a
using a homologous donor DNA (e.g., a sister chromatid or specific DNA sequence, and a domain with non-specific DNA
a plasmid). Therefore, if DSBs could somehow be directed endonuclease activity that produces DSBs. When expressed
in cells, ZNFs and TALENs bind their targeted sequence in trans-activating CRISPR RNA (tracrRNA), and the dual RNAs
genomic DNA and generate a DSB at a nearby sequence that is form a complex with a Cas protein called Cas9. The gRNA
then repaired by NHEJ or HR. In contrast, in the CRISPR-Cas then directs Cas9 to produce a DSB at a location in the for-
system, base pairing between a noncoding RNA and genomic eign DNA that contains the complementary 20-nucleotide tar-
DNA targets the Cas endonuclease to generate a DSB at a get sequence located next to a trinucleotide NGG protospacer
specific place in the genome. A major technical advantage of adjacent motif (PAM). Cas9 has two separate endonuclease
CRISPR-Cas over ZNFs and TALENs is that it is much eas- domains that cut three nucleotides upstream of the PAM; one
ier to produce RNAs than DNA-binding domains that bind a domain cuts the target strand that base pairs to the gRNA, and
unique sequence in a DNA genome. the other cuts the nontarget strand.
The CRISPR-Cas (clustered, regularly interspaced short In 2012, the laboratories of Emmanuelle Charpentier and
palindromic repeats-CRISPR-associated protein) technology is Jennifer Doudna demonstrated that when a gRNA and a
derived from a bacterial immune system that protects bacteria tracrRNA are modified to be in the same transcript, called a
against foreign plasmid and bacteriophage DNA (as discussed single guide RNA (sgRNA), the sgRNA retains the ability of
in Chapter 6). For instance, the bacterium Streptococcus pyo- the separate RNAs to assemble with Cas9, base pair to tar-
genes, which contains a relatively simple CRISPR-Cas system, geted sequences in DNA, and activate the endonuclease activ-
stores the memory of encounters with foreign DNA by inte- ities of Cas9. This simplification made it practical to employ
grating 20-nucleotide sequences from the foreign DNA into the CRISPR-Cas9 system in the laboratory to make modifica-
a particular place in the bacterial genome called a CRISPR tions to eukaryotic genomes with high efficiency and specific-
array. When S. pyogenes is attacked for a second time by the ity. Application of the CRISPR-Cas9 technology involves two
same foreign DNA, guide RNAs (gRNAs) that contain the plasmids: one expresses Cas9 protein, and the other expresses
20-nucleotide sequences are produced by cutting apart a long an sgRNA with a 20-nucleotide guide sequence that is designed
RNA transcript from the CRISPR array. A sequence common by a researcher to be complementary to a specific genomic
to all gRNAs base pairs with another noncoding RNA called a site adjacent to a PAM (Figure 10-28). After both plasmids are
5′ 3′
Donor DNA
DNA repairs
at the DSB site
Precise mutation
Substitution
Small
deletion
Random mutations
Left Right
homology homology
arm arm
GFP
Donor plasmid
GFP
Cas9
(c)
sgRNA
Mouse
Ins1 gene Genomic DNA
Left Right
homology homology
arm arm
Donor plasmid
(d)
AD
sgRNA
Mouse Ins1 gene
Genomic DNA
Cas9
FIGURE 10-29 Four examples of the use of CRISPR-Cas9 to gene. (c) Repair by homologous recombination of a single DSB in the
manipulate the mouse Ins1 gene. (a) Two sgRNAs are designed to Ins1 gene using an altered Ins1 donor DNA can generate missense
target DSBs to the ends of the Ins1 gene. The chromosome break is and nonsense mutations as well as insertions and deletions in the Ins1
then repaired by nonhomologous end joining, which deletes the Ins1 gene. (d) sgRNA-mediated targeting of a Cas9 protein with inactivated
gene and may introduce other sequence changes, indicated by the endonuclease domains fused to a transcriptional activation domain (AD)
red line. (b) Addition of a donor plasmid triggers repair by homologous yields enhanced transcription of the Ins1 gene (bold arrow).
recombination, which, in this case, replaces the Ins1 gene with the GFP ANIMATED ART CRISPR
introduced into cells or organisms, Cas9 protein and sgRNA gene. An advantage of this approach is that mice carrying
are expressed and form a complex that produces a DSB in the null alleles of Ins1 can be easily identified because they will
targeted gene. Inaccurate repair of the DSB by NHEJ causes express GFP in β cells under control of Ins1 transcriptional
gene inactivation. In contrast, specific mutations in a gene can regulatory sequences. Lastly, a donor plasmid could contain
be introduced by inclusion of a third plasmid, a donor plasmid, a missense or nonsense mutation that inactivates or alters
that is used for repair of the DSB by HR because it contains the function of the Ins1 gene (Figure 10-29c).
sequences identical to those that flank the site of cleavage in Researchers have also modified the CRISPR-Cas9 sys-
addition to the specified mutations. tem to manipulate gene expression in specific ways. The
CRISPR-Cas9 gene editing occurs in reproductive cells basic idea behind these technologies is that, when com-
of many organisms, making targeted gene modifications plexed with an sgRNA, a mutant Cas9 protein lacking its
heritable. Therefore, CRISPR-Cas9 can be used for essen- endonuclease activity can transport any protein or protein
tially all of the same reverse genetic purposes as transgen- domain to a specific place in the genome. As shown in Fig-
esis and even more. For example, to study the phenotypic ure 10-29d, transcription of the mouse Ins1 gene can be
consequences of loss of one of the two insulin genes in mice, activated by sgRNA-mediated targeting to the Ins1 gene of
a null allele of Ins1 can be generated by CRISPR-Cas9- a Cas9 protein that is converted into a transcription fac-
mediated deletion of the Ins1 gene. This is accomplished tor by the addition of a transcriptional activation domain.
by expressing two sgRNAs that create DSBs at the 5′ and Conversely, Ins1 transcription can be turned off by fusion
3′ ends of the Ins1 gene, which will cut out the Ins gene of a transcriptional repression domain to Cas9. One can
and trigger repair of the broken chromosome by NHEJ imagine that someday CRISPR-Cas9 technologies will be
(Figure 10-29a). Alternatively, the Ins1 gene can be replaced used to treat individuals with type I diabetes by manipulat-
by a reporter gene such as GFP. This is done by express- ing the sequence or expression of genes.
ing the same two sgRNAs as well as a donor plasmid con-
KEY CONCEPT The CRISPR-Cas9 system efficiently and
taining the GFP gene flanked by homology arms, sequences
specifically changes the sequence of targeted genes in an
identical to those found upstream and downstream of the
organism, and modified versions of the system alter gene
Ins1 gene (Figure 10-29b). Repair of the DSBs by HR using
expression without changing gene sequences.
the donor plasmid will replace the Ins1 gene with the GFP
SUMMARY
Recombinant DNA is constructed in the laboratory to allow short adapters containing restriction sites to their ends
researchers to manipulate and analyze DNA segments (donor before insertion into the vector. Assembly methods that do
DNA) from any genome or a DNA copy of mRNA. Three not require restriction sites have made the construction of
sources of donor DNA are (1) genomes digested with restric- recombinant DNA molecules more flexible and efficient.
tion enzymes, (2) PCR products of specific DNA regions, There are a wide variety of bacterial vectors. The choice
and (3) cDNA copies of mRNAs. Sequencing of DNA by the of vector depends largely on the size of DNA fragment to
dideoxy (Sanger) method is used to confirm the accuracy of be cloned. Plasmids are used to clone small genomic DNA
recombinant DNA molecules and also to discover informa- fragments, PCR products, or cDNAs. Intermediate-size frag-
tion stored in genomic DNA and mutant genes. ments, such as those resulting from digestion of genomic
The polymerase chain reaction is a powerful method DNA, can be cloned into modified versions of λ bacterio-
for direct amplification of a small sequence of DNA from phage (for inserts of 10–15 kb) or into phage–plasmid hybrids
within a complex mixture of DNA, without the need for a called fosmids (for inserts of 35–45 kb). Finally, bacterial arti-
host cell or very much starting material. The key is to have ficial chromosomes (BACs) are used routinely to clone very
primers that are complementary to flanking regions on large genomic fragments ( ~100–200 kb). A variety of plas-
each of the two DNA strands. These regions act as sites for mids have been developed that contain features that make it
polymerization. Multiple rounds of denaturation, anneal- easier to clone DNA fragments and to control the expression
ing, and extension amplify the sequence of interest expo- of constituent genes in different organisms.
nentially. The vector-donor DNA construct is amplified inside
To insert donor DNA into vectors, donor and vector bacterial host cells as extrachromosomal molecules that are
DNA are cut by the same restriction endonuclease, joined replicated when the host is replicating its genome. Ampli-
by annealing the sticky ends that result from digestion, and fication of plasmids, phages, and BACs results in clones
ligated to covalently join the molecules. PCR products and containing multiple copies of each recombinant DNA con-
cDNA molecules are inserted into vectors by first adding struct. In contrast, only a single fosmid is present in each
restriction sites to the 5′ end of PCR primers or by ligating bacterial cell.
Often, finding a specific clone with a gene of interest Transgenes are engineered DNA molecules that are
requires the screening of a genomic library, a set of clones, introduced and expressed in eukaryotic cells. They can be
ligated in the same vector, that together represent all regions used to engineer a novel mutation or to study the regulatory
of the genome of the organism in question. The number of sequences that constitute part of a gene. Transgenes can be
clones that constitute a genomic library depends on (1) the introduced as extrachromosomal molecules, or they can be
size of the genome in question, and (2) the insert size toler- integrated into a chromosome, either in random (ectopic)
ated by the particular cloning-vector system. Similarly, a locations or in place of the homologous gene, depending
cDNA library is a representation of the total mRNA set pro- on the system. Typically, the mechanisms used to introduce
duced by a tissue or developmental stage in a given organism. a transgene depend on an understanding and exploitation
Hybridization with single-stranded nucleic acid probes of the reproductive biology of the organism. New genome
is fundamental to both in vitro and in vivo methods for engineering methods like the CRISPR-Cas9 system are
identifying DNA fragments or RNAs of interest. These being developed whose defining features are the creation
methods include Southern blotting for DNA, Northern and repair of site-specific DNA double-strand breaks. These
blotting for RNA, and screening of genomic and cDNA methods have opened the door to new and exciting reverse
libraries. In contrast, labeled antibodies are probes for iden- genetic studies in a wide variety of eukaryotic organisms,
tifying specific proteins from complex mixtures in Western and they could potentially be used for gene-editing thera-
blotting or immunofluorescence. pies for patients with serious diseases.
KEY TERMS
antibody (p. 336) fluorescence in situ hybridization polymerase chain reaction (PCR)
autoradiography (p. 336) (FISH) (p. 337) (p. 339)
bacterial artificial chromosome (BAC) fosmid (p. 344) position effect (p. 357)
(p. 346) gel electrophoresis (p. 334) probe (p. 336)
cDNA library (p. 347) gene knockout (p. 358) quantitative PCR (qPCR) (p. 341)
chimera (chimeric) (p. 358) gene replacement (p. 358) recombinant DNA (p. 342)
clustered, regularly interspaced short genetically modified organism restriction enzyme (p. 338)
palindromic repeats (CRISPR) (GMO) (p. 355) restriction fragment (p. 338)
(p. 361) genetic engineering (p. 332) restriction map (p. 339)
complementary DNA (cDNA) genomic library (p. 347) restriction site (p. 338)
(p. 341) genomics (p. 332) reverse transcriptase (p. 341)
CRISPR-associated protein (Cas) homologous recombination (HR) reverse transcription-PCR (RT-PCR)
(p. 361) (p. 360) (p. 342)
dideoxy (Sanger) sequencing (p. 351) hybridization (p. 336) single guide RNA (sgRNA) (p. 361)
DNA amplification (p. 339) immunofluorescence (p. 337) Southern blotting (p. 334)
DNA assembly (p. 349) infection (p. 346) Ti plasmid (p. 355)
DNA cloning (p. 342) in situ hybridization (ISH) (p. 337) transduction (p. 346)
DNA ligase (p. 342) multiple cloning site (MCS) transformation (p. 346)
DNA linker (DNA adapter) (p. 347) (polylinker) (p. 342) transgene (p. 354)
DNA technologies (p. 332) nonhomologous end joining (NHEJ) transgenic organism (p. 354)
donor DNA (insert DNA) (p. 342) (p. 360) vector (p. 342)
ectopic (ectopically) (p. 354) Northern blotting (p. 334) Western blotting (p. 334)
epitope tag (p. 351) palindrome (palindromic) (p. 338)
plasmid (p. 344)
SOLVED PROBLEMS
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
WORKING WITH THE FIGURES 11. In Figure 10-8, if Sample A had a C T value of 24 and
Sample B had a C T value of 27, which sample had
(The first 33 questions require inspection of text figures.)
more DNA and how much more?
1. In the opening figure, what would happen if the cathode
12. In Figure 10-9, explain why the polymerase used for
and anode were switched during gel electrophoresis?
synthesizing DNA from an RNA template is called
2. In Figure 10-1, by what methods could the plasmid be reverse transcriptase.
introduced into bacteria?
13. In Figure 10-10a, in the second step of the procedure,
3. In Figure 10-2 , why can both DNA and RNA be label the 5′ and 3′ ends of the linearized vector and the
used as probes for both Southern and Northern blot EcoRI insert.
analysis?
14. In Figure 10-11a, which colonies (blue or white) con-
4. In Figure 10-3a, what changes occur to the human Ins tain plasmids with a DNA insert?
pre-mRNA to produce the mature Ins mRNA?
15. In Figure 10-12, determine approximately how many
5. In Figure 10-3b, what size fragment would probe 2 BAC clones are needed to provide 1× coverage of
detect in Southern blot analysis of mouse genomic
a. the yeast genome (12 Mbp).
DNA digested with Pvu II, and what site fragment
would probe 3 detect in an NsiI digest? b. the E. coli genome (4.1 Mbp).
6. In Figure 10-4, which of the seven lanes contains the c. the fruit fly genome (130 Mbp).
smallest piece of DNA? 16. In Figure 10-13, what is the difference between plas-
7. In Figure 10-5, what is the purpose behind transferring mid transformation and fosmid transduction?
the nucleic acid from a gel to a membrane? 17. In Figure 10-14, is it possible for more than one insert
8. In Figure 10-6c, what size band would be detected by to ligate into a single vector? Why or why not?
probe 1 in Southern blot analysis of human smooth 18. In Figure 10-15, how is screening a genomic library
muscle cell genomic DNA digested with EcoRI? similar to Southern blot analysis?
9. Examine Table 10-1. Draw the staggered ends pro- 19. In Figure 10-16, how would you modify the restriction
duced by digestion with NotI and the blunt ends pro- sites in the primers so that PCR products only insert
duced by MspI. into a vector in one orientation?
10. In Figure 10-7, why would PCR not work if Taq DNA 20. In Figure 10-17a, what enzymes would you use to cut
polymerase was replaced with a DNA polymerase a full-length insert back out of the vector?
from human cells? 21. In Figure 10-18a, draw a ribonucleotide that would
act as a chain terminator during transcription.
22. In Figure 10-18b, analogous to the drawing for the 32. In Figure 10-28 , how would you determine the
sequencing reaction that contains ddTTP, write the sequence change resulting from nonhomologous end
sequence of the first three termination products of the joining (NHEJ) at a double-strand break site created
sequencing reaction that contains ddCTP. by CRISPR-Cas9?
23. In Figure 10-19, what would happen to the height of 33. In Figure 10-29b, draw an analogous figure that shows
the peak at the SNP position in an individual that did a scheme for replacing the second exon in Ins1 with
not contain a SNP? the GFP gene.
24. In Figure 10-20, why are multiple procedures needed
BASIC PROBLEMS
for introducing DNA into cells?
25. In Figure 10-21, how can the marker be used to deter- 34. Why is a range of temperatures indicated in the anneal-
mine if a single or double crossover event occurred? ing step of PCR?
26. In Figure 10-22, what is the purpose of the select- 35. In the PCR process, if we assume that each cycle takes
able marker? Provide two examples of selectable 5 minutes, what fold amplification would be accom-
markers. plished in 1 hour?
27. In Figure 10-23, do all of the cells of a transgenic plant 36. Would the blot be called a Southern, Northern, or
grown from one clump of cells contain T-DNA? Justify Western blot if RNA was on the membrane and the
your answer. probe was single-stranded DNA?
28. In Figure 10-24, what is distinctive about the syncytial 37. How can genomic and cDNA sequences be used to
region that makes it a good place to inject DNA? determine where introns are located in genes? Use
Figure 10-3a to illustrate your answer.
29. In Figure 10-25, why do the fertilized eggs have two
nuclei? What is the ploidy of each nucleus? 38. In Figure 10-10, it is possible that two or more cop-
ies, rather than one copy, of the insert ligated into the
30. In Figure 10-26c, does the selection procedure distin- vector. How would you test this possibility (a) using
guish whether the targeting vector inserted in one or restriction enzymes, (b) by PCR, and (c) by Southern
both copies of the homologous gene in the diploid ES blot? Hints: The complete sequence of the vector is
cells? known, and the EcoRI site is one of many restriction
31. In Figure 10-27a , why are chimeric males, rather sites in the polylinker of the vector.
than females, used in the mating crosses to generate a 39. Write out the sequence of 20-nucleotide primers to be
homozygous mutant mouse line? used for PCR to amplify the region of interest in the
following piece of DNA:
Region of interest
5′–CCGTAACACGTCAGGGCCTAACAGG TTGACAATGCCTGGAATTCTGTAAC–3′
3′–GGCATTGTGCAGTCCCGGATTGTCC AACTGTTACGGACCTTAAGACATTG–5′
40. Draw a diagram that explains how automated dideoxy 44. In Figure 10-26, describe how positive-negative selec-
sequencing reactions are analyzed by capillary gel elec- tion is used to find rare homologous recombination-
trophoresis and laser detection. mediated gene targeting events.
41. Explain why dideoxynucleosides, which are converted 45. In T-DNA transformation of a plant with a transgene
to dideoxynucleotides in human cells, are effective from a fungus (not found in plants), the presumptive
drugs to block replication of the human immuno- transgenic plant does not display the expected pheno-
deficiency virus (HIV) genome by the HIV reverse type of the transgene. How would you determine
transcriptase. whether the transgene is in fact inserted in the plant
genome? How would you determine whether the
42. Compare and contrast the use of the word recombi- transgene mRNA and protein are expressed in the
nant as used in the phrases (a) “recombinant DNA” plant?
and (b) “recombinant frequency.” 46. Why was cDNA and not genomic DNA used to express
43. Why is DNA ligase needed to make recombinant human insulin in E. coli?
DNA? What would be the immediate consequence in 47. Based on the information presented in Figures 10-26
the cloning process if DNA ligase was not included in and 10-29, explain how CRISPR-Cas9 could be used
the reaction? to knock out the Ins1 gene in mice. In particular, what
RNAs and proteins would be expressed from plasmids suspected of being an actin mutant, how would you
that are injected into fertilized mouse eggs? use (a) restriction enzyme cloning and sequencing, (b)
PCR and sequencing, and (c) restriction enzyme clon-
CHALLENGING PROBLEMS ing and functional complementation (rescue) to deter-
mine whether your suspicion is correct?
48. Using the information in Figures 10-3 and 10-26a ,
draw a targeting vector that could be used to tag the 53. Bacterial glucuronidase converts a colorless substance
human insulin protein with GFP at the N-terminus. called X-Gluc into a bright blue indigo pigment. The
gene for glucuronidase also works in plants if given a
49. Diagram how Gibson assembly could be used to
plant promoter region. How would you use this gene as
construct the targeting vector for question 48.
a reporter gene to find the tissues in which a plant gene
50. A cloned fragment of DNA was sequenced by using that you have just cloned is normally active? (Assume
the dideoxy method. A part of the autoradiogram of that X-Gluc is easily taken up by the plant tissues.)
the sequencing gel is represented here.
54. The plant Arabidopsis thaliana was transformed using
ddA ddG ddT ddC the Ti plasmid into which a kanamycin-resistance
gene had been inserted in the T-DNA region. Two
kanamycin-resistant colonies (A and B) were selected,
and plants were regenerated from them. The plants
were allowed to self-pollinate, and the results were as
follows:
3
Plant A selfed → progeny resistant to kanamycin
4
1
progeny sensitive to kanamycin
4
15
Plant B selfed → progeny resistant to kanamycin
16
1
progeny sensitive to kanamycin
16
a. Draw the relevant plant chromosomes in both
plants.
b. Explain the two different ratios.
a. Write out the nucleotide sequence of the DNA
molecule synthesized from the primer. Label the
5′ and 3′ ends. GENETICS AND SOCIETY
b. Write out the nucleotide sequence of the DNA In 2018, a researcher claimed to have used the CRISPR-Cas9
molecule used as the template strand. Label the genome editing technique to produce the world’s first gene-
5′ and 3′ ends. edited babies. The researcher announced that they edited the
CCR5 gene in two embryos, which were then implanted in
c. Write out the nucleotide sequence of the DNA
a woman. CCR5 encodes a receptor that is expressed on the
double helix. Label the 5′ and 3′ ends.
surface of white blood cells and other cells, where it coordi-
51. Transgenic tobacco plants were obtained in which the nates immune responses. CCR5 is also the main receptor used
vector Ti plasmid was designed to insert the gene of by the human immunodeficiency virus (HIV) to gain entry
interest plus an adjacent kanamycin-resistance gene. into cells, which is necessary for its replication. Genetic vari-
The inheritance of chromosomal insertion was fol- ations in the CCR5 gene have been identified in the human
lowed by testing progeny for kanamycin resistance. population that confer natural resistance to HIV infection.
Two plants typified the results obtained generally. This includes a CCR5 allele called delta-32 that is missing
When plant 1 was backcrossed with wild-type tobacco, 32 base pairs from the coding region of the gene, causing a
50 percent of the progeny were kanamycin resistant deletion and a frameshift in the encoded CCR5 protein that
and 50 percent were sensitive. When plant 2 was back- blocks its expression on the cell surface. Therefore, to create
crossed with the wild type, 75 percent of the progeny babies that were resistant to HIV infection, the researcher used
were kanamycin resistant and 25 percent were sensi- the CRISPR-Cas9 editing technique to produce a deletion in
tive. What must have been the difference between the the CCR5 gene that was similar to delta-32. This use of the
two transgenic plants? What would you predict about CRISPR-Cas9 technique has prompted a great deal of discus-
the situation regarding the gene of interest? sion about the scientific and ethical implications of making
52. The sequence of the actin gene in the haploid fun- heritable changes to the human genome. Given your new-
gus Neurospora is known from the complete genome found insights into the CRISPR-Cas9 technique and genetic
sequence. If you had a slow-growing mutant that you phenomena, what are your concerns?
I
n December 1965, the king of Sweden presented the was arrested and executed), and André Lwoff offered him
Nobel Prize in Physiology or Medicine to François space at the Pasteur. Monod, in turn, connected Lwoff with
Jacob, Jacques Monod, and André Lwoff of the Pasteur the Resistance.
Institute for their discoveries of how gene expression is reg- After the liberation of Paris, Monod served in the
ulated (Figure 11-1). The prizes were the fruit of an excep- French army and happened on an article by Oswald Avery
tional collaboration among three superb scientists. They and colleagues demonstrating that DNA is the hereditary
were also triumphs over great odds. The chances were slim material in bacteria (see Chapter 7). His interest in genetics
that each of these three men would have lived to see that was rekindled, and he rejoined Lwoff after the war. Mean-
day, let alone earn such honors. while, Jacob’s injuries were too severe for him to pursue a
Twenty-five years earlier, Monod had been a doctoral career in surgery. Inspired by the enormous impact of anti-
student at the Sorbonne in Paris, working on a phenome- biotics introduced late in the war, Jacob eventually decided
non in bacteria called “enzymatic adaptation” that seemed to pursue scientific research. Jacob approached Lwoff sev-
so obscure to some that the director of the zoological lab- eral times for a position in his laboratory but was declined.
oratory where he worked stated, “What Jacques Monod is He made one last try and caught Lwoff in a jovial mood.
doing is of no interest whatever to the Sorbonne.” Jacob The senior scientist told Jacob, “You know, we have just
was a 19-year-old medical student intent on becoming a found the induction of the prophage. Would you be inter-
surgeon. Lwoff was by that time a well-established member ested in working on the phage?” Jacob had no idea what
of the Pasteur Institute in Paris, chief of its department of Lwoff was talking about. He stammered, “That’s just what
microbial physiology. I would like to do.”
Then came World War II. The cast was set. What unfolded in the subsequent
As France was invaded and quickly defeated, Jacob decade was one of the most creative and productive collab-
raced for the coast to join the Free French forces assembling orations in the history of genetics, whose discoveries still
in England. He served as a medic in North Africa and in reverberate throughout biology today.
Normandy until badly wounded. Monod joined the French One of the most important insights arrived not in the
Resistance while continuing his work. After a Gestapo raid laboratory but in a movie theater. Struggling with a lecture
on his Sorbonne laboratory, Monod decided that working that he had to prepare, Jacob opted instead to take his wife,
there was too dangerous (his predecessor in the Resistance Lise, to a Sunday matinee. Bored and daydreaming, Jacob
drew a connection between the work he had been doing
on the induction of prophage and that of Monod on the
induction of enzyme synthesis. Jacob became “involved by
Pioneers of gene regulation
a sudden excitement mixed with a vague pleasure. . . . Both
experiments . . . on the phage . . . and that done with Pardee
and Monod on the lactose system . . . are the same! Same
situation. Same result . . . In both cases, a gene governs the
formation . . . of a repressor blocking the expression of
other genes and so preventing either the synthesis of the
galactosidase or the multiplication of the virus. . . . Where
can the repressor act to stop everything at once? The only
simple answer . . . is on the DNA itself!”1
And so was born the concept of a repressor acting on
DNA to repress the induction of genes. It would take many
years before the hypothesized repressors were isolated
and characterized biochemically. The concepts worked
out by Jacob and Monod and explained in this chapter—
FIGURE 11-1 François Jacob, Jacques Monod, and André Lwoff messenger RNA, promoters, operators, regulatory genes,
were awarded the 1965 Nobel Prize in Physiology or Medicine
for their pioneering work on how gene expression is regulated.
1
[The Pasteur Institute.] F. Jacob, The Statue Within: An Autobiography, 1988.
370
operons, and allosteric proteins—were deduced entirely take compounds from the environment instead. Natural
from genetic evidence, and these concepts shaped the future selection favors efficiency and selects against the waste of
field of molecular genetics. resources and energy. To be economical, bacteria will syn-
Walter Gilbert, who isolated the first repressor and was thesize the enzymes necessary to produce compounds only
later awarded a Nobel Prize in Chemistry for co-inventing when there is no other option—in other words, when com-
a method of sequencing DNA, explained the effect of Jacob pounds are unavailable in their local environment.
and Monod’s work at that time: “Most of the crucial dis- Bacteria have evolved regulatory systems that couple the
coveries in science are of such a simplifying nature that expression of gene products to sensor systems that detect
they are very hard even to conceive without actually having the relevant compound in a bacterium’s local environment.
gone through the experience involved in the discovery. . . . The regulation of enzymes taking part in sugar metabolism
Jacob’s and Monod’s suggestion made things that were provides an example. Sugar molecules can be broken down
utterly dark, very simple.”2 to provide energy, or they can be used as building blocks
The concepts that Jacob and Monod illuminated went for a great range of organic compounds. However, there
far beyond bacterial enzymes and viruses. They understood, are many different types of sugar that bacteria could use,
and were able to articulate with exceptional eloquence, including lactose, glucose, galactose, and xylose. A different
how their discoveries about gene regulation pertained to import protein is required to allow each of these sugars to
the general mysteries of cell differentiation and embry- enter the cell. Further, a different set of enzymes is required
onic development in animals. The two men once quipped, to process each of the sugars. If a cell were to simultane-
“anything found to be true of E. coli must also be true ously synthesize all the enzymes that it might possibly need,
of Elephants.”3 In the next three chapters, we will see to the cell would expend much more energy and materials to
what degree that assertion is true. We’ll start in this chap- produce the enzymes than it could ever derive from break-
ter with bacterial examples that illustrate key themes and ing down prospective carbon sources. The cell has devised
mechanisms in the regulation of gene expression. We will mechanisms to shut down (repress) the transcription of all
largely focus on single regulatory proteins and the genetic genes encoding enzymes that are not needed at a given time
“switches” on which they act. Then, in Chapter 12, we’ll and to turn on (activate) those genes encoding enzymes that
tackle gene regulation in eukaryotic cells, which entails are needed. For example, if only lactose is in the environ-
more complex biochemical and genetic machinery. Finally, ment, the cell will shut down the transcription of the genes
in Chapter 13, we’ll examine the role of gene regulation encoding enzymes needed for the import and metabolism
in the development of multicellular animals. There we will of glucose, galactose, xylose, and other sugars. Conversely,
see how sets of regulatory proteins act on arrays of genetic E. coli will initiate the transcription of the genes encoding
switches to control gene expression in time and space and enzymes needed for the import and metabolism of lactose.
choreograph the building of bodies and body parts. In sum, cells need mechanisms that fulfill two criteria:
1. They must be able to recognize environmental conditions
in which they should activate or repress the transcription
11.1 GENE REGULATION of the relevant genes.
2. They must be able to toggle on or off, like a switch, the
LO 11.1 Illustrate how both positive and negative regulation transcription of each specific gene or group of genes.
control the activity of the lac operon.
Activator- Activator-
RNA
binding binding
No polymerase
site site
3 transcription Transcription
FIGURE 11-2 The binding of regulatory proteins can either activate or block transcription.
One of these DNA–protein interactions determines Both activator and repressor proteins must be able to
where transcription begins. The DNA that participates recognize when environmental conditions are appropriate
in this interaction is a DNA segment called the promoter for their actions and act accordingly. Thus, for activator
(Chapter 8, Section 8.2), and the protein that binds to this or repressor proteins to do their job, each must be able to
site is RNA polymerase. When RNA polymerase binds to exist in two states: one that can bind its DNA targets and
the promoter DNA, transcription can start a few bases another that cannot. The binding state must be appropriate
away from the promoter site. Every gene must have a pro- to the set of physiological conditions present in the cell and
moter or it cannot be transcribed. its environment. For many regulatory proteins, DNA bind-
The other type of DNA–protein interaction deter- ing is effected through the interaction of two different sites
mines whether promoter-driven transcription takes place. in the three-dimensional structure of the protein. One site is
DNA segments near the promoter serve as binding sites for the DNA-binding domain. The other site, the allosteric site,
sequence-specific regulatory proteins called activators and acts as a sensor that sets the DNA-binding domain in one of
repressors. In bacteria, most binding sites for repressors are two modes: functional or nonfunctional. The allosteric site
termed operators. For some genes, an activator protein must interacts with small molecules called allosteric effectors.
bind to its target DNA site as a necessary prerequisite for tran- In lactose metabolism, it is actually an isomer of the sugar
scription to begin. Such instances are sometimes referred to as lactose (called allolactose) that is an allosteric effector: the
positive regulation because the presence of the bound protein sugar binds to a regulatory protein that inhibits the expression
is required for transcription (Figure 11-2). For other genes, a of genes needed for lactose metabolism. In general, an allosteric
repressor protein must be prevented from binding to its target effector binds to the allosteric site of the regulatory protein in
site as a necessary prerequisite for transcription to begin. Such such a way as to change its activity. In this case, allolactose
cases are sometimes termed negative regulation because the changes the shape and structure of the DNA-binding domain of
absence of the bound repressor allows transcription to begin. a regulatory protein. Some activator or repressor proteins must
How do activators and repressors regulate transcrip- bind to their allosteric effectors before they can bind DNA.
tion? Often, a DNA-bound activator protein physically Others can bind DNA only in the absence of their allosteric
helps tether RNA polymerase to its nearby promoter so effectors. Two of these situations are shown in Figure 11-3.
that polymerase may begin transcribing. A DNA-bound
repressor protein typically acts either by physically inter- KEY CONCEPT Allosteric effectors are small molecules that
fering with the binding of RNA polymerase to its pro- bind to activator or repressor proteins and control their ability
moter (blocking transcription initiation) or by impeding to bind to their DNA target sites.
the movement of RNA polymerase along the DNA chain
(blocking transcription). Together, these regulatory proteins
and their binding sites constitute genetic switches that con- A first look at the lac regulatory circuit
trol the efficient changes in gene expression that occur in The pioneering work of François Jacob and Jacques Monod
response to environmental conditions. in the 1950s showed how lactose metabolism is genetically
regulated. Let’s examine the system under two conditions: the
KEY CONCEPT Genetic switches are proteins and DNA presence and the absence of lactose. Figure 11-4 is a simplified
sequences that control gene transcription. Activator or repres-
view of the components of this system. The cast of characters
sor proteins bind to operator sequences in the vicinity of the
for lac operon regulation includes protein-coding genes and
promoter to control its accessibility to RNA polymerase.
sites on the DNA that are targets for DNA-binding proteins.
Allosteric effectors bind to regulatory proteins Lactose is broken down into two sugars and
modified into allolactose
No effector Effector present
Allosteric site (a) HOCH2
HO O OH
Effector
Regulatory OH
b-Galactoside linkage
protein
HOCH2 OH OH
HO O Galactose
DNA-binding site O OH b-Galactosidase
OH
Activator Activator-binding site O OH H2O
HOCH2
Effector OH HOCH2 O OH
Lactose OH
HO
OH
Glucose
(b)
Repressor HOCH2
Operator O O
HO
OH CH2
FIGURE 11-3 Allosteric effectors influence the DNA-binding activi- O OH
OH
ties of activators and repressors. OH
HO
OH
Allolactose
1. The gene for the Lac repressor. A fourth gene (besides the between the promoter and the Z gene near the point at
structural genes Z, Y, and A), the I gene, encodes the Lac which transcription of the multigenic mRNA begins.
repressor protein. It is so named because it can block the
expression of the Z, Y, and A genes. The I gene happens to
The induction of the lac system The P, O, Z, Y, and
map close to the Z, Y, and A genes, but this proximity is
A segments (shown in Figure 11-6) together constitute an
not important to its function because it encodes a diffus-
operon, defined as a segment of DNA that encodes a multi-
ible protein.
genic mRNA as well as an adjacent common promoter and
2. The lac promoter site. The promoter (P) is the site on regulatory region. The lacI gene, encoding the Lac repres-
the DNA to which RNA polymerase binds to initiate sor, is not considered part of the lac operon itself, but the
transcription of the lac structural genes (Z, Y, and A). interaction between the Lac repressor and the lac operator
3. The lac operator site. The operator (O) is the site on site is crucial to proper regulation of the lac operon. The
the DNA to which the Lac repressor binds. It is located Lac repressor has a DNA-binding site that can recognize
I P O Z Y A
Structural genes
RNA polymerase
DNA
mRNA
Polypeptide
Folding
Repressor
protein
I P O Z Y A
Structural genes
RNA polymerase
DNA
mRNA
mRNA
Polypeptide
Folding
Repressor
protein Lactose
mRNA
Medium
FIGURE 11-6 Regulation of the lac operon. The I gene continually makes repressor. (a) In the
A N I M ATED A RT
absence of lactose, the repressor binds to the O (operator) region and blocks transcription. (b) The
binding of lactose changes the shape of the repressor so that the repressor no longer binds to O and Assaying lactose presence or
falls off the DNA. The RNA polymerase is then able to transcribe the Z, Y, and A structural genes, and absence through the Lac repressor
so the three enzymes are produced.
the operator DNA sequence and an allosteric site that binds in which the levels of expression differ in a wild-type
allolactose or analogs of lactose that are useful experimen- genotype, and genetic mutations that perturb the levels of
tally. The repressor will bind tightly only to the O site on expression. In other words, we need a way of describing
the DNA near the genes that it is controlling and not to wild-type gene regulation, and we need mutations that can
other sequences distributed throughout the chromosome. disrupt the wild-type regulatory process. With these ele-
By binding to the operator, the repressor prevents transcrip- ments in hand, we can analyze the expression in mutant
tion by RNA polymerase that has bound to the adjacent genotypes, treating the mutations singly and in combina-
promoter site; the lac operon is switched “off.” tion, to unravel any kind of gene-regulation event. The clas-
When allolactose or its analogs bind to the repressor sical application of this approach was used by Jacob and
protein, the protein undergoes an allosteric transition, Monod, who performed the definitive studies of bacterial
a change in shape. This slight alteration in shape in turn gene regulation.
alters the DNA-binding site so that the repressor no lon- Jacob and Monod used the lactose metabolism system
ger has high affinity for the operator. Thus, in response to of E. coli (see Figure 11-4) to genetically dissect the process
binding allolactose, the repressor falls off the DNA, allow- of enzyme induction—that is, the appearance of a specific
ing RNA polymerase to proceed (transcribe the gene): enzyme only in the presence of its substrates. This phenom-
the lac operon is switched “on.” The repressor’s response enon had been observed in bacteria for many years, but
to allolactose satisfies one requirement for such a control how could a cell possibly “know” precisely which enzymes
system—that the presence of lactose stimulates the synthesis to synthesize? How could a particular substrate induce the
of genes needed for its processing. The relief of repression appearance of a specific enzyme?
for systems such as lac is termed induction. Allolactose and In the lac system, the presence of lactose causes cells
its analogs that allosterically inactivate the repressor, lead- to produce more than 1000 times as much of the enzyme
ing to the expression of the lac genes, are termed inducers. β-galactosidase as they produced when grown in the absence
Let’s summarize how the lac switch works (Figure 11-6). of lactose. What role did lactose play in the induction phe-
In the absence of an inducer (allolactose or an analog), the nomenon? When Monod and co-workers followed the fate
Lac repressor binds to the lac operator site and prevents of radioactively labeled amino acids added to growing cells
transcription of the lac operon by blocking the movement either before or after the addition of an inducer, they found
of RNA polymerase. In this sense, the Lac repressor acts as that induction resulted in the synthesis of new enzyme
a roadblock on the DNA. Consequently, all the structural molecules, as indicated by the presence of the radioactive
genes of the lac operon (the Z, Y, and A genes) are repressed, amino acids in the enzymes. These new molecules could be
and there are very few molecules of β-galactosidase, per- detected as early as three minutes after the addition of an
mease, or transacetylase in the cell. In contrast, when an inducer. Additionally, withdrawal of lactose brought about
inducer is present, it binds to the allosteric site of each an abrupt halt in the synthesis of the new enzyme. There-
Lac repressor subunit, thereby inactivating the site that fore, it became clear that the cell has a rapid and effec-
binds to the operator. The Lac repressor falls off the DNA, tive mechanism for turning gene expression on and off in
allowing the transcription of the structural genes of the lac response to environmental signals.
operon to begin. The enzymes β-galactosidase, permease,
and transacetylase now appear in the cell in a coordinated Genes controlled together
fashion. So, when lactose is present in the environment of
When Jacob and Monod induced β-galactosidase , they
a bacterial cell, the cell produces the enzymes needed to
found that they also induced the enzyme permease, which
metabolize it. But when no lactose is present, resources are
is required to transport lactose into the cell. The analysis
not wasted.
of mutants indicated that each enzyme was encoded by
a different gene. The enzyme transacetylase (with a dis-
pensable and as yet unknown function) also was induced
11.2 DISCOVERY OF THE together with β-galactosidase and permease and was later
LAC SYSTEM: NEGATIVE shown to be encoded by a separate gene. Therefore, Jacob
and Monod could identify three coordinately controlled
REGULATION genes. Recombination mapping showed that the Z, Y, and
A genes were very closely linked on the chromosome (see
LO 11.1 Illustrate how both positive and negative regulation
Section 6.2).
control the activity of the lac operon.
LO 11.2 Infer the components of genetic switches from
experimental data and predict the effect of Genetic evidence for the operator
mutations in the different components on gene and repressor
expression.
Now we come to the heart of Jacob and Monod’s work:
To study gene regulation, ideally we need three ingredi- How did they deduce the mechanisms of gene regulation
ents: a biochemical assay that lets us measure the amount in the lac system? Their strategy was a classic genetic
of mRNA or expressed protein or both, reliable conditions approach: to examine the physiological consequences
Structure of IPTG of mutations. Thus, they haploid strain 1 in this table), even though it is heterozy-
induced mutations in the gous for mutant and wild-type Z alleles. This demonstrates
HOCH2 CH3 structural genes and regu- that the Z+ allele is dominant over its Z− counterpart.
O
HO S C H latory elements of the lac Jacob and Monod first identified two classes of reg-
OH H CH3 operon. As we will see, the ulatory mutations, called OC and I −. These were called
H H properties of mutations in constitutive mutations because they caused the lac operon
H OH
these different components structural genes to be expressed regardless of whether
Isopropyl--D-thiogalactoside of the lac operon are quite inducer was present. Jacob and Monod identified the exis-
(IPTG) different, providing impor- tence of the operator on the basis of their analysis of the OC
tant clues for Jacob and mutations. These mutations make the operator incapable of
FIGURE 11-7 IPTG is an
Monod. binding to repressor; they damage the switch such that the
inducer of the lac operon.
Natural inducers, such operon is always “on” (Table 11-1, strain 3). Importantly,
as allolactose, are not optimal for these experiments because the constitutive effects of OC mutations were restricted
they are broken down by β-galactosidase. The inducer con- solely to those lac structural genes on the same chromo-
centration decreases during the experiment, and so the mea- some as the OC mutation. For this reason, the operator
surements of enzyme induction become quite complicated. mutant was said to be cis-acting, as demonstrated by the
Instead, for such experiments, Jacob and Monod used syn- phenotype of strain 4 in Table 11-1. Here, because the wild-
thetic inducers, such as isopropyl-β-D-thiogalactoside (IPTG; type permease (Y + ) gene is cis to the wild-type operator,
Figure 11-7). IPTG is not hydrolyzed by β-galactosidase, but it permease is expressed only when lactose or an analog is
still induces β-galactosidase enzyme expression. present. In contrast, the wild-type β-galactosidase (Z+ ) gene
Jacob and Monod found that several different classes is cis to the OC mutant operator; hence, β-galactosidase is
of mutations can alter the expression of the structural expressed constitutively. This unusual property of cis action
genes of the lac operon. They were interested in assessing suggested that the operator is a segment of DNA that influ-
the interactions between the new alleles, such as which ences only the expression of the structural genes linked to
alleles exhibited dominance. But to perform such tests, one it (Figure 11-8). The operator thus acts simply as a protein-
needs diploids, and bacteria are haploid. However, Jacob binding site and makes no gene product.
and Monod were able to produce bacteria that are par- Jacob and Monod did comparable genetic tests with
tially diploid by inserting F ′ factors carrying the lac region the I − mutations (Table 11-2). A comparison of the induc-
of the genome. (An F ′ factor is a plasmid that carries one ible wild-type I + (strain 1) with I − strains shows that I −
or more bacterial genes and that can be transferred from mutations are constitutive (strain 2). That is, they cause
one bacteria to another through a process known as con- the structural genes to be expressed at all times. Strain 3
jugation; see Section 6.2.) They could then create strains demonstrates that the inducible phenotype of I + is domi-
that were heterozygous for selected lac mutations, but still nant over the constitutive phenotype of I −. This finding
haploid for the rest of the genome. These partial diploids showed Jacob and Monod that the amount of wild-type
allowed Jacob and Monod to distinguish mutations in the protein encoded by one copy of the gene is sufficient to
regulatory DNA site (the lac operator) from mutations in regulate both copies of the operator in a diploid cell. Most
the regulatory protein (the Lac repressor encoded by the significantly, strain 4 showed them that the I + gene product
I gene). is trans-acting, meaning that the gene product can regu-
We begin by examining mutations that inactivate the late all structural lac operon genes, whether residing on the
structural genes for β-galactosidase and permease (desig- same DNA molecule or on different ones (in cis or in trans,
nated Z− and Y −, respectively). The first thing that we learn respectively). Unlike the operator, the I gene behaves like a
is that Z− and Y − are recessive to their respective wild-type standard protein-coding gene. The protein product of the
alleles (Z+ and Y + ). For example, strain 2 in Table 11-1 can I gene is able to diffuse throughout a cell and act on both
be induced to synthesize β-galactosidase (like the wild-type operators in the partial diploid (Figure 11-9).
TABLE 11-1 Synthesis of β-Galactosidase and Permease in Haploid and Heterozygous Diploid Operator Mutants
β−Galactosidase ( Z ) Permease (Y)
Strain Genotype Noninduced Induced Noninduced Induced Conclusion
+ + +
1 O Z Y − + − + Wild type is inducible
+ + + + − +
2 O Z Y /F ′ O Z Y + + − + Z + is dominant to Z −
3 OC Z + Y + + + + + OC is constitutive
4 O+ Z − Y + /F ′ OC Z + Y − + + − + Operator is cis-acting
Note: Bacteria were grown in glycerol (no glucose present) with and without the inducer IPTG. Expression of maximal enzyme levels is indicated
by +. Absence or very low levels of enzyme activity is indicated by −. All strains are l +.
Repressor
× Repressor
cannot bind
Expression blocked
inducer. However, the lac
genes adjacent to an O+
to altered operator are still subject to
operator repression.
I+ P+ OC Y+
Z+ A NI MAT ED
ART
OC lac operator mutations
Expression even in absence of inducer
TABLE 11-2 Synthesis of β-Galactosidase and Permease in Haploid and Heterozygous Diploid Strains
Carrying l + and l−
β -Galactosidase (Z) Permease (Y)
Strain Genotype Noninduced Induced Noninduced Induced Conclusion
+ + + +
1 I Z Y − + − + I is inducible
− + +
2 I Z Y + + + + I− is constitutive
3 I+ Z − Y + /F ′ I− Z + Y + − + − + I+ is dominant to I−
4 I− Z − Y + /F ′ I+ Z + Y − − + − + I+ is trans-acting
Note: Bacteria were grown in glycerol (no glucose present) with and without the inducer IPTG. Expression of maximal enzyme levels is indicated by +.
Absence or very low levels of enzyme activity is indicated by −. All strains are O+ .
KEY CONCEPT Operator mutations reveal that such a site Genetic evidence for allostery
is cis-acting; that is, it regulates the expression of an adjacent Finally, Jacob and Monod were able to demonstrate allostery
transcription unit on the same DNA molecule. In contrast, through the analysis of another class of repressor mutations.
mutations in the gene encoding a repressor protein reveal that
Recall that the Lac repressor inhibits transcription of the lac
this protein is trans-acting; that is, it can act on any copy of the
operon in the absence of an inducer but permits transcription
target DNA.
when the inducer is present. This regulation is accomplished
P+ O+ A NI MAT ED
I+ Z+ Y+
ART
Expression blocked
TABLE 11-3 Synthesis of β-Galactosidase and Permease by the Wild Type and by Strains Carrying
Different Alleles of the I Gene
β -Galactosidase (Z) Permease (Y)
Strain Genotype Noninduced Induced Noninduced Induced Conclusion
1 I+ Z + Y + − + − + I+ is inducible
2 IS Z + Y + − − − − I S is always repressed
3 I S Z + Y + / F′ I + Z + Y + − − − − I S is dominant to I+
Note: Bacteria were grown in glycerol (no glucose present) with and without the inducer IPTG. Expression of maximal enzyme levels is indicated
by +. Absence or very low levels of enzyme activity is indicated by −.
through a second site on the repressor protein, the allosteric shown in Figure 11-11 as the two highly conserved regions
site, which binds to the inducer. When bound to the inducer, the at −35 and −10. Promoter mutations are cis-acting in that
repressor undergoes a change in overall structure such that its they affect the transcription of all adjacent structural genes
DNA-binding site can no longer function. in the operon. Like operators and other cis-acting elements,
Jacob and Monod isolated another class of repressor promoters are sites on the DNA molecule that are bound by
mutation, called superrepressor (I S ) mutations. I S muta- proteins and themselves produce no protein product.
tions cause repression to persist even in the presence of
an inducer (compare strain 2 in Table 11-3 with the induc- Molecular characterization of the Lac
ible wild-type strain 1). Unlike I − mutations, I S muta- repressor and the lac operator
tions are dominant over I + (see Table 11-3, strain 3). This
Walter Gilbert and Benno Müller-Hill provided a decisive
key observation led Jacob and Monod to speculate that I S
demonstration of the lac system in 1966 by monitoring the
mutations alter the allosteric site so that it can no longer
binding of the radioactively labeled inducer IPTG to purified
bind to an inducer. As a consequence, I S -encoded repres-
repressor protein. They showed that in the test tube, repressor
sor protein continually binds to the operator—preventing
protein binds to DNA containing the operator and comes off
transcription of the lac operon even when the inducer
the DNA in the presence of IPTG. (A more detailed description
is present in the cell. On this basis, we can see why I S is
of how the repressor and other DNA-binding proteins work is
dominant over I +. Mutant I S protein will bind to both copies
given later, at the end of Section 11.6.)
of the operator in the partial diploid cell, even in the presence
Gilbert and his co-workers showed that the repressor can
of an inducer and regardless of the fact that I +-encoded pro-
protect specific bases in the operator from chemical reagents.
tein may be present in the same cell (Figure 11-10).
This information allowed them to isolate the DNA segment
constituting the operator and to determine its sequence. They
Genetic analysis of the lac promoter took operon DNA to which repressor was bound and treated
Mutational analysis also demonstrated that an element essen- it with the enzyme DNase, which breaks up DNA. They were
tial for lac transcription is located between the gene for the able to recover short DNA strands that had been shielded
repressor I and the operator site O. This element, termed the from the enzyme activity by the repressor molecule. These
promoter (P), serves as the initiation site for transcription short strands presumably constituted the operator sequence.
by RNA polymerase (see Chapter 8). There are two binding The base sequence of each strand was determined, and each
regions for RNA polymerase in a typical bacterial promoter, operator mutation was shown to be a change in the sequence
FIGURE 11-11 Specific DNA sequences are important for the effi- are highly conserved in all E. coli promoters, an indication of their role
cient transcription of E. coli genes by RNA polymerase. Only the coding as contact sites on the DNA for RNA polymerase binding. Mutations in
strand (non-template strand) is shown here (see Figure 8-5). Transcrip- these regions have mild (gold) and severe (brown) effects on transcrip-
tion would proceed from left to right (5′ to 3′ ), and the mRNA transcript tion. [Data from J. D. Watson, M. Gilman, J. Witkowski, and M. Zoller,
would be homologous to the sequence shown. The boxed sequences Recombinant DNA, 2nd ed.]
The operator is a specific DNA sequence efficiency of the bacterial cell. Presumably to maximize
energy efficiency, two environmental conditions have
5′ T G G A AT T G T G A G C G G ATA AC A AT T 3′ to be satisfied for the lactose metabolic enzymes to be
3′ A C C T TA A C AC T C G C C TAT T G T TAA 5′ expressed.
One condition is that lactose must be present in the
environment. This condition makes sense because it would
A T G T T A C T be inefficient for the cell to produce the lactose metabolic
O c mutations
T AC A A T G A enzymes if there is no lactose to metabolize. We have
FIGURE 11-12 The DNA base sequence of the lactose operator already seen that the cell is able to respond to the presence
and the base changes associated with eight OC mutations. Regions of lactose through the action of a repressor protein.
of twofold rotational symmetry are indicated by color and by a dot The other condition is that glucose cannot be present
at their axis of symmetry. [Data from W. Gilbert, A. Maxam, and A.
Mirzabekov, in N. O. Kjeldgaard and O. Malløe, eds., Control of Ribosome
in the cell’s environment. Because the cell can capture more
Synthesis. Academic Press, 1976.] energy from the breakdown of glucose than it can from the
breakdown of other sugars, it is more efficient for the cell
(Figure 11-12). These results showed that the operator locus is to metabolize glucose rather than lactose. Thus, mecha-
a specific sequence of 17 to 25 nucleotides situated just before nisms have evolved that prevent the cell from synthesizing
(5′ to) the structural Z gene. They also showed the incredible the enzymes for lactose metabolism when both lactose and
specificity of repressor–operator recognition, which can be glucose are present together. The repression of the transcrip-
disrupted by a single base substitution. When the sequence of tion of lactose-metabolizing genes in the presence of glucose
bases in the lac mRNA (transcribed from the lac operon) was is an example of catabolite repression (glucose is a break-
determined, the first 21 bases on the 5′ initiation end proved to down product, or a catabolite, of lactose). The transcription
be complementary to the operator sequence that Gilbert had of genes encoding proteins necessary for the metabolism of
determined, showing that the operator sequence is transcribed. many different sugars is similarly repressed in the presence
The results of these experiments provided crucial con- of glucose. We will see that catabolite repression works
firmation of the mechanism of repressor action formulated through an activator protein.
by Jacob and Monod.
The basics of lac catabolite repression:
KEY CONCEPT The function of cis-acting elements such as
promoters and operators is determined by their DNA sequence.
choosing the best sugar to metabolize
If both lactose and glucose are present, the synthesis of
β-galactosidase is not induced until all the glucose has
been metabolized. Thus, the cell conserves its energy by
11.3 CATABOLITE REPRESSION metabolizing any existing glucose before going through
OF THE LAC OPERON: the energy-expensive process of creating new machinery
POSITIVE REGULATION to metabolize lactose. There are multiple mechanisms that
bacteria have evolved to ensure the preferential use of a
LO 11.1 Illustrate how both positive and negative regulation carbon source and optimal growth. One mechanism is to
control the activity of the lac operon. exclude lactose from the cell. A second mechanism is to reg-
ulate operon expression via catabolites.
Through a long evolutionary process, the existing lac sys- The results of studies indicate that a breakdown prod-
tem has been selected to operate for the optimal energy uct of glucose prevents activation of the lac operon by
Glucose levels control the lac operon Many DNA binding sites are symmetrical
ATP × No cAMP
(b) CAP-binding site
Low glucose
5′ GTGAGT T A GCT CAC 3′
ATP cAMP 3′ CAC T C A AT CGAGTG 5′
(b) cAMP–CAP complex activates transcription FIGURE 11-14 The DNA base sequences of (a) the lac operator, to
which the Lac repressor binds, and (b) the CAP-binding site, to which
Complex binds to promoter the CAP–cAMP complex binds. Sequences exhibiting twofold rota-
tional symmetry are indicated by the colored boxes and by a dot at
the center point of symmetry. [(a) Data from W. Gilbert, A. Maxam, and
P O A
Z Y A. Mirzabekov, in N. O. Kjeldgaard and O. Malløe, eds., Control of Ribo-
some Synthesis. Academic Press, 1976.]
CAP–cAMP
inhibited, and the cell’s concentration of cAMP increases
correspondingly (Figure 11-13a). A high concentration of
+
cAMP is necessary for activation of the lac operon. Mutants
cAMP
molecules that cannot convert ATP into cAMP cannot be induced to
CAP CAP–cAMP produce β-galactosidase because the concentration of cAMP
is not great enough to activate the lac operon.
FIGURE 11-13 Catabolite control of the lac operon. (a) Only under What is the role of cAMP in lac activation? A study of a
conditions of low glucose is cAMP (cyclic adenosine monophosphate) different set of mutants provided an answer. These mutants
formed from ATP. (b) When cAMP is present, it forms a complex with make cAMP but cannot activate the Lac enzymes because
CAP (catabolite activator protein) that activates transcription by bind-
they lack yet another protein, called catabolite activator
ing to a region within the lac promoter.
protein (CAP), encoded by the crp gene. CAP binds to a
specific DNA sequence of the lac operon (the CAP-binding
lactose—the catabolite repression just mentioned. The glu- site; see Figure 11-14b). The DNA-bound CAP is then able to
cose breakdown product is known to modulate the level interact physically with RNA polymerase and increases that
of an important cellular constituent— cyclic adenosine enzyme’s affinity for the lac promoter. By itself, CAP cannot
monophosphate (cAMP) , which is synthesized from the bind to the CAP-binding site of the lac operon. However, by
major energy source within the cell: adenosine triphos- binding to cAMP, its allosteric effector, CAP is able to bind
phate (ATP). When glucose is present in high concentra- to the CAP-binding site and activate transcription by RNA
tions, it inhibits the conversion of ATP to cAMP, so the cell’s polymerase (Figure 11-13b). By inhibiting CAP when glucose
cAMP concentration is low. As the glucose concentration is available, the catabolite-repression system ensures that the
decreases, the conversion of ATP to cAMP is no longer lac operon will be activated only when glucose is scarce.
P
RNA polymerase
I CAP site interaction site
E. coli
chromosome
Stop codon
Glu
Gln
Ser
Gly
KEY CONCEPT Operons that metabolize a nutrient, such as Binding of CAP bends DNA
the lac operon, often have an added level of control so that the
operon is inactive in the presence of its catabolic breakdown (a)
product (i.e., glucose) even if the nutrient (i.e., lactose) is present.
DNA
O Z
mRNA
fMet
Met
Thr
I P O A
Z Y
I P O A
Z Y
Active
repressor
× Inactive
repressor
R P O A B C R P O A B C
×
No transcription Transcription mRNA
×
(b) Activation
Inactive
factor
A I P X Y Z A I P X Y Z
×
Transcription mRNA No transcription
Active factor
Inducer (activator)
This increases to approximately 3000 molecules of enzyme Map of the ara operon
when lactose is present and glucose is absent. Thus, the cell
conserves its energy and resources by producing the lactose- C O I P B A D
metabolizing enzymes only when they are both needed and
useful.
Inducer–repressor control of the lac operon is an exam- Control Control Structural
gene sites genes
ple of repression, or negative regulation, in which expres-
sion is normally blocked. In contrast, the CAP–cAMP FIGURE 11-19 The B, A, and D genes together with the I and O
system is an example of activation, or positive regula- sites constitute the ara operon. O is araO and I is araI.
tion, because it acts as a signal that activates expression—
in this case, the activating signal is the interaction of the
CAP–cAMP complex with the CAP-binding site on DNA.
binding site for an activator protein. The araC gene, which
Figure 11-18 outlines these two basic types of control systems.
maps nearby, encodes an activator protein. When bound to
KEY CONCEPT Negative regulation promotes gene expres- arabinose, this protein binds to the araI site and activates
sion in the absence of the repressor, and positive regulation transcription of the ara operon, perhaps by helping RNA
promotes gene expression in the presence of an activator. polymerase bind to the promoter. In addition, the same
CAP–cAMP catabolite repression system that prevents lac
operon expression in the presence of glucose also prevents
expression of the ara operon.
11.4 DUAL POSITIVE AND NEGATIVE In the presence of arabinose, both the CAP–cAMP com-
REGULATION: THE ARABINOSE plex and the AraC–arabinose complex must bind to araI in
OPERON order for RNA polymerase to bind to the promoter and
transcribe the ara operon (Figure 11-20a). In the absence
of arabinose, the AraC protein assumes a different confor-
LO 11.3 Illustrate and compare the mechanisms that
coordinate expression of sets of genes in bacteria mation and represses the ara operon by binding both to
and bacteriophage. araI and to a second distant site, araO, thereby forming
a loop (Figure 11-20b) that prevents transcription. Thus,
As with the lac system, the control of transcription in bac- the AraC protein has two conformations, one that acts as
teria is neither purely positive nor purely negative; rather, an activator and another that acts as a repressor. The on/
both positive and negative regulation may govern individ- off switch of the operon is “thrown” by arabinose. The
ual operons. The regulation of the arabinose operon pro- two conformations, dependent on whether the allosteric
vides an example in which a single DNA-binding protein effector arabinose has bound to the protein, differ in their
may act as either a repressor or an activator—a twist on abilities to bind a specific target site in the araO region of
the general theme of transcriptional regulation by DNA- the operon.
binding proteins.
The structural genes araB, araA, and araD encode the KEY CONCEPT Operon transcription is commonly regu-
metabolic enzymes that break down the sugar arabinose. lated by both activation and repression. However, the specific
mechanisms regulating the expression of operons that control
The three genes are transcribed in a unit as a single mRNA.
the metabolism of similar compounds, such as sugars, can be
Figure 11-19 shows a map of the ara operon. Transcription
quite different.
is activated at araI, the initiator region, which contains a
(b) Repression
araO C
AraC protein
araI P B A D
11.5 METABOLIC PATHWAYS AND the growth medium, trp gene expression is high; when lev-
els of tryptophan are high, the trp operon is repressed. One
ADDITIONAL LEVELS OF mechanism for controlling the transcription of the trp operon
REGULATION: ATTENUATION is similar to the mechanism of negative regulation that we
have already seen controls the lac operon: a repressor protein
LO 11.3 Illustrate and compare the mechanisms that binds an operator, preventing the initiation of transcription.
coordinate expression of sets of genes in bacteria This repressor is the Trp repressor, the product of the trpR
and bacteriophage. gene. The Trp repressor binds tryptophan when adequate
levels of the amino acid are present, and only after binding
Coordinate control of genes in bacteria is widespread. Just
tryptophan will the Trp repressor bind to the operator and
as we saw in the preceding sections, there is a need for the
switch off transcription of the operon. This simple mecha-
cell to regulate pathways for the breakdown of specific sug-
nism ensures that the cell does not waste energy producing
ars, depending upon the availability of that sugar. Similarly,
tryptophan when the amino acid is sufficiently abundant in
the pathways that synthesize essential molecules, like amino
the environment. E. coli strains with mutations in trpR con-
acids, must be regulated so that the enzymes needed for their
tinue to express the trp mRNA and thus continue to produce
synthesis are produced by the bacteria only when amino
tryptophan when the amino acid is abundant.
acids are not available from the environment. In pathways
In studying these trpR mutant strains, Charles Yanof-
that synthesize essential molecules, the genes that encode the
sky discovered that, when tryptophan was removed from
enzymes are also organized into operons, complete with mul-
the medium, the production of trp mRNA further increased
tigenic mRNAs. Furthermore, in cases for which the sequence
several-fold. This finding was evidence that, in addition to
of catalytic activity is known, there is a remarkable congru-
the Trp repressor, a second control mechanism existed to
ence between the order of operon genes on the chromosome
negatively regulate transcription. This mechanism is called
and the order in which their products act in the metabolic
attenuation because mRNA production is normally atten-
pathway. This congruence is strikingly illustrated by the orga-
uated, meaning “decreased,” when tryptophan is plentiful.
nization of the tryptophan operon in E. coli (Figure 11-21).
Unlike the other bacterial control mechanisms described thus
The tryptophan operon contains five genes (trpE, trpD, trpC,
far, attenuation acts at a step after transcription initiation.
trpB, trpA) that encode enzymes that contribute to the syn-
The mechanisms governing attenuation were discovered by
thesis of the amino acid tryptophan.
identifying mutations that reduced or abolished attenuation.
Strains with these mutations produce trp mRNA at maximal
KEY CONCEPT In bacteria, genes that encode enzymes that
are in the same metabolic pathways are generally organized levels even in the presence of tryptophan. Yanofsky mapped the
into operons. mutations to a region between the trp operator and the trpE
gene; this region, termed the leader sequence, is at the 5′ end
of the trp operon mRNA before the first codon of the trpE
There are two mechanisms for regulating transcription of gene (Figure 11-22). The trp leader sequence is unusually long
the tryptophan operon and some other operons functioning in for a bacterial mRNA, 160 bases, and detailed analyses have
amino acid biosynthesis. One provides global control of operon revealed how a part of this sequence works as an attenuator
mRNA expression, and the other provides fine-tuned control. that governs the further transcription of trp mRNA.
The level of trp operon gene expression is governed by The key observations are that, in the absence of the
the level of tryptophan. When tryptophan is absent from TrpR repressor protein, the presence of tryptophan halts
Gene order in the trp operon corresponds to reaction order in the biosynthetic pathway
O trpE trpD trpC trpB trpA
H H H H
O O NH2
COOH COOH COOH COOH O O H
O C C CH2O P C C COOH
HO C C C CH2O P H H H H
HO L-Glutamine
NH2 PRPP N H H N N
H CH2O P N CH H
O H H
O O L-Tryptophan
H2C C COOH H H
Anthranilic CDRP Indole-3-glycerol
Phosphoribosyl phosphate N
Chorismic acid acid H L-Serine
anthranilic acid
Indole
FIGURE 11-21 The chromosomal order of genes in the trp operon of E. coli and the sequence of reac-
tions catalyzed by the enzyme products of the trp structural genes. The products of genes trpD and trpE
form a complex that catalyzes specific steps, as do the products of genes trpB and trpA. Abbreviations:
PRPP, phosphoribosylpyrophosphate; CDRP, 1-(o-carboxyphenylamino)-1-deoxyribulose 5-phosphate.
The trp mRNA leader sequence contains an attenuator region and two tryptophan codons
Trp codons
70 50
G UACCACUUAUGUGACGGGCAAAGUCCUUCACGCGGUGGUUGGAAAGUCAUGCUUUUAACGAAAGUAACAGCUAUCGGAAAAAUGCACUUGAAppp 5′
GC
U
A
A Stop codon Start codon
A
G 110 130 140 160
C
AA
UCAGAUACCCAGCCCGCCUAAUGAGCGGGCUUUUUUUUGAACAAAAUUAGAGAAUAACAAUGCAAACACAAAAACCGACUCUCGAACUGOU…
Met-Gln-Thr-Gln-Lys-Pro-Thr-Leu-Glu-Leu-Leu
Attenuator region TrpE polypeptide
FIGURE 11-22 In the trp mRNA leader sequence, the attenuator region precedes the trpE coding
sequence. Farther upstream, at bases 54 through 59, are the two tryptophan codons (shown in red)
found in the 14-amino-acid leader peptide.
transcription after the first 140 bases or so, whereas, in the it is encoded by a single codon. This pair of tryptophan
absence of tryptophan, transcription of the operon contin- codons is therefore an unusual feature. Second, the trp
ues. The mechanism for terminating or continuing tran- mRNA leader sequence consists of four segments that form
scription consists of two key elements. First, the trp mRNA stem-and-loop RNA structures that are able to alternate
leader sequence encodes a short, 14-amino-acid peptide between two conformations. One of these conformations
that includes two adjacent tryptophan codons. Tryptophan favors the termination of transcription, while the other
is one of the least abundant amino acids in proteins, and favors the continuation of transcription (Figure 11-23).
FIGURE 11-23 (a) Proposed secondary structures in the conforma- pair to form a stem-loop that causes RNA polymerase to terminate
tion of trp leader mRNA that favors termination of transcription. Four transcription. (c) In contrast, when tryptophan is scarce, the ribosome
regions can base pair to form three stem-and-loop structures, but is stalled at the codons of segment 1. Thus, segment 2 can interact
only two regions base pair with one another at a given time. (b) When with segment 3, and so segments 3 and 4 cannot pair. Consequently,
tryptophan is abundant, segment 1 of the trp mRNA is translated. transcription continues. [Data from D. L. Oxender, G. Zurawski, and
Segment 2 enters the ribosome, enabling segments 3 and 4 to base C. Yanofsky, Proc. Natl. Acad. Sci. USA 76, 1979, 5524.]
The regulatory logic of the operon pivots on the abun- 11.6 BACTERIOPHAGE LIFE
dance of tryptophan. When tryptophan is abundant, there
is a sufficient supply of aminoacyl-tRNATrp to allow trans-
CYCLES: MORE REGULATORS,
lation of the 14-amino-acid leader peptide. Recall that COMPLEX OPERONS
transcription and translation in bacteria are coupled; so
ribosomes can engage mRNA transcripts and initiate trans- LO 11.3 Illustrate and compare the mechanisms that
lation before transcription is complete. The engagement of coordinate expression of sets of genes in bacteria
the ribosome alters trp mRNA conformation to the form and bacteriophage.
that favors termination of transcription. Because segment LO 11.4 Explain the roles of sequence-specific DNA-
binding proteins and DNA regulatory sequences
1 of the trp leader mRNA is translated in the presence of
in coordinating the expression of sets of genes in
tryptophan, segment 2 of the leader mRNA will enter the bacteria and bacteriophage.
ribosome. This allows base pairing between the attenua-
tor region found in segments 3 and 4 of the leader mRNA, In that Paris movie theater, François Jacob had a flash of
which leads to the termination of transcription by RNA insight that the phenomenon of prophage induction might
polymerase (Figure 11-23b). However, when tryptophan is be closely analogous to the induction of β-galactosidase syn-
scarce, the ribosome is stalled at the tryptophan codons in thesis. He was right. Here, we are going to see how the life
segment 1, such that segments 2 and 3 base pair, and tran- cycle of the bacteriophage λ is regulated. Although its regu-
scription is able to continue (Figure 11-23c). This mecha- lation is more complex than that of individual operons, it is
nism is exquisitely sensitive to the levels of tryptophan in controlled by now-familiar modes of gene regulation.
the environment because the number of transcripts pro-
duced will be directly related to the number of stalled ribo-
somes, which in turn is related to the amount of tryptophan Regulation of the bacteriophage λ life cycle
in the cell. Thus, attenuation provides a way for the bacte- Bacteriophage λ is a so-called temperate phage that has
ria to rapidly fine-tune the synthesis of tryptophan depend- two alternative life cycles ( Figure 11-25). When a normal
ing upon environmental conditions. bacterium is infected by a wild-type λ phage, two possi-
Other operons for enzymes in biosynthetic pathways ble outcomes may follow: (1) the phage may replicate and
have similar attenuation controls. One signature of amino eventually lyse the cell (the lytic cycle) or (2) the phage
acid biosynthesis operons is the presence of multiple codons genome may be integrated into the bacterial chromosome
for the amino acid being synthesized in a separate peptide as an inert prophage (the lysogenic cycle). In the lytic state,
encoded by the 5′ leader sequence. For instance, the phe most of the phage’s 71 genes are expressed at some point,
operon has seven phenylalanine codons in a leader peptide whereas in the lysogenic state, most genes are inactive.
and the his operon has seven tandem histidine codons in its What decides which of these two pathways is taken?
leader peptide (Figure 11-24). The physiological control of the decision between the lytic
or lysogenic pathway depends on the resources available in
KEY CONCEPT A second level of regulation in operons that the host bacterium. If resources are abundant, the lytic cycle
control amino acid biosynthesis is attenuation of transcription is preferred because then there are sufficient nutrients to
mediated by the abundance of the amino acid and translation
make many copies of the virus. If resources are limited, the
of a leader peptide.
lysogenic pathway is taken. The virus then remains present
(b) phe operon Met - Lys - His - Ile - Pro - Phe - Phe - Phe - Ala - Phe - Phe - Phe - Thr - Phe - Pro - Stop
5′ AUG - AAA - CAC - AUA - CCG - UUU - UUU - UUC - GCA - UUC - UUU - UUU - ACC - UUC - CCC - UGA 3′
(c) his operon Met - Thr - Arg - Val - Gln - Phe - Lys - His - His - His - His - His - His - His - Pro - Asp
5′ AUG - ACA - CGC - GUU - CAA - UUU - AAA - CAC - CAC - CAU - CAU - CAC - CAU - CAU - CCU - GAC 3′
FIGURE 11-24 (a) The translated part of the trp leader region contains two consecutive tryptophan codons,
(b) the phe leader sequence contains seven phenylalanine codons, and (c) the his leader sequence contains
seven consecutive histidine codons.
Chromosome + λ DNA
(in head)
Infection
1. Lytic 2. Lysogenic
cycle cycle
λ prophage
Cell lysis
Cell lysis
Lysogenic growth
FIGURE 11-25 Whether bacteriophage λ enters the lytic cycle immediately or enters the lysogenic
cycle depends on the availability of resources. The lysogenic virus inserts its genome into the bacte-
rial chromosome, where it remains quiescent until conditions are favorable.
as a prophage until conditions improve. The inert prophage recognized by the appearance of infected plaques on a lawn
can be induced by ultraviolet light to enter the lytic cycle— of bacteria. When wild-type phage particles are placed on
the phenomenon studied by Jacob. The lytic and lysogenic a lawn of sensitive bacteria, clearings (called “plaques”)
states are characterized by very distinct programs of gene appear where bacteria are infected and lysed, but these
expression that must be regulated. Which alternative state plaques are cloudy because bacteria that are lysogenized
is selected is determined by a complex genetic switch com- grow within them (Figure 11-26). Mutant phages that are
prising several DNA-binding regulatory proteins and a set unable to lysogenize cells form clear plaques.
of operator sites. Such clear mutants (designated by c) turn out to be
Just as they were for the lac and other regulatory sys- analogous to the I and O mutants of the lac system. These
tems, genetic analyses of mutants were sources of crucial mutants were often isolated as temperature-sensitive
insights into the components and logic of the λ genetic mutants that had clear phenotypes at higher temperatures
switch. Jacob used simple phenotypic screens to iso- but wild-type phenotypes at lower temperatures. Three
late mutants that were defective in either the lytic or classes of mutants led to the identification of the key reg-
the lysogenic pathway. Mutants of each type could be ulatory features of phage λ . In the first class, mutants
Clear and cloudy bacteriophage plaques on a lawn TABLE 11-4 Major Regulators of Bacteriophage λ
of E. coli host bacteria Life Cycle
Gene Protein Promotes
cI λ repressor lysogenic pathway
cro Cro repressor lytic pathway
N positive regulator cII, cIII expression
Cloudy plaque
cII activator cI expression
cIII protease inhibitor cII activity
Clear plaque
Excisionase xis PL
Integrase int
att
Head
genes
Tail genes
FIGURE 11-27 Map of phage λ in the circular form. The genes for recombination, integration and
excision, replication, head and tail assembly, and cell lysis are clustered together and coordinately
regulated. Transcription of the right side of the genome begins at PR , and that of the leftward genes
begins at PL . Key regulatory interactions governing the lysogenic-versus-lytic decision take place at
operators between the cro and the cI genes.
However, if resources are limited, cII is more active and 11-28). Recall that the cI gene encodes the λ repressor. The
more λ repressor is produced. In this case, the genes tran- three operator sites are similar but not identical in sequence,
scribed from PL and PR are repressed by the λ repressor and although Cro and λ repressor can each bind to any
and the lysogenic cycle is entered. The cII protein is also one of the operators, they do so with different affinities:
responsible for activating the transcription of int, a gene λ repressor binds to OR1 with the highest affinity, whereas
that encodes an additional protein required for lysogeny— Cro binds to OR3 with the highest affinity. The λ repressor’s
an integrase required for the λ genome to integrate into the occupation of OR1 blocks transcription from PR and thus
host chromosome. The cIII protein shields cII from degra- blocks the transcription of genes for the lytic cycle. Cro’s
dation; so it, too, contributes to the lysogenic decision. occupation of OR3 blocks transcription from PRM and thus
blocks maintenance of cI transcription. Hence, no λ repres-
sor is produced, and transcription of genes for the lytic
Molecular anatomy of the genetic switch cycle can continue. The occupation of the operator sites
To see how the decision is executed at the molecular level, therefore determines the lytic-versus-lysogenic patterns of λ
let’s turn to the activities of λ repressor and Cro. The OR gene expression (Figure 11-29).
operator lies between the two genes encoding these proteins After a lysogen has been established, it is generally
and contains three sites, OR1, OR2 , and OR3, that overlap stable. But the lysogen can be induced to enter the lytic
two opposing promoters: PR, which promotes transcription cycle by various environmental changes. Ultraviolet light
of lytic genes, and PRM (for repressor maintenance), which induces the expression of host genes. One of the host genes
directs transcription of the cI gene (see Figures 11-27 and encodes a protein, RecA, that stimulates cleavage of the
Infection
(a) cIII N
PL PRM PRE
int cIII N OL cI OR cro cII Lytic genes
PR
cro cII
PL PRM PRE
int cIII N OL cI OR cro cII cII Lytic genes
PR
cIII cII
No degradation of cII
(c)
λ
PL PRM PRE
int cIII N OλL cI λ cro cII Lytic genes
PR
PL PRM PRE
int cIII N OL cI cro cro cII Lytic genes
PR
Lytic genes
Degraded cII
cro
FIGURE 11-28 On infection (a), host RNA polymerase initiates tran- remains active, cI transcription proceeds at a high level, and the Int
scription at PL and PR , expressing the N and cro genes, respectively. protein integrates the phage chromosome. Eventually (c), the cI pro-
Antiterminator protein N enables transcription of the cIII gene and tein (λ repressor) shuts off all genes except itself. The phage will then
recombination genes (see Figure 11-27, left), and the cII gene and remain in the lysogenic state. However, if resources and proteases are
other genes. Next, (b) the cII protein, protected by the cIII protein, abundant (d), the cII protein is degraded, Cro represses transcription of
turns on cI by activating transcription at PRE and also activates the cI from PRM and activates transcription of Cro and lytic genes from PL
transcription of int. If resources and proteases are not abundant, cII and PR , and the lytic cycle continues.
λ repressor, thus crippling maintenance of lysogeny and lysogen induction makes sense in that this type of radi-
resulting in lytic growth. Prophage induction, just as Jacob ation damages host DNA and stresses the bacteria; the
and Monod surmised, requires the release of a repressor phage replicates and leaves the damaged, stressed cell for
from DNA. The physiological role of ultraviolet light in another host.
Lysogen λ repressor
RNA polymerase
cl OR3 OR2 OR1 cro
PRM On PR Off
Lytic growth
Cro
PRM Off PR On
FIGURE 11-29 Lysogeny is promoted by λ repressor binding to OR1 and OR2, which prevents
transcription from PR . On induction or in the lytic cycle, the binding of Cro to OR3 prevents transcription of
the cI gene from PRM. [Data from M. Ptashne and A. Gann, Genes and Signals, p. 30, Fig. 1-13.]
KEY CONCEPT The phage λ genetic switch illustrates that Crystallographic analysis has identified a common
the regulatory logic that underlies the control of gene expres- structural feature of the DNA-binding domains of λ and
sion in response to physiological signals is conserved. Just as Cro. Both proteins make contact with DNA through a
in the lac, ara, trp, and other systems, the alternative states of helix-turn-helix domain that consists of two helices joined
gene expression in λ phage are determined by the interaction by a short flexible linker region (Figure 11-30). One helix,
of a few key DNA-binding regulatory proteins with control sites the recognition helix, fits into the major groove of DNA. In
on the DNA. The order and orientation of these genetic ele-
that position, amino acids on the helix’s outer face are able
ments in the genome is important to the switch function in all
to interact with chemical groups on the DNA bases. The
of these systems.
specific amino acids in the recognition helix determine the
affinity of a protein for a specific DNA sequence.
T A C C T C T G T A T C C C T T
OR1 OR3
A T G G A G A C C A T A G G G A A C
FIGURE 11-31 Interactions between amino acids and bases determine the specificity and
affinity of DNA-binding proteins. The amino acid sequences of the recognition helices of the λ
repressor and Cro proteins are shown. Interactions between the glutamine (Gln), serine (Ser),
and alanine (Ala) residues of the λ repressor and bases in the OR1 operator determine the
strength of binding. Similarly, interactions between the glutamine, serine, asparagine (Asn),
and lysine (Lys) residues of the Cro protein mediate binding to the OR3 operator. Each DNA
sequence shown is that bound by an individual monomer of the respective repressor; it is half
of the operator site occupied by the repressor dimer. [Data from M. Ptashne, A Genetic Switch:
Phage l and Higher Organisms, 2nd ed.]
The recognition helices of the λ repressor and Cro have responses to changes in the environment require the coor-
similar structures and some identical amino acid residues. dinated expression of large sets of unlinked genes located
Differences between the helices in key amino acid residues throughout the genome to bring about dramatic physio-
determine their DNA-binding properties. For example, in logical and even morphological changes. Analyses of these
the λ repressor and Cro proteins, glutamine and serine side processes have revealed another twist in bacterial gene reg-
chains contact the same bases, but an alanine residue in the ulation: the control of large numbers of genes by alternative
λ repressor and lysine and asparagine residues in the Cro sigma (σ) factors of RNA polymerase. One such example,
protein impart different binding affinities for sequences in the process of sporulation in Bacillus subtilis, has been ana-
OR1 and OR3 (Figure 11-31). lyzed in great detail in the past few decades. Under stress,
The Lac and TrpR repressors, as well as the AraC acti- the bacterium forms spores that are remarkably resistant to
vator and many other proteins, also bind to DNA through heat and desiccation.
helix-turn-helix motifs of differing specificities, depending Early in the process of sporulation, the bacterium
on the primary amino acid sequences of their recognition divides asymmetrically, generating two components of
helices. In general, other domains of these proteins, such unequal size that have very different fates. The smaller
as those that bind their respective allosteric effectors, are compartment, the forespore, develops into the spore. The
dissimilar. larger compartment, the mother cell, nurtures the develop-
ing spore and lyses when spore morphogenesis is complete
KEY CONCEPT The biological specificity of gene regulation to liberate the spore (Figure 11-32a). Genetic dissection of
is due to the chemical specificity of amino acid–base interac- this process has entailed the isolation of many mutants
tions between individual regulatory proteins and discrete DNA that cannot sporulate. Detailed investigations have led to
sequences. the characterization of several key regulatory proteins that
directly regulate programs of gene expression that are spe-
cific to either the forespore or the mother cell. Four of these
proteins are alternative σ factors.
11.7 ALTERNATIVE SIGMA FACTORS Recall from Chapter 8 that transcription initiation in
REGULATE LARGE SETS OF bacteria includes the binding of the σ subunit of RNA poly-
merase to the −35 and −11 regions of gene promoters. The
GENES σ factor disassociates from the complex when transcription
begins and is recycled. In B. subtilis, two σ factors, σ A and
LO 11.4 Explain the roles of sequence-specific DNA-
binding proteins and DNA regulatory sequences
σ H , are active in vegetative cells. During sporulation, a dif-
in coordinating the expression of sets of genes in ferent σ factor, σ F, becomes active in the forespore and acti-
bacteria and bacteriophage. vates a group of more than 40 genes. One gene activated by
σ F is a secreted protein that in turn triggers the proteolytic
Thus far, we have seen how single switches can control the processing of the inactive precursor pro-σ E , a distinct σ fac-
expression of single operons or two operons containing tor in the mother cell. The σ E factor is required to activate
as many as a couple of dozen genes. Some physiological sets of genes in the mother cell. Two additional σ factors, σ K
(a)
Forespore
σF
σG
σAσH σE σK Spore
Vegetative Mother
cell cell
(b)
σE–regulated promoters
–35 –10
ybaN TCG G TTATAT T CA AT T G T – C C AT GCT C ATAAG AT …
ydcC G TCT GCATAT T A GGGAAA–C C C CACT C AT A TAT T…
ydcA TA C G TACTATT TAAAT G G – T T T T TCTC ATAAAC G …
σF–regulated promoters
yrrR AT C T G TTTA GCAGCGAAACACCTCGTCCACAATG…
ytfT CCGGG T T TAT T T T T T T –AGGAAT TGGCGATA AT G…
yuiC T T T T GAATA AT GCT C T CTCCACTTGGGAACAATG…
and σG, are subsequently activated in the mother cell and regulate more than 250 genes, and σ F binds to at least 36
forespore, respectively (Figure 11-32a). The expression of promoters to regulate 48 genes.
distinct σ factors allows for the coordinated transcription
of different sets of genes, or regulons, by a single RNA KEY CONCEPT Sequential expression of alternative σ fac-
polymerase. tors that recognize alternative promoter sequences enables the
New approaches for characterizing the expression of coordinated expression of large numbers of independent oper-
all genes in a genome (see Section 14.7) have made it pos- ons and unlinked genes during the developmental program of
sible to monitor the transcription of each B. subtilis gene sporulation.
during vegetative growth and spore formation and in dif-
ferent compartments of the spore. Several hundred genes Alternative σ factors also play important roles in the
have been identified in this fashion that are transcription- virulence of human pathogens. For example, bacteria of
ally activated or repressed during spore formation. the genus Clostridium produce potent toxins that are
How are the different sets of genes controlled by each responsible for severe diseases such as botulism, tetanus,
σ factor? Each σ factor has different sequence-s pecific and gangrene. Key toxin genes of C. botulinum, C. tetani,
DNA-binding properties. The operons or individual and C. perfringens have recently been discovered to be
genes regulated by particular σ factors have character- controlled by related, alternative σ factors that recognize
istic sequences in the −35 and −11 regions of their pro- similar sequences in the −35 and −10 regions of the toxin
moters that are bound by one σ factor and not others genes. Understanding the mechanisms of toxin-gene regu-
(Figure 11-32b). For example, σ E binds to at least 121 lation may lead to new means of disease prevention and
promoters, within 34 operons and 87 individual genes, to therapy.
SUMMARY
Gene regulation is often mediated by proteins that react to Negative regulatory control is exemplified by the lac system.
physiological signals from within and around the cell. The Negative regulation is one very straightforward way for the
proteins respond by raising or lowering the transcription lac system to shut down metabolic genes in the absence of
rates of specific genes. The logic of this regulation is straight- appropriate sugars in the environment. In positive regulatory
forward. For regulation to operate appropriately, the regula- control, protein factors are required to activate transcription.
tory proteins have built-in sensors that continually monitor Positive regulatory control is exemplified by repression of the
cellular conditions. The activities of these proteins would lac system in the presence of its catabolite breakdown prod-
then depend on the right set of environmental conditions. uct, glucose. By contrast, repression of operons that synthe-
In bacteria and their viruses, the control of several struc- size amino acids is often controlled by attenuation.
tural genes may be coordinated by clustering the genes Many regulatory proteins are members of families of
together into operons on the chromosome so that they are proteins that have very similar DNA-binding motifs, such
transcribed into multigenic mRNAs. Coordinated control as the helix-turn-helix domain. Other parts of the proteins,
simplifies the task for bacteria because one cluster of regula- such as their protein–protein interaction domains, tend to
tory sites per operon is sufficient to regulate the expression be less similar. The specificity of gene regulation depends
of all the operon’s genes. Alternatively, coordinate control on chemical interactions between the side chains of amino
can also be achieved through discrete σ factors that regulate acids and chemical groups on DNA bases.
dozens of independent promoters simultaneously. The mechanisms of the regulatory control of gene
In negative regulatory control, a repressor protein expression can be inferred from analyzing the physiological
blocks transcription by binding to DNA at the operator site. effects of genetic mutations.
KEY TERMS
activator (p. 372) cis-acting (p. 376) lysogenic cycle (p. 386)
allosteric effector (p. 372) constitutive mutation (p. 376) lytic cycle (p. 386)
allosteric site (p. 372) coordinately controlled genes (p. 373) negative regulation (p. 372)
allosteric transition (p. 375) cyclic adenosine monophosphate operator (p. 372)
antiterminator (p. 388) (cAMP) (p. 380) operon (p. 374)
attenuation (p. 384) DNA-binding domain (p. 372) partial diploid (p. 376)
attenuator (p. 384) genetic switch (p. 372) positive regulation (p. 372)
catabolite (p. 379) inducer (p. 375) promoter (p. 372)
catabolite activator protein (CAP) induction (p. 375) regulon (p. 393)
(p. 380) initiator (p. 383) repressor (p. 372)
catabolite repression (p. 379) leader sequence (p. 384) trans-acting (p. 376)
SOLVED PROBLEMS
This set of four solved problems, which are similar to Problem Z gene Y gene
15 in the Basic Problems at the end of this chapter, is designed
Genotype No inducer Inducer No inducer Inducer
to test understanding of the operon model. Here, we are given
several diploids and are asked to determine whether Z and
SOLVED PROBLEM 1
Y gene products are made in the presence or absence of an
inducer. Use a table similar to the one in Problem 15 as a basis I − P− OC Z+ Y +
for your answers, except that the column headings will be as I + P+ O+ Z− Y −
follows:
I+ P+ O+ Z– Y–
I+ P+ OC Z– Y+
× × Repressor cannot
×
bind to O C operator ×
No No Active enzyme
RNA polymerase active active No in presence
can bind enzyme enzyme active and absence
enzyme of an inducer
I+ P– O+ Z+ Y–
−
The first chromosome is P , and so transcription is blocked
and no Lac enzyme can be synthesized from it. The second × No transcription:
chromosome (P+ ) can be transcribed, and thus transcription RNA polymerase
cannot bind
is repressible (O+ ). However, the structural genes linked to
the good promoter are defective; thus, no active Z product
or Y product can be generated. The symbols to add to your SOLVED PROBLEM 4
table are “−, −, −, −.”
I S P+ O+ Z+ Y −
I − P+ OC Z− Y +
SOLVED PROBLEM 2
SOLUTION
I + P− O+ Z+ Y + In the presence of an I S repressor, all wild-type operators
I − P+ O+ Z+ Y − are shut off, both with and without an inducer. Therefore,
SOLUTION the first chromosome is unable to produce any enzyme.
The first chromosome is P− , and so no enzyme can be syn- However, the second chromosome has an altered (OC ) oper-
thesized from it. The second chromosome is O+ , and so ator and can produce enzyme in both the absence and the
transcription is repressed by the repressor supplied from the presence of an inducer. Only the Y gene is wild type on the
first chromosome, which can act in trans through the cyto- OC chromosome, and so only permease is produced consti-
plasm. However, only the Z gene from this chromosome is tutively. The entries in the table should be “−, −, +, +.”
intact. Therefore, in the absence of an inducer, no enzyme is
made; in the presence of an inducer, only the Z gene prod-
uct, β-galactosidase, is generated. The symbols to add to the
table are “−, +, −, −.” No transcription
IS P+ O+ Z+ Y–
RNA polymerase
IS repressor binds
cannot bind:
to operator even in
× no transcription
× presence of IPTG
No active Repressor cannot
P–
×
repressor bind to O C operator
I+ O+ Z+ Y+ P+
I– OC Z– Y+
×
Repression in No Active enzyme
absence of IPTG active in presence and
I– P+ O+
enzyme absence of IPTG
Z+ Y–
×
Active No
Induction enzyme in active
in presence presence enzyme
of IPTG of IPTG
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a Problem Solving Video. Problems with the icon have an Unpacking the Problem exercise.
www
www
WORKING WITH THE FIGURES 12. Explain why I − alleles in the lac system are normally
(The first 10 questions require inspection of text figures.) recessive to I + alleles and why I + alleles are recessive
to I S alleles.
1. Based on Figure 11-6, why does the binding of the
repressor protein to the operator sequence in the 13. What do we mean when we say that OC mutations in
absence of lactose prevent expression of the structural the lac system are cis-acting?
genes? Why does the absence of binding of the repres- 14. The symbols a, b, and c in the table below represent the
sor protein to the operator sequence in the presence of E. coli lac system genes for the repressor (I), the oper-
lactose allow expression of the structural genes? ator (O) region, and the β-galactosidase (Z), although
2. Compare the structure of IPTG shown in Figure 11-7 not necessarily in that order. Furthermore, the order in
with the structure of galactose shown in Figure 11-5. which the symbols are written in the genotypes is not
Why is IPTG bound by the Lac repressor but not bro- necessarily the actual sequence in the lac operon. www
a. Galactose c. Allolactose
b . Glucose d. Isothiocyanate
β-Galactosidase Permease
Genotype No lactose Lactose No lactose Lactose
I + +
P O Z Y /I + + + + +
P O Z Y + + + − + − +
− + C + − + + + − +
a. I P O Z Y /I P O Z Y
b. I + P− OC Z− Y + /I − P+ OC Z+ Y −
c. I S P+ O+ Z+ Y − /I + P+ O+ Z− Y +
d. I S P+ O+ Z+ Y + /I − P+ O+ Z+ Y +
e. I − P+ OC Z+ Y − /I − P+ O+ Z− Y +
f. I − P− O+ Z+ Y + /I − P+ OC Z+ Y −
g. I + P+ O+ Z− Y + / I − P+ O+ Z+ Y −
16. Explain why it makes sense for the cell to synthesize 28. Predict the effect of a mutation that eliminates the
β-galactosidase only when levels of lactose are high DNA-binding activity of the σ E protein on spore for-
and levels of glucose are low. mation in Bacillus subtilis.
17. Explain the fundamental differences between negative
regulation and positive regulation of transcription in CHALLENGING PROBLEMS
bacteria. 29. An interesting mutation in lacI results in repressors
18. Which molecule regulates both the lac operon and the with 110-fold increased binding to both operator
ara operon? and nonoperator DNA. These repressors display a
“reverse” induction curve, allowing β-galactosidase
19. Compare the mechanisms of negative and positive
synthesis in the absence of an inducer (IPTG) but
regulation in the lac operon with those in the ara
partly repressing β-galactosidase expression in the
operon.
presence of IPTG. How can you explain this? (Note
20. Mutants that are lacY − retain the capacity to synthe- that, when IPTG binds a repressor, it does not com-
size β-galactosidase. However, even though the lacI pletely destroy operator affinity; rather, it reduces
gene is still intact, β-galactosidase can no longer be affinity 110-fold. Additionally, as cells divide and new
induced by adding lactose to the medium. Explain. operators are generated by the synthesis of daughter
21. What is the function of the two tryptophan codons in strands, the repressor must find the new operators by
the 14-amino-acid leader peptide in the regulation of searching along the DNA, rapidly binding to nonoper-
trp operon? ator sequences and dissociating from them.)
22. Could the attenuation mechanism found in the trp 30. Certain lacI mutations eliminate operator binding by
operon regulate gene expression in eukaryotic cells? the Lac repressor but do not affect the aggregation
of subunits to make a tetramer, the active form of the
23. What are the similarities between the mechanisms con-
repressor. These mutations are partly dominant over
trolling the lac operon and those controlling bacterio-
wild type. Can you explain the partly dominant I −
phage λ genetic switches?
phenotype of the I − /I + heterodiploids?
24. Compare the arrangement of cis-acting sites in the
31. You are examining the regulation of the lactose operon
control regions of the lac operon and bacteriophage λ .
in the bacterium Escherichia coli. You isolate seven
25. Which regulatory protein induces the lytic phase genes new independent mutant strains that lack the products
of the bacteriophage λ life cycle? of all three structural genes. You suspect that some
a. cI of these mutations are lacI S mutations and that other
mutations are alterations that prevent the binding of
b. Cro
RNA polymerase to the promoter region. Using what-
c. Int ever haploid and partial diploid genotypes that you
d. cIII think are necessary, describe a set of genotypes that
will permit you to distinguish between the lacI and
26. What protein in bacteriophage λ serves as a readout of
lacP classes of uninducible mutations.
the level of resources in the cell?
32. You are studying the properties of a new kind of reg-
27. What is the function of the cIII protein in the bacterio-
ulatory mutation of the lactose operon. This muta-
phage λ genetic switch?
tion, called S, leads to the complete repression of the
lacZ, lacY, and lacA genes, regardless of whether lac-
tose is present. The results of studies of this mutation
in partial diploids demonstrate that this mutation is following mutant strains when the cells are grown
completely dominant over wild type. When you treat in the presence of tryptophan? In its absence?
bacteria of the S mutant strain with a mutagen and 1. R+ O+ A+ (wild type)
select for mutant bacteria that can express the enzymes
encoded by lacZ, lacY, and lacA genes in the presence 2. R− O+ A+ /R+ O+ A−
of lactose, some of the mutations map to the lac opera- 3. R+ O− A+ /R+ O+ A−
tor region and others to the lac repressor gene. On the 34. The activity of the enzyme β-galactosidase produced by
basis of your knowledge of the lactose operon, provide wild-type cells grown in media supplemented with dif-
a molecular genetic explanation for all these properties ferent carbon sources is measured. In relative units, the
of the S mutation. Include an explanation of the con- following levels of activity are found:
stitutive nature of the “reverse mutations.”
33. The trp operon in E. coli encodes enzymes essential for Glucose Lactose Lactose + glucose
the biosynthesis of tryptophan. The general mechanism 0 100 1
for controlling the trp operon is similar to that observed
with the lac operon: when the repressor binds to the Predict the relative levels of β-galactosidase activity in
operator, transcription is prevented; when the repressor cells grown under similar conditions when the cells are
does not bind to the operator, transcription proceeds. The lacI −, lacI S, lacOC , and crp− .
regulation of the trp operon differs from the regulation of 35. A bacteriophage λ is found that is able to lysogenize
the lac operon in the following way: the enzymes encoded its E. coli host at 30°C but not at 42°C. What genes
by the trp operon are synthesized not when tryptophan is may be mutant in this phage?
present, but rather when it is absent. In the trp operon,
the repressor has two binding sites: one for DNA, and 36. What would happen to the ability of bacteriophage λ
the other for the effector molecule, tryptophan. The trp to lyse a host cell if it acquired a mutation in the OR
repressor must first bind to a molecule of tryptophan binding site for the Cro protein? Why?
before it can bind effectively to the trp operator. 37. Sketch the effects of exposure of host cells to UV on
a. Draw a map of the tryptophan operon, indicating the the bacteriophage λ genetic switch.
promoter (P), the operator (O), and the first struc- 38. Contrast the effects of mutations in genes encod-
tural gene of the tryptophan operon (trpA). In your ing sporulation-specific σ factors with mutations in
drawing, indicate where on the DNA the repressor the −35 and −10 regions of the promoters of genes
protein binds when it is bound to tryptophan. in their regulons. Would functional mutations in the
b. The trpR gene encodes the repressor; trpO is the σ-factor genes or in the individual promoters have the
operator; trpA encodes the enzyme tryptophan greater effect on sporulation?
synthetase. A trpR− repressor cannot bind trypto-
phan, a trpO− operator cannot be bound by the GENETICS AND SOCIETY
repressor, and the enzyme encoded by a trpA− How might an understanding of the regulation of gene expres-
mutant gene is completely inactive. Do you expect sion in bacteria be important for the treatment or prevention
to find active tryptophan synthetase in each of the of human disease?
Regulation of Transcription
in Eukaryotes
399
I
n Chapters 8 and 11, you learned that transcription in directly bind DNA and those that do not. The first set of reg-
bacteria is often regulated by single activator or repres- ulatory proteins consists of transcription factors that directly
sor proteins that directly bind DNA. Initial expectations bind regulatory DNA sequences called enhancers. Enhancers
were that eukaryotic transcription would be regulated by that are located close to the core promoter are part of
similar means. However, in most eukaryotes, multiple pro- proximal promoters and are called proximal enhancers,
teins and DNA sequences work together to control tran- and those that are a considerable distance from the pro-
scription. A key additional difference between bacteria and moter are part of distal enhancers and are called enhancers
eukaryotes is that in eukaryotes access to transcription ( Figure 12-1 ). In addition, some general transcription
regulatory sequences in DNA is restricted by the packag- factors (GTFs) directly bind DNA regulatory sequences
ing of DNA with proteins in the nucleus. Gene regulation within core promoters that surround transcription start sites.
in eukaryotes involves proteins that promote or restrict The second set of regulatory proteins consists of coreg-
access of RNA polymerases to gene promoters. This chap- ulators, which do not directly bind DNA. There are two
ter will focus on the transcription of protein-coding genes types of coregulators: coactivators and corepressors .
by RNA polymerase II and thereby provide the foundation Coacti vators and corepressors, respectively, increase or
for understanding the regulation of transcription in time decrease the amount of transcription through binding or
and space that choreographs the process of development enzymatically modifying other transcription regulatory fac-
described in Chapter 13. tors. For example, some coactivators serve to bridge tran-
scription factors and RNA polymerase II (Figure 12-2), while
others alter the structure of chromatin, which is described
later in this chapter.
12.1 TRANSCRIPTION FACTORS
REGULATE TRANSCRIPTION KEY CONCEPT Distal and proximal enhancers are DNA
sequences that regulate the transcription of genes. Coregula-
tors, which bind transcription factors, control the recruitment
LO 12.1 Diagram how transcription factors and DNA and access to DNA of general transcription factors and RNA
enhancer elements control the transcription
polymerase II.
of individual genes.
The machinery required for generating the distinct pat- Eukaryotic transcription regulatory mechanisms have
terns of gene transcription that occur in eukaryotic cells been discovered through both biochemical and genetic
has many components, including trans-acting regulatory approaches. The latter has been advanced in particular by
proteins and cis-acting regulatory DNA sequences. The studies of the single-celled yeast Saccharomyces cerevisiae
regulatory proteins can be divided into two sets, those that (see the Model Organism box on page 404). This organism,
400
3.0
1.0
FIGURE 12-3 Point mutations throughout the proximal enhancer and core promoter regions of the β-globin gene
were analyzed for their effects on transcription level. The height of each line represents the transcription level relative
to a wild-type gene (set to 1.0). Only the base substitutions that lie within the three labeled elements changed the
level of transcription. Wild-type sequences are shown for the proximal enhancer and core promoter elements.
Positions with black dots were not tested. [Data from T. Maniatis, S. Goodbourn, and J. A. Fischer, “Regulation of
Inducible and Tissue-Specific Gene Expression,” Science 236, 1987, 1237.]
UDP-glucose
Transcription factors: lessons from the Gal7
yeast GAL system
Glucose-1-phosphate
Yeast make use of extracellular galactose (gal) by import-
ing and converting it into a form of glucose that can be
metabolized. Several genes— GAL1, GAL2, GAL7, and
GAL10—in the yeast genome encode enzymes that cata- Glycolysis
lyze steps in this metabolic pathway (Figure 12-5). Three
additional genes—GAL3, GAL4, and GAL80— encode FIGURE 12-5 Galactose is A N I M ATED
converted into glucose1phosphate A RT
proteins that regulate transcription of the enzyme-encoding in a series of biochemical steps.
genes. Just as in the lac system of E. coli, the abundance Galactose metabolism
These steps are catalyzed by the
of the sugar determines the level of transcription in the enzymes Gal1, Gal2, Gal7, and in yeast
metabolic pathway. In yeast cells growing in media lack- Gal10, which are encoded by the
ing galactose, the GAL genes are largely transcriptionally genes GAL1, GAL2, GAL7, and
GAL10, respectively.
silent. But, in the presence of galactose (and the absence of
glucose), the GAL genes are transcriptionally induced. Just
Between the GAL1 and GAL10 transcription start sites is
as for the lac operon, genetic and molecular analyses of
a 118-base-pair region that contains four Gal4-binding sites.
mutants have been key to understanding how transcription
Each Gal4-binding site is 17 base pairs long and is bound by
of genes in the galactose pathway is controlled.
a homodimer of Gal4 proteins (two Gal4 proteins bound
Gal4 binds enhancers called upstream activation together). There are also two Gal4-binding sites upstream of
sequences The key regulator of GAL gene transcription the GAL2 gene and another two upstream of the GAL7 gene.
is the Gal4 transcription factor, a sequence-specific DNA- These binding sites are required for transcription activation.
binding protein. In the presence of galactose, mRNA levels If the binding sites are deleted, the genes are transcription-
for the GAL1, GAL2, GAL7, and GAL10 genes are about ally silent, even in the presence of galactose. Because the Gal4
1000-fold higher than in its absence. However, in GAL4 enhancers are located upstream of the genes they regulate, they
mutants they are unchanged, indicating that Gal4 is required are also called upstream activation sequence (UAS) elements.
for transcription of these genes. Each of the four genes has
two or more Gal4-binding sites (i.e., enhancers) located at KEY CONCEPT Transcription factors coordinately regulate
the transcription of multiple genes involved in the same
some distance 5′ (upstream) of its promoter (Figure 12-6).
biological process by binding enhancers that are common to
Consider the GAL10 and GAL1 genes, which are adja-
the genes.
cent to each other and transcribed in opposite directions.
Activation domain
Gal4 Dimerization domain
DNA-binding domain
UAS
FIGURE 12-6 Gal4 activates the transcription of target genes by binding upstream activation A N I M ATED A RT
sequence (UAS) elements (red rectangles). The Gal4 protein has three functional domains. The
protein binds as a dimer to inverted repeats of the same sequence (arrows show directionality of the Gal4 regulation of galactose-
sequence) located upstream of the promoters of Galpathway genes. Some of the GAL genes are on metabolizing enzymes
the same chromosome (GAL7, GAL10, GAL1), whereas others are on different chromosomes (GAL2).
In addition to its action in yeast cells, Gal4 has been genes (see Chapter 10) that contained Gal4-binding sites
shown to activate transcription of UAS-containing genes (i.e., UAS) or LexA-binding sites (i.e., LexA site) upstream
when they are introduced into insect cells, human cells, and of a promoter and the E. coli lacZ gene coding region. The
many other eukaryotic organisms. This versatility suggests level of transcription of lacZ in yeast cells was determined
that transcription machineries and mechanisms of gene by measuring the level of its encoded protein product
activation are common to a broad array of eukaryotes, and β-galactosidase. Full-length Gal4 activated transcrip-
that features revealed in yeast are generally present in other tion when bound to the UAS (Figure 12-7a) but the Gal4
eukaryotes and vice versa. Furthermore, because of their DNA-binding domain lacking the activation domain did not
versatility, Gal4 and its UAS elements have become favored (Figure 12-7b). Similarly, the LexA DNA-binding domain did
tools for manipulating gene expression in a wide variety of not activate transcription from LexA sites (Figure 12-7c), but
model organisms. a protein fusion of the Gal4 activation domain and the LexA
DNA-binding domain did (Figure 12-7d). Likewise, a fusion
KEY CONCEPT The ability of Gal4 to function in a variety of of the Gal4 binding domains to another activation domain
eukaryotes indicates that eukaryotes generally have common was able to activate transcription (not shown).
transcription regulatory machineries and mechanisms. Researchers have used the modularity of transcription
factors to develop technologies such as the yeast two-hybrid
system that is used to detect protein-protein interactions in
Gal4 domains function independently vivo (Chapter 14). The modularity of transcription factors is
of one another also the cause of some cancers such as acute promyelocytic
In addition to the DNA-binding domain and dimeriza- leukemia (APL), a cancer of early blood-forming cells. In
tion domain, Gal4 has an activation domain. A series almost all cases of APL, a chromosome translocation creates
of simple, elegant experiments demonstrated that the a gene fusion between the activation domain of PML (pro-
DNA-binding and activation domains of Gal4 as well as myelocytic leukemia) and the DNA-binding and ligand bind-
other transcription factors are modular; that is, they func- ing domains of RARA (retinoic acid receptor α ). The fusion
tion independently of one another ( Figure 12-7 ). In this protein assembles with corepressor proteins, instead of coac-
study, researchers fused the Gal4 activation domains to tivator proteins, to block transcription of normal RARA gene
the DNA-binding domain from the E. coli transcription targets that control the differentiation of myeloid (blood)
factor LexA. Transcription was measured using reporter cells, which leads to uncontrolled proliferation of these cells.
+
a/a
a/a (2n)
Ascus Meiosis (2n)
a(n) a(n)
a(n) a(n)
beer, bread, and wine. Yeast has many features that make it Culture Culture
an ideal model organism. As a unicellular eukaryote, it can colony colony
be grown on agar plates and, with yeast’s life cycle of just The life cycle of baker’s yeast. The nuclear alleles MATa and MATα determine
mating type. Ploidies, n and 2n, are indicated in parentheses.
90 minutes, large quantities can be cultured in liquid media.
It has a very compact genome with only about 12 megabase
pairs of DNA (compared with almost 3000 megabase pairs
for humans) containing approximately 6000 genes that are isolate mutants using a forward genetic approach, hap-
distributed on 16 chromosomes. It was the first eukaryote to loid cells are mutagenized (with X rays, for example) and
have its genome sequenced. screened on plates for mutant phenotypes. This procedure is
The yeast life cycle makes it very versatile for laboratory usually done by first plating cells on a rich medium on which
studies. Cells can be grown as either diploids or haploids. In all cells grow and by copying, or replica plating, the colonies
both cases, the mother cell produces a bud containing an from this master plate onto replica plates containing selec-
identical daughter cell. Diploid cells either continue to grow tive media or special growth conditions. For example, tem-
by budding or are induced to undergo meiosis, which pro- perature-sensitive mutants will grow on the master plate at
duces four haploid spores held together in an ascus (also the permissive temperature but not on a replica plate at a
called a tetrad). Haploid spores of opposite mating type (a restrictive temperature. Comparison of the colonies on mas-
or α) will fuse and form a diploid. Spores of the same mating ter and replica plates will reveal the temperature-sensitive
type will continue growth by budding. mutants. Using reverse genetics, scientists can replace any
Yeast has been called the E. coli of eukaryotes because yeast gene of known or unknown function with a mutant ver-
of the ease of forward and reverse mutant analysis. To sion to understand the nature of the gene product.
KEY CONCEPT Eukaryotic transcription factors are modular, and directly inhibits Gal4 activity. Specifically, Gal80 binds
having separable domains for DNA binding, activation/ within the Gal4 activation domain, blocking its ability to
repression, dimerization, and ligand-binding. promote transcription. Gal80 is expressed continuously,
so it is always acting to repress transcription of the GAL
structural genes unless stopped.
Regulation of Gal4 The role of Gal3 is to release the GAL structural genes
Gal4 activity is regulated by the Gal80 and Gal3 proteins from their repression by Gal80 when galactose is present.
(Figure 12-8). In GAL80 mutants, the GAL structural genes Gal3 is thus both a sensor and inducer. When Gal3 binds
(GAL1, GAL2, GAL7, and GAL10) are transcriptionally galactose and ATP, it undergoes a conformational change
active even in the absence of galactose. This suggests that the that promotes binding to Gal80, which in turn causes Gal80
normal function of Gal80 is to inhibit GAL gene transcrip- to be released from Gal4, which is then able to interact with
tion. Conversely, in GAL3 mutants, the GAL structural genes coactivators and RNA polymerase II to activate transcrip-
are not active in the presence of galactose, suggesting that tion. Thus, Gal80, Gal3, and Gal4 are all part of a switch,
Gal3 normally promotes transcription of the GAL genes. whose state is determined by the presence or absence of
Extensive biochemical analyses revealed that Gal80 is galactose (Figure 12-8). In this switch, DNA binding by the
a corepressor of Gal4. Gal80 binds Gal4 with high affinity transcription factor is not the physiologically regulated step
404
lacZ
OFF
UAS Active Gal4
GAL1
(c) LexA lacking the activation domain ON
UAS
DNA-binding
domain lacZ
FIGURE 12-8 Gal4 activity is A N I MAT ED
OFF
regulated by Gal80 and Gal3. (Top) A RT
LexA site
In the absence of galactose, Gal4
Gal4 regulation of galactose-
is inactive, even though it can bind
UAS elements upstream of GAL metabolizing enzymes
(d) Gal4–LexA hybrid genes such as GAL1. Gal4 activity
is repressed by the binding of Gal80. (Bottom) In the presence of
Gal4 activation
galactose, Gal3 induces a conformation change in Gal80, releasing it
domain
from Gal4, which can then activate GAL gene transcription.
LexA
DNA-binding
domain lacZ
of genes. The activation or repression of sets of genes must
ON
LexA site therefore be coordinated in the making of particular cell types.
One of the best-understood examples of cell type-specific reg-
FIGURE 12-7 Transcription factors AN IM ATED ulation in eukaryotes is the regulation of mating type in yeast.
have multiple, separable domains. A RT This regulatory system has been dissected by a combination
(a) Fulllength Gal4 has three of genetics, molecular biology, and biochemistry. Mating type
domains and activates transcription Gal4 modularity and gene
induction in non-yeast serves as an excellent model for understanding the logic of
from UAS sites. (b) Removal of
transcription regulation in multicellular animals.
the Gal4 activation domain shows
that dimerization and DNA binding is not sufficient for transcription The yeast Saccharomyces cerevisiae can exist in any of three
activation. (c) Similarly, the LexA DNAbinding domain cannot activate different cell types known as a, α(alpha), and a /α. The two cell
transcription, but (d) when fused to the Gal4 activation domain, it can types a and α are haploid and contain only one copy of each
activate transcription through LexAbinding sites. chromosome. The a /α cell is diploid and contains two copies of
each chromosome. Although the two haploid cell types cannot
(as is the case in the lac operon and bacteriophage λ ); rather, be distinguished by their appearance, they can be differentiated
the ability of the activation domain to perform its function is by a number of specific cellular characteristics, principally their
regulated. mating type (see the Model Organism box on page 404). An
α cell mates only with an a cell, and an a cell mates only with
KEY CONCEPT Environmental signals such as galactose alter an α cell. An α cell secretes an oligopeptide pheromone, or sex
the activity of eukaryotic transcription factors by controlling hormone, called α factor that arrests a cells in the cell cycle.
their interactions with other proteins. Similarly, an a cell secretes a pheromone, called a factor, that
arrests α cells. Cell arrest of both participants is necessary for
successful mating. The diploid a /α cell does not mate, is larger
Combinatorial control of transcription: than a and α cells, and does not respond to mating hormones.
lessons from yeast mating type Genetic analysis of mutants defective in mating has
Thus far, we have focused on transcription regulation of single shown that cell type is controlled by a single genetic locus,
genes or a few genes in one pathway. In multicellular organ- the mating-type locus, MAT. There are two alleles of the
isms, distinct cell types differ in the transcription of hundreds MAT locus: haploid a cells have the MATa allele, and
a1 a1
MAT locus
α2 α1 α2 α1
Transcription factors a1 α2 a1
α1 α2
MCM1 MCM1 MCM1
α2 α2 α2 α2
a-specific genes ON OFF OFF
MCM1 MCM1 MCM1
α1
𝛂-specific genes OFF ON OFF
MCM1
a1 α2
Haploid-specific genes ON ON OFF
FIGURE 12-9 Control of cell type-specific transcription in yeast. The three cell types of S. cerevisiae
are determined by differential expression of the transcription factors a1, α1, and α2, which regulate
different subsets of target genes. The MCM1 transcription factor is expressed in all three cell types,
but its function is changed by interactions with α1 and α2.
haploid α cells have the MATα allele. The a /α diploid has in haploid cells but not diploid cells. The a1 transcription
both alleles. Yeast can switch mating type by a homologous factor encoded by MATa has a part to play at last. a1 can
recombination event that replaces one MAT allele with the bind α2 and alter its binding specificity such that the a1–α2
other MAT allele. These two alleles activate different sets of complex binds enhancers found upstream of haploid-specific
genes because they encode different transcription factors. genes and silences these genes. In diploid cells, then, the α2
In addition, a transcription factor not encoded by the MAT protein exists in two forms: (1) as an α2–MCM1 complex
locus, called MCM1, plays a key role in regulating cell type. that represses a-specific genes, and (2) in a complex with the
The simplest case is the a cell type (Figure 12-9a). The a1 protein that represses haploid-specific genes. Moreover,
MATa locus encodes a single transcription factor, a1. How- the a1–α2 complex also represses expression of the α1 gene,
ever, a1 has no effect in haploid cells, only in diploid cells. which is thus no longer present to turn on α-specific genes.
In a haploid a cell, the transcription factor MCM1 turns on Thus, cell type-specific transcription of genes that control
the expression of the structural genes needed by an a cell by mating type in yeast is achieved by multiple transcription fac-
binding enhancers for a-specific genes. tors working in different combinations.
In an α cell, the α-specific structural genes must be
transcribed, but, in addition, MCM1 must be prevented KEY CONCEPT The control of yeast mating type is an
from activating the a-specific genes (Figure 12-9b). The example of how cell type-specific patterns of transcription
DNA sequence of the MATα allele encodes two transcrip- in eukaryotes can be governed by different combinations of
tion factors, α1 and α2, that are produced by separate interacting transcription factors.
genes. These two proteins have different regulatory roles in
the cell. The α1 protein is an activator of α-specific tran-
scription. It binds in concert with the MCM1 protein to 12.2 CHROMATIN STRUCTURE
an enhancer that controls α-specific genes. The α2 protein
represses transcription of the a-specific genes by binding LO 12.2 Draw a segment of chromatin, labeling each
with MCM1 to enhancers upstream of a-specific genes. histone, a nucleosome, and the structural features
In a diploid yeast cell, transcription factors encoded by that are important to their function in transcription.
each MAT locus are expressed (Figure 12-9c). This results in
repression of all genes involved in cell mating and a sepa- In eukaryotic cells, DNA is packaged with proteins to cre-
rate set of genes, called haploid specific, that are expressed ate chromatin. In the cell’s nucleus, DNA in chromatin
is compacted over 10,000-fold compared to its linear the central region of histone proteins. They are approxi-
form. The structure of chromatin serves to fit DNA into mately 70 amino acids in length and are made up of three
the nucleus, and it also serves as a substrate for reversible α -helices separated by loops (Figures 12-10b and c).
changes in protein-DNA and protein-protein interactions Hydrophobic contacts between α-helices of histone folds
that regulate transcription. In this section, we describe are critical for the specific pairing of H2A with H2B and
the structure of histones (the major protein components H3 with H4. Histone-fold extensions also make a signif-
of chromatin), nucleosomes (the basic structural units of icant contribution to the specificity of histone pairing.
chromatin), and higher-order chromatin structures (three- Finally, as the name suggests, flexible tails are located at
dimensional assemblies of nucleosomes with one another). the ends of histone proteins (Figure 12-10d). The tails are
Because higher-order chromatin structures in eukaryotic largely unstructured and are involved in interactions with
cells can make DNA inaccessible to binding by transcrip- non-histone proteins as well as neighboring nucleosomes.
tion factors, an understanding of chromatin structure The structure of linker histone H1 is substantially dif-
is essential for understanding how transcription is ferent than that of core histones. It is larger (~21 kDa)
regulated. and it has much greater sequence and structural diversity.
For example, in humans, histone H1 has three domains, a
central domain of approximately 80 amino acids flanked
Histones by unstructured N- and C-terminal tails of approximately
Eukaryotic cells express five types of histone proteins: H1, 20 and 100 amino acids, respectively; whereas in yeast, his-
H2A, H2B, H3, and H4. Histones H2A, H2B, H3, and H4 tone H1 has only a single, unstructured domain.
are known as core histones because they form a core com-
plex around which DNA is wrapped to form nucleosomes. Nucleosomes
Histone H1 is known as a linker histone because it binds
Nucleosomes are the basic structural units of chromatin.
the DNA that links adjacent nucleosomes. In addition to
They contain 146 base pairs of DNA that wrap about 1.7
these canonical histones that package the newly replicated
times around a histone octamer (eight proteins) consisting
genome, there are variant histones that are incorporated
of two copies of each of the four core histones H2A, H2B,
into nucleosomes in a DNA-replication independent man-
H3, and H4 (only one copy is shown in Figure 12-10c).
ner. As an example, the histone variant H2A-Z is 60 percent
The stability of nucleosomes is due to many protein-
identical in sequence to canonical histone H2A. H2A-Z
protein interactions within the histone octamer and elec-
replaces H2A in nucleosomes at promoters of both tran-
trostatic and hydrogen bonds between histones and
scriptionally active and silent genes. In contrast, another
DNA. The flexible tails of histones extend away from the
H2A variant called H2A-X is incorporated into nucleo-
nucleosomal DNA and are involved in interactions with
somes at sites of DNA damage (Chapter 15). Variants of
adjacent nucleosomes and numerous nuclear factors
H1, H2B, and H3 also play specialized roles, but, as yet, no
(Figure 12-10d).
variants of H4 have been identified.
Molecular machines that assemble and disassemble
Histone proteins have unusual features that are rele-
nucleosomes play important roles in regulating transcrip-
vant to their roles as structural and regulatory components
tion. During DNA replication, nucleosome formation
of chromatin. They are extremely abundant. In mamma-
begins with assembly of an H3/H4 tetramer (two H3/H4
lian cells, histones constitute approximately 70 percent
dimers joined together) on DNA followed by sequential
of the protein complement of chromatin. There are about
addition of two H2A/H2B dimers. Binding of histone H1
10 million copies of each core histone per cell and about
to nucleosomes organizes an additional 20 base pairs of
half this amount of histone H1. Core histone proteins are
linker DNA to form a complete nucleosome. Removal of
small (11–15 kDa), unusually basic (at least 20 percent of
histones from DNA occurs in the reverse order, beginning
their amino acids are lysine or arginine), and positively
with sequential removal of the H2A/H2B dimers, followed
charged at neutral pH. Electrostatic interactions between
by removal of the H3/H4 tetramer. Neighboring nucleo
positively charged amino acids and the negatively charged
somes are separated from one another by ~20–75 base pairs
phosphate backbone of DNA play an important role in
of linker DNA.
determining the structure of chromatin. The sequences of
core histone proteins are among the most highly conserved KEY CONCEPT In eukaryotes, DNA is packaged with his-
in evolution. From yeast to humans, both H2A and H2B tones in chromatin. Nucleosomes, the units of chromatin,
sequences are more than 70 percent identical, and both H3 contain two copies of each of the core histones (H2A, H2B,
and H4 are more than 90 percent identical. Because of this H3, and H4) around which is wrapped 146 base pairs of DNA.
conservation, studies of histones in genetically controlla- Complete nucleosomes also contain histone H1 and linker
ble organisms such as yeast and Drosophila have provided DNA of variable length.
considerable insights into the function of histones in higher
eukaryotic organisms such as humans.
Core histone proteins have three types of structural Chromatin folding
domains: histone folds, histone-fold extensions, and Wrapping of DNA around histone octamers forms a struc-
flexible tails (Figure 12-10a). Histone folds are located in ture of ~11 nanometers (nm) in diameter and compacts
(a) (b) L2
α1 L1 α2 L2 α3
L1
H3
H3
H4 α1
H2A
α3
α2 α2
H2B
α1
Flexible tail
Histone-fold extension α3
L1
Histone-fold
H4 H3
L2
(c) (d)
H2A
H4
H2A
H2B
FIGURE 12-10 (a) Primary structures of the four core histones. α1, α 2, and α3 are alphahelices and
L1 and L2 are loops. (b) Interactions between the histonefold structures of histones H3 and H4. (c)
The structure of a nucleosome, showing the histonefold regions of one copy of each of the four core
histones wrapped by DNA (gray). (d) The same view of the nucleosome as in (c) with the addition of
H2A and H3 histonefold extensions and flexible tails for all four core histones that extend beyond the
wrapped DNA. [PDB ID 5y0c and 1aoi.]
DNA ~sixfold (Figure 12-11). This does not come close to chromosome arms, is gene-rich, and, through mechanisms
the 10,000-fold compaction that occurs in eukaryotic cells. described later in this chapter, can lose its compact struc-
To achieve higher levels of compaction, nucleosomes fold ture and become transcriptionally active euchromatin.
upon themselves. The next order of chromatin folding pro-
duces the 30-nm fiber, a structure ~30 nm in diameter, and KEY CONCEPT Regions of the genome with few genes, such
there are even more compact structures. as centromeres and telomeres, are compacted into hetero-
During the cell cycle, chromosomes vary in their level chromatin throughout the cell cycle, whereas regions that are
of compaction. Nucleosomes in mitosis are much more gene-rich vary in their level of chromatin compaction. Typically,
genes are transcriptionally silent when compacted into heter-
highly compacted than in interphase (see Appendix 2-1).
ochromatin, and they can be transcriptionally active when less
Even in interphase, regions of chromosomes vary in their
compacted into euchromatin.
level of compaction. More compacted regions are called
heterochromatin, and less compacted regions are called
euchromatin . Heterochromatin constitutes a significant Over the past 20 years, new technologies such as chro-
fraction of some eukaryotic genomes—approximately 20 matin immunoprecipitation (ChIP, Chapter 14) have made
percent for humans and 30 percent for Drosophila—but it possible to determine the genome-wide distribution of
very little of others—less than 1 percent for the yeast. nucleosomes. As an example, in the yeast S. cerevisiae, Frank
Chromatin that remains heterochromatic throughout the Pugh and colleagues found that approximately 70,000
cell cycle is called constitutive heterochromatin, is concen- nucleosomes occupy 81 percent of the genome and are
trated at centromeres and telomeres, and is rich in repet- typically separated by an 18-base-pair linker. Furthermore,
itive sequences such as transposons but poor in genes. In nucleosomes are not equally distributed in the genome; they
contrast, facultative heterochromatin is spread along cover 87 percent of transcribed regions, but only 53 percent
(a)
Naked DNA
Insulator Enhancer Promoter Gene body Insulator
Repressed −1 +1
NFR
Active
NFR NFR
(b)
CTCF
TF
TAD
GTF
CTCF
FIGURE 12-12 Genes contain insulator, enhancer, and promoter elements that regulate transcription.
(a) Nucleosomes, indicated by purple circles, cover most of the transcribed region of genes (i.e., the
gene bodies) but are excluded from enhancers and promoters of active genes to create nucleosome
free regions (NFRs). (b) In chromosomes, groups of genes are segregated from one another by
topologically associating domains (TADs) that are formed by the interaction of insulator binding proteins
such as CTCF. Within TADs, enhancers bound by transcription factors (TFs) are positioned to act on
general transcription factors (GTFs) at particular promoters.
defined by specialized regulatory sequences called insulators Histone acetylation and deacetylation reactions
or boundaries that interact with one another or possibly the
nuclear envelope through their associated proteins. In mam- Acetyl-CoA
mals, most of the known insulator sequences are bound by CH3 Neutral
a zinc-finger DNA-binding protein called CTCF (CCCTC- Positive charge
C O
binding factor). Therefore, insulators divide chromosomes charge CH3
S SH
into precisely defined loops that determine which enhancer- +
NH3
Coenzyme A Coenzyme A
C O
promoter interactions are allowed and which are prevented. CH2 NH
HAT
CH2 CH2
KEY CONCEPT The wrapping of DNA enhancer elements CH2 CH2
into nucleosomes can prevent binding by transcription factors. CH2 HDAC CH2
Insulators prevent enhancers and their associated transcription CH3 H2O CH2
X N C C X
factors from activating the transcription of genes outside
H Histone C O X N C C X
a TAD.
Lysine OH H Histone
Acetylated lysine
12.3 CHROMATIN REGULATES
TRANSCRIPTION FIGURE 12-13 The positively charged side chain of lysine residues in
histones is neutralized by post-translational addition of an acetyl group.
Acetylation is catalyzed by histone acetyltransferases (HATs) that use
LO 12.3 Compare and contrast how chromatin modifying acetyl-CoA as the donor of the acetyl group, indicated in red. The
and chromatin remodeling mechanisms contribute reverse reaction, deacetylation, is catalyzed by histone deacetylases
to gene-specific transcription regulation. (HDACs).
writers, HDACs are erasers, and bromodomains are readers arginine can be methylated one time (monomethylargine)
of histone acetylation (Figure 12-14). or two times in symmetric or asymmetric configurations
(dimethylarginine) (Figure 12-15). Methylation is controlled
KEY CONCEPT Acetylation of lysines in histones by HATs by writers (histone methyltransferases, HMTs), erasers
(1) loosens interactions within and between nucleosomes and (histone demethylases, HDMs), and readers (proteins that
(2) creates a binding site for bromodomains, found in some contain a chromodomain or a plant homeodomain (PHD)
transcription coregulators.
finger). There are also writers, erasers, and readers for
phosphorylation, ubiquitination, and other modifications,
and both histones and DNA can be modified.
In summary, histone acetylation plays a crucial role in
stepwise mechanisms that activate transcription: a tran-
KEY CONCEPT Transcription is regulated by chemical mod-
scription factor binds an enhancer, a HAT such as GCN5 ifications of amino acids in histones and nucleotides in DNA.
binds the transcription factor, the HAT acetylates his- Modifications are added by writer enzymes, removed by eraser
tones in nucleosomes at the promoter, a bromodomain enzymes, and bound by reader proteins.
protein such as TAF1 binds acetylated histones, and RNA
polymerase II is recruited either directly or indirectly by
the bromodomain protein. Similarly, an activated gene is
turned off by transcription factor-mediated binding of an The histone code hypothesis
HDAC, which deacetylates histones and blocks recruitment Post-translational modifications occur in all parts of
of bromodomain-containing proteins. Acetylation affects histone proteins but are concentrated in the tails. They
transcription initiation, and elongation, by being targeted are experimentally detected in vivo and in vitro using
to nucleosomes positioned in different regions of genes and modification-specific antibodies, and in vitro by mass spec-
thereby affecting recruitment of bromodomain-containing trometry of histones purified from cells. Unfortunately,
initiation and elongation factors (Chapter 8). these methods are largely unable to detect the extent to
Acetylation is one of many histone modifications that which modifications coexist on an individual histone pro-
affect transcription. Other abundant modifications include tein. This information may be very important because
methylation of lysine and arginine residues; phosphoryla- different combinations of histone modifications may con-
tion of serine, threonine, and tyrosine residues; and ubi vey the information to bring about different transcription
quitination of lysine residues. Furthermore, lysine can be outputs. In 2000, this idea was formalized by David Allis
methylated one, two, or three times (monomethyllysine, and Thomas Jenuwein in the histone code hypothesis,
dimethyllysine, and trimethylysine, respectively); while which proposes that multiple histone modifications, acting
Histone tail
FIGURE 12-14 The information content of histones and DNA is changed by writer enzymes that add and eraser
enzymes that remove chemical modifications. Reader proteins interpret the information by binding chemical
modifications. The text provides detailed descriptions of writer, eraser, and reader proteins for histone acetylation,
methylation, and phosphorylation as well as DNA methylation.
(b)
sequentially or in combination on one or several histone 14-3-3 proteins, and acetylation of lysine 14 on histone H3
tails, specify unique transcription outcomes. (H3K14ac), which is bound by bromodomain proteins, sig-
In support of the histone code hypothesis, nucleo- nals transcription activation. The potential information con-
somes at promoters of transcriptionally active genes are tent of histone modifications is enormous. For example, there
commonly trimethylated (me3) on lysine (K) 4 of histone are more than two million possible combinations of modifi-
H3 (H3K4me3), whereas promoters of transcriptionally cations that can occur on the N-terminal tail of histone H3.
repressed genes are trimethylated on H3K9 (H3K9me3)
(Figure 12-16). H3K4me3 activates transcription by serv- KEY CONCEPT The histone code hypothesis posits that
ing as a binding site for transcription coactivators such as different combinations of histone modifications create unique
the PHD finger-containing TAF3 subunit of TFIID. In con- binding sites that can be read by transcription coregulators,
trast, H3K9me3 represses transcription by serving as a thereby conferring a variety of transcriptional outcomes.
binding site for transcription corepressors such as the chro-
modomain-containing protein heterochromatin protein 1
DNA modification: another type of
(HP1), which promotes the formation of heterochromatin.
Combinations of modifications are also hallmarks of tran- chromatin modification
scription activity—the combination of phosphorylation (P) Like histone modifications, DNA modifications affect
of serine (S) 10 on histone H3 (H3S10P), which is bound by transcription. In vertebrates, the predominant DNA
14-3-3
PHD CD BD
R T K R K S T K R K K R K S K H3
2 3 4 8 9 10 11 14 17 18 23 26 27 28 36
FIGURE 12-16 The N-terminal flexible tail of histone H3 contains lysine (K), arginine (R), serine (S), and threonine (T) amino acids that are
post-translationally acetylated, methylated, or phosphorylated. Different combinations of modifications, termed a histone code, are thought
to alter the level of transcription to different extents by conveying different information to the transcription machinery. The information is read
by bromodomains (BD), chromodomains (CD), PHD fingers (PHD), 14-3-3 proteins, and other proteins and protein domains.
(b)
Sequence read after the
Original sequence sodium bisulfite reaction
modification is 5-methylcytosine (5mC), where methyla- In vertebrates, CpGs occur much less frequently than
tion by a DNA methyltransferase (DNMT) occurs at the would be expected based on the C + G content of genomes.
fifth carbon in the cytosine ring of the dinucleotide CpG This is due to widespread methylation of CpGs and subse-
(cytidine-phosphodiester bond-guanosine) (Figure 12-17a). quent conversion by deamination over evolutionary time
DNMTs use S-adenosyl methionine (SAM) as the methyl to TpG. Approximately 85 percent of CpGs are methylated
donor. In contrast to vertebrates, in plants, 5mC occurs in and are scattered throughout the genome (Figure 12-18). The
the dinucleotide CpG as well as other nucleotide contexts, remaining CpGs are unmethylated, and many of these are
and in Drosophila, C. elegans, and S. cerevisiae, little or no highly clustered in 200- to 4000-base-pair regions called CpG
DNA methylation has yet been detected. islands. Approximately half of all CpG islands are located in
CpG methylation is reversed by enzymes in the TET gene promoters, and the remaining half are roughly equally
(ten-eleven translocation) family (Figure 12-17a). Reversal divided between intragenic and intergenic locations. The
comprises three steps, with each step catalyzed by a TET majority of gene promoters are associated with a CpG island.
enzyme; the second and third intermediates are rapidly Unmethylated CpG islands at promoters are generally cor-
excised by a mechanism described in Chapter 15. In con- related with open chromatin and active transcription, whereas
trast, the first intermediate is more stable and is particularly methylated islands are associated with closed chromatin and
abundant in embryonic stem cells and adult neurons where repressed transcription. These effects on transcription are
it is bound by reader proteins that regulate genes involved mediated by proteins that distinctly bind unmethylated or
in development and tumorigenesis. methylated CpGs. In humans, the protein Cfp1 (CxxC finger
5mC is detected in the lab by the sodium bisulfite protein 1) binds unmethylated CpGs and recruits a histone
reaction. DNA isolated from cells is treated with sodium methyltransferase (HMT) that produces the transcription
bisulfite, which in single-stranded DNA efficiently con- activating histone modification H3K4me3. In contrast, meth-
verts cytosine to uracil but inefficiently converts 5mC to ylation of CpGs represses transcription by interfering with
thymine. After the sodium bisulfite reaction, the DNA is transcription factor binding to enhancers and by serving as a
sequenced, and cytosines that are converted to uracil are binding site for methyl binding domain (MBD) proteins that
read as thymine (T), whereas 5mC that are unchanged are recruit transcription repressors such as HDACs that deacetyl-
read as cytosine (C) (Figure 12-17b). ate lysine or HMTs that produce H3K9me3. Thus, DNMTs
are writers, TETs are erasers, and a variety of proteins are remodeling complex. In one study, mutagenized yeast cells
readers of unmethylated and methylated CpGs. were screened for the lack of growth on sucrose (s ugar
nonfermenting mutants, snf, pronounced “sniff”). In the
KEY CONCEPT Methylation of cytosine in CpG islands at other study, mutagenized yeast cells were screened for
gene promoters is correlated with the repression of transcrip- defective switching of mating type ( switch mutants, swi,
tion. Like modifications of histone proteins, CpG methylation pronounced “switch”). Many mutants for different loci
of DNA represses transcription by altering the affinity of tran-
were recovered in each screen, but one mutant gene was
scription factors, coregulators, and general transcription fac-
found to cause both phenotypes. Mutants at the swi2/
tors for chromatin.
snf2 locus could neither use sucrose effectively nor switch
mating type because the transcription of specific genes
was blocked. The protein encoded in the swi2/snf2 locus
Chromatin remodeling was found to be the ATPase subunit of the multisubunit
Chromatin remodeling is the workhorse in the process SWI/SNF (“switch-sniff”) chromatin remodeling complex.
of altering chromatin structure to regulate transcription. SWI/SNF affects transcription by remodeling nucleo-
Remodeling means changing histone-DNA interactions in somes in two steps, initially removing an H2A/H2B dimer
nucleosomes to render DNA either more or less accessible from DNA, followed by removal of the rest of the histone
to transcription regulators (Figure 12-19). To permit factors octamer. The gene specificity of the SWI/SNF complex is
to bind enhancer and promoter elements, remodeling moves provided by binding to transcription factors such as Gal4
nucleosomes to new locations. On moderately active genes, and through binding of a bromodomain-containing subunit
remodeling displaces H2A/H2B dimers in front of RNA of the complex to acetylated lysine. Thus, transcription fac-
polymerase II and replaces them behind the polymerase tors, histone modifying enzymes, and chromatin remodel-
during every round of transcription. H3/H4 tetramers may ing factors function in concert to regulate transcription.
only be displaced and replaced on highly active genes. In In contrast to the other chromatin remodeling complexes
contrast, eviction of histones from DNA may not be nec- that slide, eject, or replace histone octamers on DNA, the
essary to permit RNA polymerase II passage through genes SWR1 (pronounced “swur one”) complex remodels chroma-
being transcribed at a low level. tin by assembling the variant histone H2A-Z into chroma-
Chromatin remodeling complexes use energy from ATP tin. SWR1 does this by exchanging H2A-Z/H2B dimers for
hydrolysis to disrupt non-covalent histone-DNA inter- H2A/H2B dimers in histone octamers. This activity is tar-
actions. Eukaryotic organisms contain four families of geted to enhancers and promoters of specific genes by the
ATP-dependent chromatin remodeling complexes, which bromodomain-containing subunit of SWR1, which binds
can be characterized as being associated with transcription specific acetylated lysines on histone tails. Nucleosomes that
activation or repression. contain H2A-Z are particularly prone to disassembly by
Two genetic screens in yeast for mutants in seemingly other chromatin remodeling complexes, leading to increased
unrelated processes led to the discovery of a chromatin access of transcription regulators to DNA.
KEY CONCEPT Chromatin is dynamic; nucleosomes are not • H3S10P peaks at 6 hours, the time at which H3K14ac is
necessarily in fixed positions on the chromosome. Chromatin first detected (see Figure 12-20a), suggesting that H3S10P
remodeling complexes change nucleosome density, position, is required for GCN5 to acetylate H3K14 (Figure 12-20e).
and subunit composition to control access of the transcription This illustrates that modification of one amino acid can
machinery to DNA. promote or inhibit modification of other amino acids, a
process called crosstalk.
• TBP binding is first detected at 6 hours, which is the same
Connecting chromatin structure time that IFN-β transcription starts (see Figure 12-20a).
to transcription: lessons from the The TFIID complex is recruited through direct binding of
TBP to the TATA promoter element as well as binding of
interferon-β gene
TAF1 bromodomains to H3K9ac and H3K14ac (Figure
The human interferon-beta (IFN-β) gene, which encodes an 12-20e). In addition, the SWI/SNF chromatin remodeling
antiviral protein, is one of the best-characterized genes in complex is recruited by binding to H4K8ac as well as
eukaryotes. Its transcription is normally switched off but, interactions with another HAT called CBP (CREB-binding
upon viral infection, is activated to very high levels. A cen- proteins), which replaces GCN5 at the enhanceosome. To
tral feature of activation of this gene is assembly of mul- initiate IFN-β transcription, TFIID recruits RNA poly-
tiple, different transcription factors into an enhanceosome merase II to the promoter and SWI/SNF remodels nucleo-
about 100 base pairs upstream of the TATA box promoter somes to allow the polymerase to initiate transcription.
element and transcription start site. This illustrates that histone modifications aid in assembly
A study of the IFN-β gene illustrates how regulated of the transcription preinitiation complex as well as fac-
changes in chromatin structure affect transcription. Dimi- tors that remodel chromatin structure.
tris Thanos and colleagues used the chromatin immunopre-
• Last, transcription is turned off at 24 hours, at which
cipitation (ChIP) technique (Chapter 14) to identify histone
time activating histone modifications have mostly been
modifications and other events that occur at the IFN-β
removed by eraser enzymes (Figure 12-20g). This illus-
promoter as the gene shifts from transcriptionally inactive
trates the rapid reversibility of chromatin-mediated
to active and back to inactive over a 24-hour period in
control of transcription.
response to virus infection of human cells (Figure 12-20a).
They also used reverse transcription-PCR (RT-PCR) to Lessons learned from the IFN-β gene are generalizable.
quantify IFN-β mRNA levels. The following bullet points The molecular mechanisms that alter chromatin structure
walk through the temporal pathway of molecular events turn transcription of particular genes on or off in response
uncovered in the study and they highlight the general mech- to developmental and environmental signals by permitting
anisms by which chromatin structure affects transcription: or preventing transcription regulatory proteins access to
DNA. The information that controls transcription consists
• IFN-β mRNA is first detected 6 hours after viral infection,
of DNA regulatory sequences (e.g., enhancers, promoters,
but histone modifications are detected as early as 3 hours
and insulators), the histone code (e.g., various chemical
after viral infection (Figure 12-20a). This illustrates that
modifications of histone amino acids), and DNA modifi-
histone modifications occur prior to the first transcription
cations (e.g., unmodified CpG and 5mC) (Figure 12-21a).
initiation event to generate a chromatin structure that is
Proteins read the information through physical interactions
conducive to transcription initiation and elongation.
with the information-containing elements (Figure 12-21b).
• The earliest event is cooperative binding of a suite of DNA sequences are bound by transcription factors via
transcription factors to the IFN-β proximal enhancer
their DNA-binding reader domain. Modified histones
(Figure 12-20b) to form an enhanceosome (Figure 12-20c). and modified DNA are bound by other coregulators that
The histone acetyltransferase GCN5 then binds the contain a variety of reader domains (e.g., bromodomain,
assembled enhanceosome and acetylates H4K8 and H3K9 chromodomain, and methyl-binding domain). Once bound
and a kinase binds and phosphorylates H3S10 in nucleo- to chromatin, reader proteins serve as scaffolds for assem-
somes near the IFN-β promoter (Figure 12-20d). This bly of enzymes that change chromatin structure (Figure
illustrates that transcription factors recruit histone modi- 12-21c). Reader proteins are either part of a stable com-
fying enzymes to generate specific histone modifications plex that contains enzymes (e.g., SWI/SNF) or they recruit
in particular nucleosomes. enzymes through protein-protein interactions. The enzymes
• H4K8ac and H3S10P are last detected at 8 hours (see then either edit the information at the gene by modifying
Figure 12-20a), even though transcription occurs up to histones or DNA, or they change the chromatin structure
15 hours. In contrast, H3K9ac is last detected at 19 hours. by sliding nucleosomes, ejecting or replacing histone oc-
Unknown HDACs and phosphatases are involved in remov- tamers or parts of octamers, or exchanging variant histones
ing these histone marks. This illustrates that both writing for canonical histones (Figure 12-21d). Collectively, these
and erasing of histone modifications as well as the relative events alter the access to DNA by RNA polymerase II, tran-
timing of histone modifications are important for transcrip- scription initiation factors (e.g., TFIID), and transcription
tion regulation. elongation factors (e.g., P-TEFb) (Figure 12-21e). Lastly,
TBP (TFIID)
(b) Prior to viral infection (transcription OFF) (e) 6–8 hours after viral infection (transcription ON)
H3 H3
K9 K9 ac
S10
SWI/SNF S10 p
H4 K8 H4 K8 ac
K14 GCN5 K14 ac RNA pol II
K9 ac K9
H3 H3
(c) 0–2 hours after viral infection (transcription OFF) (f) 10–19 hours after viral infection (transcription ON)
H3 H3
K9 K9 ac
S10
SWI/SNF S10
Transcription H4 K8 H4 K8
factors HDAC
K14 GCN5 K14 ac RNA pol II
K9 Phosphatase ac K9
H3 H3
(d) 3–5 hours after viral infection (transcription OFF) (g) 24 hours after viral infection (transcription OFF)
H3 H3
K9 ac K9
S10 p S10
H4 K8 ac H4 K8
HDAC
GCN5 K14 K14 ac
K14 ac K14
Kinase K8 ac K8
p S10 H4 S10 H4
ac K9 K9
H3 H3
FIGURE 12-20 (a) ChIP and RTPCR analysis of the IFN-β gene. For interactions and enzymatic events that occur over time after viral infection.
ChIP, the intensity of the black bands is proportional to the amount of Purple circles indicate nucleosomes. Dotted purple circles indicate
modified histone or TBP associated with the IFN-β promoter. For RTPCR, nucleosomes altered by chromatin remodeling. Note that transcription is on
the intensity of the black bands indicates the amount of IFN-β mRNA. only 6 to 19 hours after viral infection. [Data from Agalioti et al., “Deciphering
Descriptions in parentheses indicate the functional consequence of the the Transcriptional Histone Acetylation Code for a Human Gene,” Cell 111,
experimental outcome. (b–g) Models that illustrate the sequential molecular 2002, 381–392.]
ELON
INIT
Pol II
FIGURE 12-21 A model describing how the structural features of chromatin regulate gene-specific
transcription in response to signals. Red and green indicate factors that, respectively, activate or
repress transcription, and purple circles indicate nucleosomes.
since these mechanisms are reversible, transcription can be since trimethyl-lysine and acetyl-lysine cannot occur at
rapidly turned on or off in response to developmental and the same time on H3K27, trimethylation by the Polycomb
environmental signals. complex blocks the activating acetylation by the Trithorax
complex. As with other transcription coregulators, target-
KEY CONCEPT IFN-β transcription exemplifies how chroma- ing of Polycomb and Trithorax complexes to chromatin is
tin regulatory strategies are used by cells to alter the transcrip- influenced by transcription factors, histone modifications,
tion of specific genes in response to signals. DNA methylation, and long noncoding RNAs (for exam-
ple, see X-chromosome inactivation later in this section).
12.4 CHROMATIN IN EPIGENETIC KEY CONCEPT Polycomb and Trithorax group proteins
REGULATION work in opposition to maintain the repressed and active
transcription states of parent cells in daughter cells.
LO 12.4 Provide examples of the chromatin-based
mechanisms that maintain gene expression over
cellular or organismal generations. Position-effect variegation
In 1930, Hermann Muller discovered an interesting genetic
When cells divide, information stored in the sequence of phenomenon while studying Drosophila. He found that the
DNA is faithfully replicated and transferred to daugh- expression of genes can be silenced when they are experi-
ter cells. Similarly, information stored in the structure mentally “relocated” to another region of a chromosome.
of chromatin is inherited through cell divisions. This In these experiments, flies were irradiated with X rays to
form of inheritance is given a special name—epigenetic induce mutations in their germ cells, and the progeny of
i nheritance—because it affects the traits of daughter cells the irradiated flies were screened for unusual phenotypes.
without altering DNA sequence. In this section, we describe Among the collection of mutants, Muller found flies with
four examples of the epigenetic control of transcription: eyes that had patches of red and white color. This is unusual
cellular memory, position-effect variegation, genomic because wild-type flies have uniform red eyes, and flies that
imprinting, and X-chromosome inactivation. In each case, are mutant for the white gene, which is required for the
the collection of genes that are transcribed in a parent cell production of red pigment, have uniform white eyes.
is reproduced in daughter cells through the maintenance of Cytological examination revealed a chromosomal rear-
chromatin structure by histone and DNA modifying and rangement in the mutant flies: a region of the X-chromosome
chromatin remodeling mechanisms. containing the white gene was inverted ( Figure 12-22 ).
Inversions and other chromosomal rearrangements will be
Cellular memory discussed in Chapter 17. In this rearrangement, the white
Unlike DNA sequence, chromatin structure can change gene, which is normally located in a euchromatic region of
during the life of a cell, and the changes can be inherited in the X-chromosome, now is near the heterochromatic cen-
successive generations of cell division. Changes in cell fate tromere. The patchy eye phenotype of Muller’s flies is due
are based on short-lived signals that affect the transcrip- to spreading of heterochromatin into the wild-type white
tion of specific genes. Even after the signal goes away, the gene and silencing of white transcription in some cells but
cell fate does not change because the effect on transcription not others. Patches of white tissue in the eye are derived
stays. For instance, once an embryonic cell differentiates from descendants of a single cell in which the white gene is
into an intestinal cell, with its intestinal cell-specific spec- silenced and remains silenced through future cell divisions.
trum of transcriptionally active and inactive genes, it usu- In contrast, red patches arise from cells in which heter-
ally remains an intestinal cell as long as it lives. ochromatin has not spread into the white gene, and so the
Studies that were initially performed in Drosophila white gene remains active in all its descendants.
identified two groups of proteins that function to maintain The existence of red and white patches of cells in the eye
the cellular memory of transcription, Polycomb group pro- of a single organism dramatically illustrates two features
teins and Trithorax group proteins. Polycomb and Tritho- of epigenetic transcription regulation. First, as described
rax proteins often function in opposition to one another, earlier, differences in chromatin structure across chromo-
with Polycomb proteins maintaining genes in a transcrip- somes can be inherited from one cell generation to the next.
tionally repressed state and Trithorax proteins maintaining Second, differences in chromatin structure across chromo-
genes in a transcriptionally active state. Members of the somes affect the expression of resident genes.
Polycomb and Trithorax groups are components of multi- Findings from subsequent studies in Drosophila and
protein complexes that post-translationally modify histones yeast demonstrated that many active genes are silenced in
and remodel chromatin. For example, a Polycomb complex this mosaic fashion when they are relocated to neighbor-
trimethylates H3K27 (a histone modification commonly hoods near centromeres or telomeres that are heterochro-
associated with transcription silencing), while a Trithorax matic. Furthermore, the effect of local chromatin structure
complex acetylates H3K27 (a histone modification com- on transcription is not limited to centromeres and telo-
monly associated with transcription activation). Note that meres. In mouse cells the degree of chromatin compaction
Chromosome
white+
white+ gene
Wild-type eye
expressed
Telomere Centromere
white+
white+ gene
Red cells
expressed
Heterochromatin spreads
at the site of integration of a transgene correlates with the Among more than 300 Drosophila mutants identi-
level of transcription of the transgene and accounts for fied by these screens was Su(var)2-5, which encodes a
about a 1000-fold variation in transcription level across the histone reader protein heterochromatin protein 1 (HP1),
genome. This phenomenon has been called position-effect and Su(var)3-9, which encodes a histone methyltrans-
variegation (PEV). It provides powerful evidence that chro- ferase (Figure 12-24). HP1 contains a chromodomain that
matin structure is able to regulate the expression of genes— binds H3K9me3 and a chromoshadow domain involved
in this case, determining whether genes with identical DNA in dimerization of HP1 proteins and recruitment of a vari-
sequence will be active or silent. ety of chromatin-modifying factors. On the other hand,
Geneticists reasoned that PEV could be exploited to Su(var)3-9 trimethylates H3K9. HP1 and Su(var)3-9
identify the proteins necessary for forming heterochroma- interact with one another to create a feed-forward loop
tin. To this end, they isolated mutations that either sup- that spreads heterochromatin. HP1 binds H3K9me3 and
pressed or enhanced the variegated pattern (Figure 12-23). dimerizes with another HP1 molecule; the dimer recruits
A Suppressor of variegation (Su(var)) is a gene that when Su(var)3-9, which generates H3K9me3; and HP1 binds
mutated reduces the spread of heterochromatin, mean- H3K9me3 to continue the process.
ing that the wild-type product of this gene is required for In the absence of any barriers, heterochromatin might
spreading. In contrast, an Enhancer of variegation (E(var)) spread into adjoining regions and inactivate genes in some
is a gene that when mutated increases the spread of het- cells but not in others. One can imagine that the spread-
erochromatin and normally functions to block spreading. ing of heterochromatin into active gene regions could be
Su(var) and E(var) genes have proved to be a treasure trove disastrous for an organism because active genes would be
for scientists interested in the proteins that are required to silenced as they are converted into heterochromatin. To
establish and maintain the heterochromatic state. avert this potential disaster, boundary/insulator elements,
Spreading enhanced.
More cells with silenced white+.
E(var)
K K K
H3 K K K
9 H3 9 9
9 9 9
which were discussed earlier in the context of topologi- KEY CONCEPT Proteins involved in the spread of hetero
cally associating domains (TADs, Figure 12-12) prevent chromatin include writers, readers, and erasers of histone
the spreading of heterochromatin by creating a local envi- modifications.
ronment that is not favorable to heterochromatin forma-
tion. Insulator-binding proteins may block the spread of
heterochromatin by recruiting activating enzymes such as Genomic imprinting
histone acetyltransferases, H3K4 methyltransferases, and The phenomenon of genom ic i mpr i nt i ng was dis-
SWI/SNF chromatin remodelers, or they may block access covered about 35 years ago in mammals. In genomic
to histones by directly binding them. imprinting, certain autosomal genes are expressed in a
Xist RNA Xi
FIGURE 12-26 RNA fluorescent in situ hybridization (FISH; see silences transcription. [From: J.T. Lee et al., “Lessons from X-chromosome
Chapter 10) for Xist RNA performed on a metaphase chromosome inactivation: long ncRNA as guides and tethers to the epigenome,” Genes
spread of a female fibroblast cell. Xist (labeled with a red fluorescent Dev., 23 (16), 2009, 1831–1842, Fig. 2 © Cold Spring Harbor Laboratory
dye) covers one of the two X chromosomes. DNA (blue) is visualized Press. Photography by Jeannie Lee.]
with DAPI. Binding of Xist is part of the X-inactivation mechanism that
silencing of one of the two X chromosomes in females through the choice of whether to inactivate the maternal or paternal
a process called X-chromosome inactivation or X-inactivation, X chromosome is random, but once an X chromosome is
for short. The inactivated X chromosome, called a Barr inactivated, it will remain inactive for the lifetime of the cell
body, can be seen in the nucleus as a darkly staining, highly and its daughter cells.
condensed, heterochromatic structure (Figure 12-26). This is A 17-kilobase-long noncoding RNA (lncRNA) called
a classic example of epigenetic regulation because the two Xist (X-inactive specific transcript) plays a central role in
X chromosomes in female cells are nearly identical in sequence; initiating silencing of one of the X chromosomes, as does
however, one is transcriptionally active and the other is an antisense transcript Tsix (“Xist” spelled backward) from
silenced by the formation of heterochromatin. In human cells, the same locus (Figure 12-27). Early on in development of the
1 2 3
PRC2
Xa* Xa
Xist
Tsix
4 5
Xi* Xi Xi Xi
Xist PRC2
Xist
YY1
Tsix
Nucleation center
Spreading Spreading
FIGURE 12-27 A model showing how the Xist lncRNA, YY1, and PRC2 act to inactivate one
X chromosome by forming heterochromatin. The five steps illustrate the progression of events that
begin early in embryonic development to silence the transcription of one X chromosome (Xi) and
maintain transcription on the other (Xa). Spreading of Xist across Xa initiates at a nucleation center at
the Xist locus. Xi* and Xa* indicate the future inactive and active X chromosomes, respectively.
embryo, when both female X chromosomes are transcrip- Dosage compensation also takes place in Drosophila, but
tionally active, Tsix is expressed from both alleles. At the in this case, the X chromosome is transcriptionally upregu-
beginning of X-chromosome inactivation, transient pairing lated two-fold in males to equal the transcription of the two
of the X chromosomes represses the transcription of Tsix X chromosomes in females. As in mammals, dosage com-
from one allele, establishing the future inactive X chromo- pensation in Drosophila involves lncRNAs (roX1 and roX2)
some (Xi). Transcription that persists from the other allele that are transcribed from the X chromosome in males as well
blocks activation of Xist transcription and this establishes as alterations in chromatin structure that include acetyla-
the future active X chromosome (Xa). Xist lncRNA spreads tion of H4K16, phosphorylation of H3S10, and nucleosome
along the future Xi and induces silencing. Spreading is remodeling by a chromatin remodeling complex. Twofold
nucleated at the Xist locus by the YY1 (Yin-Yang 1) protein, transcription upregulation of X-chromosome genes in males
which binds both Xist lncRNA and DNA. As it spreads, Xist is probably achieved by precisely balancing activities that
recruits PRC2 (Polycomb Repressive Complex 2), which condense and decondense the X chromosome.
catalyzes the heterochromatin-associated modification
H2K27me3. Other lncRNAs and structural changes in chro- KEY CONCEPT In X-inactivation, epigenetic mechanisms
matin occur to establish and/or maintain X-chromosome enacted early in embryonic development silence an entire
inactivation, including H3K9 methylation, histone deacetyl- chromosome.
ation, DNA CpG island methylation, and incorporation of
the histone H2A variant macroH2A into nucleosomes.
SUMMARY
This chapter has focused on the roles that transcription of these activities have affinity for specific genes because of
factors and chromatin structure play in directing cell type- interactions with transcription factors and chromatin mod-
specific transcription of eukaryotic genes. Transcription ifications themselves. Histone modifications are concen-
factors bind distal and proximal enhancers and alter tran- trated in the flexible tails of core histones. There are many
scription by recruiting coactivators and corepressors. A different types of modifications, including acetylation,
single transcription factor such as Gal4 can affect the tran- methylation, and phosphorylation, and an enormous num-
scription of multiple genes by binding an enhancer shared ber of different combinations of modifications can occur
by the genes. In addition, as demonstrated by MCM1 in on single histone tails and within a histone octamer. Thus,
mating-type specification, one transcription factor can histone modifications have the capacity to convey complex
affect the activation/repression activity and transcriptional instructions, dubbed the histone code, for how the tran-
targets of other transcription factors. Transcription factors scription machinery should operate. Similarly, DNA modifi-
are spurred into action by environment signals such as the cations, mainly cytosine methylation in CpG dinucleotides,
presence of galactose, hormones, and viruses or develop- provide an additional set of instructions to the transcription
mental signals such as those that specify the intestinal cell machinery in some eukaryotic organisms, including plants
fate. Furthermore, cells control transcription by regulating and mammals. Histone and DNA modifications control
the expression, cellular localization, stability, and activity recruitment of transcription initiation factors such as TFIID
(DNA binding, dimerization, ligand-binding, and interac- and elongation factors such as P-TEFb to specific genes. In
tions) of transcription factors. addition, they recruit ATP-dependent chromatin remodeling
In eukaryotic cells, chromatin, not naked DNA, is the complexes such as SWI/SNF that change factor accessibil-
substrate for transcription factors, general transcription ity to DNA. Chromatin remodeling complexes make DNA
factors, and RNA polymerase II. The wrapping of DNA more or less accessible by sliding, ejecting, replacing, or
around an octamer of core histones (H2A, H2B, H3, and exchanging histones on DNA. Studies such as those of the
H4) and the binding of linker histones (H1) to form nucleo- IFN-β gene demonstrate how transcription factors, chro-
somes as well as the gathering of nucleosomes into compact matin modifying factors, and chromatin remodeling factors
structures is generally repressive to transcription. Reversible function cooperatively to adjust the level of transcription of
chromatin modifying and chromatin remodeling activities particular genes in response to a signal. In the end, the level
work together to overcome the repressive effects of chroma- of transcription is determined by the balance of activating
tin by changing its structure. and repressive mechanisms that act on a gene.
Chromatin modification encompasses modifications DNA replication faithfully copies both DNA sequence
to histone amino acids and DNA nucleotides. Modifica- and chromatin structure from parent to daughter cells.
tion systems are made up of coactivator and corepressor Newly formed cells inherit both genetic information, inher-
proteins that add (write), remove (erase), and bind (read) ent in the nucleotide sequence of DNA, and epigenetic infor-
modifications. Protein complexes containing one or several mation, built into histone and DNA modifications. Cellular
memory, position-effect variegation, genomic imprinting, group factors, suppressor of variegation ( Su(var) ) and
and X-chromosome inactivation are examples of epigenetic enhancer of variegation (E(var)) factors, insulator elements,
phenomenon where the transcription state of single genes, and long noncoding RNAs (lncRNAs). Thus, the nucleotide
multiple genes, and even whole chromosomes is inherited sequence of genomes is not sufficient for understanding the
without changing the sequence of DNA. Epigenetic inher- inheritance of normal and disease states of transcription.
itance mechanisms involve Polycomb group and Trithorax
KEY TERMS
activation domain (p. 401) enhancer (p. 400) linker histone (p. 407)
Barr body (p. 422) epigenetic inheritance (p. 418) maternal imprinting (p. 421)
boundary (p. 410) euchromatin (p. 408) nucleosome (p. 407)
canonical histone (p. 407) facultative heterochromatin (p. 408) nucleosome free region (NFR)
chromatin (p. 406) flexible tail (p. 407) (p. 409)
chromatin modification (p. 410) genomic imprinting (p. 420) paternal imprinting (p. 421)
chromatin remodeling (p. 410) general transcription factor position-effect variegation (PEV)
coactivator (p. 400) (GTF) (p. 400) (p. 419)
constitutive heterochromatin (p. 408) heterochromatin (p. 408) post-translational modification (p. 410)
core histone (p. 407) histone (p. 407) proximal enhancer (p. 400)
core promoter (p. 400) histone acetyltransferase (HAT) (p. 410) proximal promoter (p. 400)
corepressor (p. 400) histone code (p. 411) repression domain (p. 401)
CpG island (p. 413) histone deacetylase (HDAC) (p. 410) topologically associating domain
dimerization domain (p. 401) histone fold (p. 407) (TAD) (p. 409)
distal enhancer (p. 400) histone-fold extension (p. 407) transcription factor (p. 400)
DNA-binding domain (p. 401) histone octamer (p. 407) upstream activation sequence (UAS)
DNA modification (p. 412) histone modification (p. 411) (p. 402)
dosage compensation (p. 421) insulator (p. 410) variant histone (p. 407)
enhanceosome (p. 415) ligand-binding domain (p. 402) X-chromosome inactivation (p. 422)
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
WORKING WITH THE FIGURES 7. In Figure 12-7, what effect would a Gal4 protein that
(The first 27 questions require inspection of text figures.) lacks the DNA-binding domain have on transcription
of the UAS-lacZ reporter gene, and why?
1. In Figure 12-1, name the cis-acting sequence elements
and trans-acting proteins that regulate transcription. 8. In Figure 12-8, is Gal3 a transcription factor, coactiva-
tor, corepressor, or none of these?
2. In Figure 12-2, how might a corepressor block tran-
scription by RNA polymerase II? 9. In Figure 12-9, hypothesize why MCM1 does not bind
and activate α -specific genes in a cells and a /α cells.
3. In Figure 12-3, what proteins bind the GGCCAATC
and the ATATAA sequences? 10. In Figure 12-10, several protein subunits of the TFIID gen-
eral transcription factor contain a histone-fold domain.
4. In Figure 12-4, what is the function of each transcrip-
Based on the function of the histone fold in histones, pro-
tion factor domain?
pose a function for the histone fold in TFIID proteins.
5. In Figure 12-5, how does Gal4 regulate the transcrip-
11. In Figure 12-11, how might the structure of chromatin
tion of four different GAL genes at the same time?
bring enhancer and promoter elements close together
6. In Figure 12-6, how many individual Gal4 proteins that are far apart in linear DNA?
can bind the DNA between the GAL10 and GAL1
12. In Figure 12-12, what features of chromatin structure
genes?
are shared between enhancers and promoters?
13. In Figure 12-13, what effect might reduced acetyl-CoA 30. The Lugar and Richmond crystal structure of the
levels have on transcription? nucleosome used Xenopus laevis (toad) histones. Why
14. Lysines in histone tails can be propionylated. The propi- is the structure thought to be a good representation of
onyl group is similar in structure to an acetyl group. Using human nucleosomes?
the categories shown in Figure 12-14, what would you call 31. Why are histone tails not visible in the crystal struc-
enzymes that regulate propionyl addition and removal? ture of the nucleosome core particle?
15. In Figure 12-15, what are the implications to the his- 32. How might higher-order structures of chromatin acti-
tone code of the different lysine and arginine methyla- vate, rather than repress, transcription?
tion types?
33. What are the two general mechanisms by which his-
16. I n Fi gure 12-16 , how m any d iffer en t co d es tone acetylation affects transcription?
could be produced on the histone H3 tail just by
34. What functions might be served by modifications of
phosphorylation?
amino acids in the histone-fold domain?
17. Based on Figure 12-17, what sequence would be read
35. How is the function of histone tails similar to that of
after the sodium bisulfite reaction, if all of the CpGs
the C-terminal domain (CTD) of RNA polymerase II?
in the sequence 5′-GGCGTCGAAGTCGAA-3′ were
methylated? 36. By what two mechanisms could histone acetylation
levels increase at a gene promoter?
18. In Figure 12-18, how might a CpG island function dif-
ferently than an isolated CpG? 37. Explain how phosphorylation of histone H3 serine 10
(H3S10P) might increase acetylation of histone H4
19. In Figure 12-19, what steps would need to occur to
lysine 16 (H4K16ac)?
exchange a variant H2A for a canonical H2A in a
nucleosome? 38. Vertebrate histone H1 can be phosphorylated on many
amino acids in the C-terminal unstructured domain.
20. In Figure 12-20 , describe two ways in which the
What effect would you expect histone H1 phosphory-
HDAC might be recruited to the IFN-β gene.
lation to have on chromatin structure?
21. In Figure 12-21a, which of the transcription instructions,
in the form of binding sites in chromatin, are reversible? 39. What type of factors would you expect to be involved
in the regulation of histone H1 phosphorylation?
22. In Figure 12-22, will all flies with a white gene inver-
sion have the same pattern of white and red cells as the 40. What molecular interactions must by broken by chro-
eye shown at the bottom? Why or why not? matin remodeling complexes to remove a histone oc-
tamer from DNA?
23. In Figure 12-23, name a type of gene that might be an
E(var) and explain your answer. 41. Why is the order of assembly and disassembly of
nucleosomes important for understanding transcrip-
24. In Figure 12-24, how is this mechanism similar to the tion regulation?
mechanism by which transcription factors regulate
transcription (for example, as in Figure 12-2)? 42. Why might insertion of a transgene at different places
in the Drosophila genome cause the transgene to be
25. In Figure 12-25, what mechanisms might position the transcribed at different levels?
enhancer of the paternal allele to act on the Igf2 pro-
moter that is >50 kilobases away, and why might this 43. How would you modify a transgene so that its expres-
not happen for the maternal allele? sion was not affected by position-effect variegation
(PEV)?
26. In Figure 12-26, why is it specified in the figure legend
that this is a female cell? 44. What purpose might be served by the long half-life of
core histones?
27. In Figure 12-27, what histone modification is expected
to be enriched on the inactive X chromosome relative 45. What is meant by the term epigenetic inheritance?
to the active X chromosome, and why? Describe two examples of such inheritance.
46. Give three functions of insulator elements.
BASIC PROBLEMS
47. How many nucleosomes would be needed to cover the
28. Do all nucleosomes have the same eight core histones? human genome (3 × 109 base pairs), if the average linker
Why or why not? distance between nucleosomes was 50 base pairs?
29. Why might binding of a transcription factor to DNA 48. Why might the concentration of ATP in cells affect the
be inhibited for DNA that is part of a nucleosome? structure of chromatin?
O
f all the phenomena in biology, few if any inspire Homeotic mutants of Drosophila melanogaster
more awe than the formation of a complex animal
from a single-celled egg. In this spectacular trans- (a)
formation, unseen forces organize the dividing mass of cells
into a form with a distinct head and tail, various append-
ages, and many organs. The great geneticist Thomas Hunt
Morgan was not immune to its aesthetic appeal:
A transparent egg as it develops is one of the most
fascinating objects in the world of living beings.
The continuous change in form that takes place
(b)
from hour to hour puzzles us by its very simplicity.
The geometric patterns that present themselves
at every turn invite mathematical analysis. . . .
This pageant makes an irresistible appeal to the
emotional and artistic sides of our nature.1
Yet, for all its beauty and fascination, biologists were
stumped for many decades concerning how biological form
is generated during development. Morgan also said that
(c)
“if the mystery that surrounds embryology is ever to come
within our comprehension, we must . . . have recourse to
other means than description of the passing show.”
The long drought in embryology lasted well beyond
Morgan’s heyday in the 1910s and 1920s, but it was even-
tually broken by geneticists working very much in the
tradition of Morgan-style genetics and with his favorite,
most productive genetic model, the fruit fly Drosophila
FIGURE 13-1 In homeotic mutants, the identity of one body
melanogaster. structure has been changed into another. (a) Normal fly with one pair of
The key catalysts to understanding the making of animal forewings on the second thoracic segment and one pair of small hind
forms were the discoveries of genetic “monsters”—mutant wings on the third thoracic segment. (b) Mutations in the Ultrabithorax
fruit flies with dramatic alterations of body structures gene lead to loss of Ubx function in the posterior thorax, which
(Figure 13-1). In the early days of Drosophila genetics, rare causes the development of forewings in place of the hind wings.
(c) Antennapedia mutant in which the antennae are transformed into
mutants arose spontaneously or as by-products of other legs. [Sean Carroll.]
experiments with spectacular transformations of body parts.
In 1915, Calvin Bridges, then Morgan’s student, isolated a
fly having a mutation that caused the tiny hind wings (hal-
Antennapedia mutant in which legs develop in place of the
teres) of the fruit fly to resemble the large forewings. He
antennae (Figure 13-1c).
dubbed the mutant bithorax. The transformation in bitho-
The spectacular effects of homeotic mutants inspired
rax mutants is called homeotic (Greek homeos, meaning
what would become a revolution in embryology, once the
same or similar) because one part of the body (the hind
tools of molecular biology became available to understand
wing) is transformed to resemble another (the forewing), as
what homeotic genes encoded and how they exerted such
shown in Figure 13-1b. Subsequently, several more homeotic
enormous influence on the development of entire body
mutants were identified in Drosophila, such as the dramatic
parts. Surprisingly, these strange fruit-fly genes turned out
to be a passport to the study of the entire animal kingdom,
1
T. H. Morgan, Experimental Embryology. Columbia University Press, as counterparts to these genes were discovered that played
1927. similar roles in almost all animals. Furthermore, the same
428
regulatory logic that underpins development in animals is normal digits (Figure 13-2b). These transplanted regions
also used to control development in plants. of the amphibian embryo and chick limb bud were termed
The study of animal and plant development is a very organizers because of their remarkable ability to organize
large and still-growing discipline. As such, we do not the development of surrounding tissues. The cells in the
attempt a comprehensive overview. Rather, in this chapter, organizers were postulated to produce morphogens, mol-
we will focus on a few general concepts that illustrate the ecules that induced various responses in surrounding tissue
logic of the genetic control of animal development. We will in a concentration-dependent manner.
explore how the information for building complex struc-
tures is encoded in the genome. In contrast to the control KEY CONCEPT Organizers are groups of cells in an embryo
of gene regulation in single bacterial or eukaryotic cells, the that have the remarkable ability to instruct the development
genetic control of body formation and body patterning is of other cells in an embryo via the production of morphogens,
which are molecules that act in a concentration-dependent
fundamentally a matter of gene regulation in three-dimen-
manner. Cells in close proximity to the organizer are exposed
sional space and over time. Yet we will see that the prin-
to high concentrations of morphogens and therefore develop
ciples governing the genetic control of development are into different structures from cells located further from the
connected to those already presented in Chapters 11 and organizer.
12, governing the physiological control of gene expression
in bacteria and single-celled eukaryotes.
be mutated regardless of the amount of product made by To address these questions, strategies had to be devised
a gene. And, third, the genetic approach can uncover phe- to identify, catalog, and analyze genes that control develop-
nomena for which there is no biochemical or other bioassay. ment. One of the first considerations in the genetic analysis
From the genetic viewpoint, there are four key ques- of animal development was which animal to study. Of the
tions concerning the number, identity, and function of genes millions of living species, which offered the most promise?
taking part in development: The fruit fly Drosophila melanogaster emerged as the lead-
1. Which genes are important in development? ing genetic model of animal development because its ease
of rearing, rapid life cycle, cytogenetics, and decades of
2. Where in the developing organism and at what times
classical genetic analysis (including the isolation of many
are these genes active?
very dramatic mutants) provided important experimental
3. How is the expression of developmental genes regu- advantages (see the Model Organism box on Drosophila
lated? melanogaster above). The nematode worm Caenorhabditis
4. Through what molecular mechanisms do gene products elegans also presented many attractive features, most par-
affect development? ticularly its simple construction and well-studied cell
430
Adult
Eclosion Larva
Imaginal disks
4 days
1st instar (1 day)
3rd instar
Overview of Drosophila development. The larva forms in 1 day and then undergoes A N I M ATED A RT
several stages of growth during which the imaginal disks and other precursors of adult
structures proliferate. These structures differentiate during pupation, and the adult fly Drosophila embryonic
Drosophila Model organism box
hatches (eclosion) and begins the cycle again. development
Using Knowledge from One Model strands without disrupting the proper hydrogen bonding of
nearby base pairs. Some of these treasure hunts were carried
Organism to Fast-Track Developmental
out in the Drosophila genome itself, in looking for more family
Gene Discovery in Others members. Others searched for similar genes in other animals,
With the discovery of genes that regulate development within the by means of zoo blots (Southern blots of restriction-enzyme-
Drosophila genome, similarities among the DNA sequences of digested DNA from different animals), by using radioactive Dro-
these genes could be exploited in treasure hunts for other mem- sophila DNA as the probe (see Chapter 10). This approach led to
bers of the gene family. These hunts depend on DNA base-pair the discovery of homologous gene sequences in many different
complementarity. For this purpose, DNA hybridizations were animals, including humans and mice. Now homologous genes
carried out under moderate stringency conditions, in which there are typically identified by computational searches of genome
could be some mismatch of bases between the hybridizing sequences (see Chapter 14).
lineages (see the Model Organism box on Caenorhabditis toolkit—the set of genes that control the development of
elegans on page 451). Among vertebrates, the development the bodies, body parts, and cell types of several different
of targeted gene disruption techniques opened up the lab- animal species—has been defined. We will first focus on the
oratory mouse Mus musculus to more systematic genetic genetic toolkit of Drosophila melanogaster because its iden-
study, and the zebrafish Danio rerio has recently become tification was a source of major insights into the genetic
a favorite model owing to the transparency of the embryo control of development; its discovery catalyzed the identi-
and to advances in its genetic study. Among plants, Ara- fication of the genetic toolkit of other animals, including
bidopsis thaliana has played a similar role as Drosophila humans.
in illuminating fundamental mechanisms in plant develop-
ment. More information about the most common model KEY CONCEPT Genetic model organisms, particularly Dro-
organisms can be found in “A Brief Guide to Model Organ- sophila melanogaster, have played a key role in the identifica-
tion of the genetic toolkit for development. Remarkably, many
isms” at the end of this book.
of the toolkit genes discovered in model organisms play funda-
Through systematic and targeted genetic analysis, as
mental roles in human development and disease.
well as comparative genomic studies, much of the genetic
431
13.2 THE GENETIC TOOLKIT FOR cause some discrete defect in either the embryonic or the
adult body pattern, or both. It has proved useful to group
DROSOPHILA DEVELOPMENT the genes affected by mutations into several categories
based on the nature of their mutant phenotypes. Many
LO 13.1 Outline experimental approaches to identify and toolkit genes can be classified according to their function
characterize members of the genetic toolkit for
development in different animal phyla. in controlling the identity of body parts (for example, of
different segments or appendages), the formation of body
LO 13.2 Differentiate members of the genetic toolkit for
development from other genes. parts (for example, of organs or appendages), the number
of body parts, the formation of cell types, and the organiza-
tion of the primary body axes (the anteroposterior, or A–P,
Animal genomes typically contain about 13,000 to 22,000
and dorsoventral, or D–V, axes; see the Model Organism
genes. Many of these genes encode proteins that function in
Box on page 430).
essential processes in all cells of the body (for example, in
We will begin our inventory of the Drosophila toolkit
cellular metabolism or the biosynthesis of macromolecules).
by examining the genes that control the identity of seg-
Such genes are often referred to as housekeeping genes.
ments and appendages. We do so for both historical and
Other genes encode proteins that carry out the specialized
conceptual purposes. The genes controlling segmental and
tasks of various organ systems, tissues, and cells of the body
appendage identity were among the very first toolkit genes
such as the globin proteins in oxygen transport or antibody
identified. Subsequent discoveries about their nature were
proteins that mediate immunity. Here, we are interested in a
sources of profound insights into not just how their prod-
different set of genes, those concerned with the building of
ucts work, but also the content and workings of the tool-
organs and tissues and the specification of cell types—the
kits of most animals. Furthermore, their spectacular mutant
genetic toolkit for development that determines the overall
phenotypes indicate that they are among the most globally
body plan and the number, identity, and pattern of body
acting genes that affect animal form.
parts.
Toolkit genes of the fruit fly have generally been iden-
tified through the monstrosities or catastrophes that arise Homeotic genes and segmental
when they are mutated. Toolkit-gene mutations from two identity
sources have yielded most of our knowledge. The first
source consists of spontaneous mutations that arise in lab- Among the most fascinating abnormalities to be described
oratory populations, such as those found in the Morgan in animals are those in which one normal body part is
lab. The second source comprises mutations induced at replaced by another. Such homeotic transformations have
random by treatment with mutagens (such as chemicals or been observed in many species in nature, including sawflies
radiation) that greatly increase the frequency of damaged in which a leg forms in place of an antenna and frogs in
genes throughout the genome. Elegant refinements of the which a thoracic vertebra forms in place of a cervical ver-
latter approach have made possible systematic searches, tebra (Figure 13-3). Whereas only one member of a bilateral
called genetic screens, in which organisms are treated with pair of structures is commonly altered in many naturally
a mutagen and allowed to reproduce, and then the off- occurring variants, both members of a bilateral pair of
spring are examined for visible defects in a phenotype of structures are altered in homeotic mutants of fruit flies (see
interest. Such screens have identified many members of the Figure 13-1). In the former case, the alteration is not her-
fly’s genetic toolkit. The members of this toolkit constitute itable, but homeotic mutants breed true from generation to
only a small fraction, perhaps several hundred genes, of the generation.
roughly 14,000 genes in the fly genome. The scientific fascination with homeotic mutants stems
from three properties. First, it is amazing that a single gene
KEY CONCEPT The genetic toolkit for animal development is mutation can alter a developmental pathway so dramati-
composed of a small fraction of all genes. Only a small subset cally. Second, it is striking that the structure formed in the
of the entire complement of genes in the genome affect devel- mutant is a well-developed likeness of another body part.
opment in discrete ways. And, third, it is important to note that homeotic mutations
transform the identity of serially reiterated structures.
Insect and many animal bodies are made of repeating parts
of similar structure, like building blocks, arranged in a
Classification of genes by developmental series. The forewings and hind wings, the segments, and the
function antennae, legs, and mouthparts of insects are sets of serially
One of the first tasks following the execution of a genetic reiterated body parts. Homeotic mutations transform iden-
screen for mutations is to sort out those of interest. Many tities within these sets.
mutations are lethal when hemi- or homozygous because A mutation may cause a loss of homeotic gene func-
cells cannot survive without products affected by these tion where the gene normally acts, or it may cause a
mutations. The more interesting mutations are those that gain of homeotic function where the homeotic gene does
(a) (b)
FIGURE 13-3 A late-nineteenth-century drawing from one of the first studies of homeotic
transformations in nature. (a) Homeosis in a sawfly, with the left antenna transformed into a leg.
(b) Homeosis in a frog. The middle specimen is normal. The specimen on the left has extra structures
growing out of the top of the vertebral column, transforming a cervical vertebra into a thoracic vertebra
(red arrow). The specimen on the right has an extra set of vertebrae (red arrow). [From W. Bateson,
Material for the Study of Variation. Macmillan, 1894.]
not normally act. For example, the Ultrabithorax (Ubx) Organization and expression of Hox genes
gene acts in the developing hind wing to promote hind-
A most intriguing feature of Hox genes is that they are
wing development and to repress forewing development.
clustered together in two gene complexes that are located
L oss-of-function mutations in Ubx transform the hind
on the third chromosome of Drosophila. The Bithorax
wing into a forewing (see Figure 13-1). Dominant gain-of-
complex contains three Hox genes, and the Antennapedia
function mutations in Ubx transform the forewing into a
complex contains five Hox genes. Moreover, the order of
hind wing. Similarly, the antenna-to-leg transformations of
the genes in the complexes and on the chromosome corre-
Antennapedia (Antp) mutants are caused by the dominant
sponds to the order of body regions, from head to tail, that
gain of Antp function in the antenna (see Figure 13-1). In
are influenced by each Hox gene (Figure 13-4).
addition to these transformations in appendage identity,
The relation between the structure of the Hox-gene
homeotic mutations can transform segment identity, caus-
complexes and the phenotypes of Hox-gene mutants was
ing one body segment of the adult or larva to resemble
illuminated by the molecular characterization of the genes.
another.
Molecular cloning of the sequences encompassing each Hox
Although homeotic genes were first identified through
locus provided the means to analyze where in the develop-
spontaneous mutations affecting adult flies, they are
ing animal each gene is expressed. These spatial aspects of
required throughout most of a fly’s development. System-
gene expression and gene regulation are crucial to under-
atic searches for homeotic genes have led to the identifica-
standing the logic of the genetic control of development. In
tion of eight loci, now referred to as Hox genes, that affect
regard to the Hox genes and other toolkit genes, the devel-
the identity of segments and their associated appendages
opment of technology that made possible the visualization
in Drosophila. Generally, the complete loss of any Hox-
of gene and protein expression was crucial to understand-
gene function is lethal in early development. The dominant
ing the relation among gene organization, gene function,
mutations that transform adults are viable in heterozygotes
and mutant phenotypes.
because the wild-type allele provides normal gene function
Two principal technologies for the visualization of
to the developing animal.
gene expression in embryos or other tissues are (1) the
Hox genes regulate the identity of body parts the dark blue shading in Figure 13-6 indicates where the
Ubx gene is expressed. This Hox gene is expressed in the
posterior thoracic and most of the abdominal segments of
the embryo. The development of these segments is altered in
Ubx mutants. Ubx is also expressed in the developing hind
wing but not in the developing forewing ( Figure 13-7), as
one would expect knowing that Ubx promotes hind-wing
development and represses forewing development in this
appendage.
Add enzyme
substrate.
FIGURE 13-5 The two principal technologies
for visualizing where a gene is transcribed Visualize mRNA Visualize protein
or where the protein that it encodes is expression in expression in
expressed are (left) in situ hybridization of light microscope. fluorescence
microscope.
complementary RNA probe to mRNA and
(right) immunolocalization of protein expression.
The procedures for each method are outlined.
Expression patterns may be visualized as
the product of an enzymatic reaction or of a
chromogenic substrate or with fluorescently
labeled compounds.
Hox genes are expressed in spatially restricted function of the proteins that they encode are of special
domains interest. Edward Lewis, a pioneer in the study of homeotic
genes, noted early on that the clustering of Bithorax com-
(a)
plex genes suggested that the multiple loci had arisen by
tandem duplication of an ancestral gene. This idea led
researchers to search for similarities in the DNA sequences
of Hox genes. They found that all eight Hox genes of the
two complexes have a short region of sequence similarity,
180 bp in length. Because this stretch of DNA sequence
similarity is present in homeotic genes, it was dubbed the
Head Thorax Abdomen homeobox. The homeobox encodes a protein domain, the
homeodomain, containing 60 amino acids. The amino acid
(b)
sequence of the homeodomain is very similar among the
Hox proteins (Figure 13-8).
Although the discovery of a common protein motif in
each of the Hox proteins was very exciting, further anal-
ysis of the structure of the homeodomain revealed that
it forms a helix-turn-helix motif—the structure common
to the Lac repressor, the λ repressor, Cro, and the α2 and
a1 regulatory proteins of the yeast mating-type loci! This
similarity suggested immediately (and it was subsequently
FIGURE 13-6 Expression of Hox genes in the Drosophila embryo.
(a) Schematic representation of Drosophila embryo showing regions borne out) that Hox proteins are sequence-specific
where eight individual Hox genes are expressed. (b) Actual image of DNA-binding proteins and that they exert their effects
the expression of seven Hox genes visualized by in situ hybridization. by controlling the expression of genes within developing
Colors indicate expression of labial (turquoise), Deformed (lavender), segments and appendages. Thus, the products of these
Sex combs reduced (green), Antennapedia (orange), Ultrabithorax
remarkable genes function through principles that are
(dark blue), Abdominal-A (red), and Abdominal-B (yellow). The embryo
is folded so that the posterior end (yellow) appears near the top center. already familiar from Chapters 11 and 12—by binding to
[(b) Dave Kosman, Ethan Bier, and Bill McGinnis.] regulatory elements of other genes to activate or repress
their expression. We will see that it is also true of many
limbs also can form, but they have antennal identity; and, other toolkit genes: a significant fraction of these genes
similarly, wings can form, but they have forewing identity. encode transcription factors that control the expression
Other genes control the formation of segments, limbs, and of other genes.
wings and will be described later. First, we must under-
stand how Hox genes exert their dramatic effects on fly KEY CONCEPT Homeotic transformations result from
development. mutations in Hox genes, which are genes that contain a
conserved sequence called the homeobox. This sequence
The homeobox encodes a protein domain called the homeodomain, which is
similar to the helix-turn-helix motif found in many other tran-
Because Hox genes have large effects on the identities of
scription factors.
entire segments and other body structures, the nature and
Consensus -RRGRT-YTR-QTLELEKEFHFNRYLTRRRRIEIAHALCLTERQIKIWFQNRRMK-KKE-
sequence Helix 1 Helix 2 Helix 3
FIGURE 13-8 Sequences of fly homeodomains. All eight Drosophila Hox genes encode proteins
containing a highly conserved 60-amino-acid domain, the homeodomain, composed of three α helices.
Helices 2 and 3 form a helix-turn-helix motif similarly to the Lac repressor, Cro, and other DNA-binding
proteins. Residues common to the Hox genes are shaded in yellow; divergent residues are shaded in
red; those common to subsets of proteins are shaded in blue or green. [Data from S. B. Carroll,
J. K. Grenier, and S. D. Weatherbee, From DNA to Diversity: Molecular Genetics and the Evolution of Animal
Design 2e, Blackwell, 2005.]
We will examine how Hox proteins and other toolkit since their last common ancestor, the extent of sequence
proteins orchestrate gene expression in development a little similarity indicates very strong pressure to maintain the
later. First, there is one more huge discovery to describe, sequence of the homeodomain.
which revealed that what we learn from fly Hox genes has The existence of Hox genes with homeoboxes through-
very general implications for the animal kingdom. out the animal kingdom was entirely unexpected. Why dif-
ferent types of animals would possess the same regulatory
genes was not obvious, which is why biologists were further
Clusters of Hox genes control development surprised by the results when the organization and expres-
in most animals sion of Hox genes was examined in other animals. In ver-
When the homeobox was discovered in fly Hox genes, it tebrates, such as the laboratory mouse, the Hox genes also
raised the question whether this feature was some peculiar- are clustered together in four large gene complexes on four
ity of these bizarre fly genes or was more widely distrib- different chromosomes. Furthermore, the order of the genes
uted, in other insects or segmented animals, for example. in the mouse Hox complexes parallels the order of their
To address this possibility, researchers searched for homeo- most related counterparts in the fly Hox complexes, as well
boxes in the genomes of other insects, as well as earth- as in each of the other mouse Hox clusters (Figure 13-10a).
worms, frogs, cows, and even humans. They found many This correspondence indicates that the Hox complexes of
homeoboxes in each of these animal genomes. insects and vertebrates are related and that some form of
The similarities in the homeobox sequences from dif- Hox complex existed in their distant common ancestor. The
ferent species were astounding. Over the 60 amino acids of four Hox complexes in the mouse arose by duplications of
the homeodomain, some mouse and fish Hox proteins were entire Hox complexes (perhaps of entire chromosomes) in
identical with the fly sequences at as many as 54 of the 60 vertebrate ancestors.
positions (Figure 13-9). In light of the vast evolutionary dis- Why would such different animals have these sets of
tances between these animals, more than 500 million years genes in common? Their deep, common ancestry indicates
FIGURE 13-9 The sequences of the Drosophila Deformed protein homeodomain and of several
members of the vertebrate Hox group 4 genes are strikingly similar. Residues in common are shaded
in yellow; divergent residues are shaded in red; residues common to subsets of proteins are shaded
in blue. The very similar C-terminal flanking regions outside of the homeodomain are shaded in green.
[Data from S. B. Carroll, J. K. Grenier, and S. D. Weatherbee, From DNA to Diversity: Molecular Genetics and
the Evolution of Animal Design 2e, Blackwell, 2005.]
Mouse embryo
that Hox genes play some fundamental role in the develop- The Hox-gene expression patterns of vertebrates sug-
ment of most animals. That role is apparent from analyses gested that they also specify the identity of body regions,
of how the Hox genes are expressed in different animals. In and subsequent analyses of Hox-gene mutants have borne
vertebrate embryos, adjacent Hox genes also are expressed this suggestion out. For example, mutations in the Hoxa11
in adjacent or partly overlapping domains along the antero- and Hoxd11 genes cause the homeotic transformation of
posterior body axis. Furthermore, the order of the Hox sacral vertebrae to lumbar vertebrae ( Figure 13-11). Thus,
genes in the complexes corresponds to the head-to-tail as in the fly, the loss or gain of function of Hox genes in
order of body regions in which the genes are expressed vertebrates causes transformation of the identity of seri-
(Figure 13-10b). ally repeated structures. Such results have been obtained in
several classes, including mammals, birds, amphibians, and provided by the female to
fish. Furthermore, clusters of Hox genes have been shown the egg are called maternal-
to govern the patterning of other insects and to be deployed effect genes. Mutant phe-
in regions along the anteroposterior axis in annelids, mol- notypes of strict maternal-
luscs, nematodes, various arthropods, primitive chordates, effect genes depend only on
flatworms, and other animals. Therefore, despite enor- the genotype of the mother
mous differences in anatomy, the possession of one or more (Figure 13-12, top).
clusters of Hox genes that are deployed in regions along In these screens, genes
the main body axis is a common, fundamental feature of were identified that were
at least all bilateral animals. Indeed, the surprising lessons necessary to make the
from the Hox genes portended what turned out to be a gen- proper number and pat-
eral trend among toolkit genes; that is, most toolkit genes tern of larval segments, to
are common to different animals. make its three tissue layers
(ectoderm, mesoderm, and
KEY CONCEPT Despite great differences in anatomy, a endoderm), and to pattern
broad array of different animal phyla have many toolkit genes the fine details of an ani-
in common. mal’s anatomy. The power
of the genetic screens was
their systematic nature. By Christiane Nüsslein-Volhard and
Now let’s take an inventory of the rest of the toolkit to saturating each of a fly’s Eric Wieschaus at the European
see what other general principles emerge. chromosomes (except the Molecular Biology Laboratory.
[Christiane Nüsslein-Volhard.]
small fourth chromosome)
with chemically induced mutations, the researchers were
13.3 DEFINING THE ENTIRE able to identify most genes that were required for the build-
TOOLKIT ing of the fly. For their pioneering efforts, Nüsslein-Volhard,
Wieschaus, and Lewis shared the 1995 Nobel Prize in Phys-
LO 13.1 Outline experimental approaches to identify and iology or Medicine.
characterize members of the genetic toolkit for The most striking and telling features of the newly iden-
development in different animal phyla.
tified mutants were that they showed dramatic but discrete
LO 13.2 Differentiate members of the genetic toolkit for defects in embryo organization or patterning. That is, the
development from other genes.
dead larva was not an amorphous carcass but exhibited
LO 13.3 Predict both the phenotypic effects of mutations
specific, often striking patterning defects. The Drosophila
in toolkit genes based on their expression during
development as well as the expression patterns of larval body has various features whose number, position,
toolkit genes based on the phenotypic effects of or pattern can serve as landmarks to diagnose or classify
mutations in toolkit genes.
the abnormalities in mutant animals. Each locus could thus Bicoid mutants are missing the anterior region
be classified according to the body axis that it affected and
the pattern of defects caused by mutations. Each class of
genes appeared to represent different steps in the progressive
refinement of the embryonic body plan—from those that
affect large regions of the embryo to those with more limited
realms of influence.
FIGURE 13-14 Classes of Drosophila segmentation-gene mutants. These diagrams depict A N I M ATED A RT
representative gap, pair-rule, and segment-polarity mutants. The red trapezoids are the dense bands
of exoskeleton seen in Figure 13-13. The boundary of each segment is indicated by a dotted line. Drosophila embryonic
The left-hand diagram of each pair depicts a wild-type larva, and the right-hand diagram depicts the development
pattern formed in a given mutant. The shaded light orange regions on the wild-type diagrams indicate
the domains of the larva that are missing or affected in the mutant.
stripes correspond to the periodicity of defects in mutant making of the body plan is a step-by-step process, with
larvae), as shown in Figure 13-15c. Many segment-polarity major subdivisions of the body outlined first and then
genes are expressed in stripes of cells within each segment, refined until a fine-grain pattern is established. The order
14 stripes in all corresponding to 14 body segments (Figure of gene action further suggests that the expression of one
13-15d). Note that the domains of gene expression become set of genes might govern the expression of the succeeding
progressively more refined as development proceeds: genes set of genes.
are expressed first in large regions (gap proteins), then in One clue that this progression is indeed the case comes
stripes from three to four cells wide (pair-rule proteins), from analyzing the effects of mutations in toolkit genes
and then in stripes from one to two cells wide (segment- on the expression of other toolkit genes. For example, in
polarity proteins). embryos from Bicoid mutant mothers, the expression of
In addition to what we have learned from the spa- several gap genes is altered, as well as that of pair-rule
tial patterns of toolkit-gene expression, the order and segment-polarity genes. This finding suggests that the
of toolkit-gene expression over time is logical. The Bicoid protein somehow (directly or indirectly) influences
maternal-effect Bicoid protein appears before the zygotic the regulation of gap genes.
gap proteins, which are expressed before the 7-striped Another clue that the expression of one set of genes
patterns of pair-rule proteins appear, which in turn pre- might govern the expression of the succeeding set of genes
cede the 14-striped patterns of segment-polarity pro- comes from examining the protein products. Inspection
teins. The order of gene expression and the progressive of the Bicoid protein sequence reveals that it contains
refinement of domains within the embryo reveal that the a homeodomain, related to but distinct from those of
Expression of anteroposterior-axis-patterning proteins Hox proteins. Thus, Bicoid has the properties of a DNA-
binding transcription factor. Each gap gene also encodes
(a) a transcription factor, as does each pair-rule gene, several
segment-polarity genes, and, as described earlier, all Hox
genes. These transcription factors include representatives
of most known families of sequence-specific DNA-binding
proteins; so, although there is no restriction concerning
to which family they may belong, many early-acting tool-
(b)
kit proteins are transcription factors. Those that are not
transcription factors tend to be components of signaling
pathways (Table 13-1). These pathways, shown in generic
form in Figure 13-16 , mediate ligand-induced signaling
processes between cells, and their output generally leads
to gene activation or repression. Thus, most toolkit pro-
(c) teins either directly (as transcription factors) or indi-
rectly (as components of signaling pathways) affect gene
regulation.
TABLE 13-1 Examples of Drosophila A–P Axis Genes That Contribute to Pattern Formation
Gene symbol Gene Name Protein function Role(s) in early development
bcd Bicoid Transcription factor—homeodomain protein Maternal-effect gene
hb-z hunchback-zygotic Transcription factor—zinc-finger protein Gap gene
Kr Krüppel Transcription factor—zinc-finger protein Gap gene
kni knirps Transcription factor—steroid receptor-type protein Gap gene
eve even-skipped Transcription factor—homeodomain protein Pair-rule gene
ftz fushi tarazu Transcription factor—homeodomain protein Pair-rule gene
opa odd-paired Transcription factor—zinc-finger protein Pair-rule gene
prd paired Transcription factor—paired class homeodomain protein Pair-rule gene
en engrailed Transcription factor—homeodomain protein Segment-polarity gene
wg wingless Signaling protein-secreted ligand Segment-polarity gene
hh hedgehog Signaling protein-secreted ligand Segment-polarity gene
ptc patched Signaling protein-transmembrane receptor Segment-polarity gene
lab labial Transcription factor—homeodomain protein Segment-identity gene
Dfd Deformed Transcription factor—homeodomain protein Segment-identity gene
Antp Antennapedia Transcription factor—homeodomain protein Segment-identity gene
Ubx Ultrabithorax Transcription factor—homeodomain protein Segment-identity gene
Nucleus
TF
Enhancer Promoter
that is deposited in the egg and localized at the anterior Several zygotic genes, including gap genes, are regulated
pole. Because the early Drosophila embryo is a syncytium by different levels of the Bicoid protein. For example, the
with all nuclei in one cytoplasm, and lacks any cell mem- hunchback gene is a gap gene activated in the zygote in
branes that would impede the diffusion of protein mole- the anterior half of the embryo. This activation is through
cules, the Bicoid protein can diffuse through the cytoplasm. direct binding of the Bicoid protein to three sites 5′ of the
This diffusion establishes a protein concentration gradient promoter of the hunchback gene. Bicoid binds to these sites
(Figure 13-17a): the Bicoid protein is highly concentrated at cooperatively; that is, the binding of one Bicoid protein
the anterior end, and this concentration gradually decreases molecule to one site facilitates the binding of other Bicoid
as distance from that end increases, until there is very little molecules to nearby sites.
Bicoid protein beyond the middle of the embryo. This con- In vivo experiments can demonstrate that the activa-
centration gradient provides positional information about tion of hunchback depends on the concentration gradient.
the location along the anteroposterior axis. A high concen- These tests require linking gene regulatory sequences to a
tration means anterior end, a lower concentration means reporter gene (an enzyme-encoding gene such as the LacZ
middle, and so on. Thus, a way to ensure that a gene is gene or the green fluorescent protein of jellyfish; see Chap-
activated in only one location along the axis is to link gene ter 10), introducing the DNA construct into the fly germ
expression to the concentration level. A case in point is the line, and monitoring reporter expression in the embryo off-
gap genes, which must be activated in specific regions along spring of transgenic flies (a general overview of the method
the axis. is shown in Figure 13-18). The wild-type sequences 5′ of the
hunchback gene are sufficient to drive reporter expression
Gap genes are activated by specific maternally in the anterior half of the embryo. Importantly, deletions
provided proteins of Bicoid-binding sites in this cis-acting regulatory element
reduce or abolish reporter expression (Figure 13-17b).
(a)
Bicoid gradient Hunchback expression More than one Bicoid site must be occupied to generate
a sharp boundary of reporter expression, which indicates
that a threshold concentration of Bicoid protein is required
to occupy multiple sites before gene expression is activated.
A gap gene with fewer binding sites will not be activated at
(b)
locations with lower concentration of Bicoid protein.
Bicoid proteins Reporter expression
Each gap gene contains cis-acting regulatory elements
with different arrangements of binding sites, and these bind-
ing sites may have different affinities for the Bicoid protein.
hunchback 5′ reporter Consequently, each gap gene is expressed in a unique dis-
cis-acting regulatory gene tinct domain in the embryo, in response to different levels
element with
Bicoid-binding sites of Bicoid and other transcription-factor gradients. A similar
theme is found in the patterning of the dorsoventral axis:
cis-acting regulatory elements contain different numbers
and arrangements of binding sites for the maternally sup-
plied Dorsal protein and other zygotic transcription factors.
Consequently, genes are activated in discrete domains along
the dorsoventral axis.
Analysis of cis-acting regulatory elements with reporter genes second stripe expressed by the even-skipped
gene (Figure 13-19a). This stripe lies within
Toolkit gene
the broad region of hunchback expression
A B C and on the edges of the regions of expres-
Promoter Coding region sion of two other gap proteins, Giant and
Cis-acting regulatory elements Krüppel (Figure 13-19b). Thus, within the
area of the future stripe, there will be large
Isolate cis-acting regulatory amounts of Hunchback protein and small
DNA fragments.
amounts of Giant protein and Krüppel pro-
tein. There will also be a certain concentra-
A B C tion of the maternal-effect Bicoid protein.
No other stripe of the embryo will contain
Clone fragments into DNA vector with these proteins in these proportions. The for-
general promoter and reporter gene. mation of stripe 2 is controlled by a specific
cis-acting regulatory element, an enhancer,
Fragments
containing that contains a number of binding sites for
A, B, or C these four proteins (Figure 13-19c). Detailed
Promoter Reporter gene
analysis of the eve stripe 2 cis-acting regu-
latory element revealed that the position
of this “simple” stripe is controlled by the
binding of these four aperiodically distrib-
uted transcription factors, including one
maternal protein and three gap proteins.
Specifically, the eve stripe 2 element
contains multiple sites for the maternal
Bicoid protein and the Hunchback, Giant,
and Krüppel gap proteins (Figure 13-19d).
Inject recombinant constructs into Mutational analyses of different combina-
host embryos (make transgenic by
insertion into germ line); analyze spatial
tions of binding sites revealed that Bicoid
expression of reporter gene by staining and Hunchback activate the expression
for enzyme or by fluorescence. of the eve stripe 2 element over a broad
region. The Giant and Krüppel proteins
are repressors that sharpen the boundaries
of the stripe to just a few cells wide. The
eve stripe 2 element acts, then, as a genetic
switch, integrating multiple regulatory pro-
tein activities to produce one stripe from
Fly embryos
three to four cells wide in the embryo.
FIGURE 13-18 Toolkit loci (such as hunchback, as described in the text) often contain multiple The entire seven-striped periodic pat-
independent cis-acting regulatory elements that control gene expression in different places or tern of even-skipped expression is the
at different times during development or both (for example, A, B, C, here). These elements are sum of different sets of inputs into sepa-
identified by their ability, when placed in cis to a reporter gene and inserted back into a host
rate cis-acting regulatory elements. The
genome, to control the pattern, timing, or level, or all three, of reporter-gene expression. In
this example, each element drives a different pattern of gene expression in a fly embryo. Most enhancers for other stripes contain differ-
reporter genes encode enzymes or fluorescent proteins that can be easily visualized. ent combinations of protein binding sites.
Activators
Bcd-5 Bcd-4 Bcd-3 Bcd-2 Hb-3 Bcd-1
larva. How are the different segmental identities established about 20 cells in different segments. Different structures
by Hox proteins? This process has two aspects. First, the develop from the different segments of the head and tho-
Hox genes are expressed in different domains along the rax, whereas the abdomen is limbless. The first sign of the
anteroposterior axis. Hox-gene expression is largely con- development of these structures is the activation of regula-
trolled by segmentation proteins, especially gap proteins, tory genes within small clusters of cells, which are called
through mechanisms that are similar to those already the appendage primordia. The expression of the Distal-
described herein for hunchback and eve stripe 2 (as well less (Dll) gene marks the start of the development of the
as some cross-regulation by Hox proteins of other Hox appendages. This gene is one of the key targets of the Hox
genes). The regulation of Hox genes will not be considered genes, and its function is required for the subsequent devel-
in depth here. The second aspect of Hox control of segmen- opment of the distal parts of each of these appendages. The
tal identity is the regulation of target genes by Hox pro- small clusters of cells expressing Distal-less arise in several
teins. We will examine one example that nicely illustrates head segments and in each of the three thoracic segments,
how a major feature of the fruit fly’s body plan is con- but not in the abdomen (Figure 13-20a).
trolled through the integration of many inputs by a single How is Distal-less expression restricted to the more
cis-acting regulatory element. anterior segments? Several lines of evidence have revealed
The paired limbs, mouthparts, and antennae of Dro- that the Distal-less gene is repressed in the abdomen by
sophila each develop from initially small populations of two Hox proteins—the Ultrabithorax and Abdominal-A
A8 A7 A6 A5
A4
(a) Wild type
A3
Slp Hox1 Exd En Hth Hox2 Cl
An
A2
Md Mx A1
Lb T1 T2 T3
Repressed in A1–A7
A8 A7 A6 A5
A4
(b) Hox mutations
A3
X X Cl
An
A2
Md Mx A1
Lb T1 T2 T3
Derepressed in A1–A7
A8 A7 A6 A5
(c) Slp mutation A4
A3
X Cl
An
A2
Md Mx Lb T1 T2 T3 A1
Derepressed in aA1–aA7
A8 A7 A6 A5
(d) En mutation A4
A3
X Cl
An
A2
Md Mx Lb T1 T2 T3 A1
Derepressed in pA1–pA7
A8 A7 A6 A5
(e) Slp, En mutations A4
A3
X X Cl
An
A2
Md Mx A1
Lb T1 T2 T3
Derepressed in A1–A7
A8 A7 A6 A5
A4
(f) Exd, Hth mutations
A3
X X Cl
An
A2
Md Mx A1
Lb T1 T2 T3
Derepressed in A1–A7
FIGURE 13-21 Integration of Hox and segmentation-protein inputs by patterns in the abdomen. Binding sites are: Slp, Sloppy-paired; Hox1 and
a cis-acting regulatory element. (a) Left: A cis-acting regulatory element of Hox2, Ultrabithorax and Abdominal-A; Exd, Extradenticle; En, Engrailed;
Introduction
the to Genetic
Dll gene governs Analysis,of11e
the repression Dll expression in the abdomen by Hth, Homothorax. [Data from B. Gebelein, D. J. McKay, and R. S. Mann,
aFigure
set of 13.22 #1340factors. (a) Right: Dll expression (red) extends to the
transcription “Direct Integration of Hox and Segmentation Gene Inputs During Drosophila
06/09/14
thorax but not into the abdomen in a wild-type embryo. (b–f) Mutations Development,” Nature 431, 2004, 653–659.]
Dragonfly
in Media
the respective Groupsites shown derepress Dll expression in various
binding
447
functions. The evolution of gene regulation plays a major proteins. The various genetic factors that influence Dsx
role in the evolution of development and morphology. We expression and sex determination have been identified by
will return to this topic in Chapter 20. mutations that affect the sexual phenotype.
One key regulator is the product of the transformer (tra)
gene. Whereas null mutations in tra have no effect on males,
XX female flies bearing tra mutations are transformed into
13.5 POST-TRANSCRIPTIONAL the male phenotype. The Tra protein is an alternative splic-
REGULATION OF ing factor that affects the splice choices in the dsx RNA
GENE EXPRESSION IN transcript. In the presence of Tra (and a related protein
Tra2), a splice occurs that incorporates exon 4 of the dsx
DEVELOPMENT gene into the mature dsxF transcript (Figure 13-22), but not
exons 5 and 6. Males lack the Tra protein; so this splice
LO 13.4 Infer how spatially and temporally restricted does not occur, and exons 5 and 6 are incorporated into the
patterns of gene expression are generated during
development from analyses of genetic mutations. dsxM transcript, but not exon 4.
The Tra protein explains how alternative forms of Dsx
Although transcriptional regulation is a major means of are expressed, but how is Tra expression itself regulated to
restricting the expression of gene products to defined areas differ in females and males? The tra RNA itself is alternatively
during development, it is not at all the exclusive means of spliced. In females, a splicing factor encoded by the Sex-lethal
doing so. Alternative RNA splicing also contributes to gene (Sxl) gene is present. This splicing factor binds to the tra RNA
regulation, and so does the regulation of mRNA translation and prevents a splicing event that would otherwise incorpo-
by proteins and microRNAs (miRNAs). In each case, regu- rate an exon that contains a stop codon. In males, no Tra pro-
latory sequences in RNA are recognized—by splicing fac- tein is made because this stop codon is present.
tors, mRNA-binding proteins, or miRNAs—and govern the The production of the Sex-lethal protein is, in turn,
structure of the protein product, its amount, or the location regulated both by RNA splicing and by factors that alter
where the protein is produced. We will look at one example the level of transcription. The level of Sxl transcription is
of each type of regulatory interaction at the RNA level. initially governed by activators on the X chromosome and
repressors on the autosomes. In females, which have two
X chromosomes and therefore a double dose of activators,
RNA splicing and sex determination Sxl activation prevails and the Sxl protein is produced,
in Drosophila which regulates tra RNA splicing and feeds back to regu-
A fundamental developmental decision in sexually repro- late the splicing of Sxl RNA itself. In females, a stop codon
ducing organisms is the specification of sex. In animals, is spliced out so that Sxl protein production can continue.
the development of many tissues follows different paths, However, in males, which have only one X chromosome
depending on the sex of the individual animal. In Dro- and therefore only half the dose of X-linked activators,
sophila, many genes have been identified that govern sex transcription of Sxl is initially repressed. Later, Sxl tran-
determination through the analysis of mutant phenotypes scription is activated in males, but the absence of Sxl pro-
in which sexual identity is altered or ambiguous. tein means that the stop codon is still present in unspliced
The doublesex (dsx) gene plays a central role in gov- Sxl RNA transcript and no Sxl protein can be produced.
erning the sexual identity of somatic (non-germ-line) tissue. This cascade of sex-specific RNA splicing in D. mela-
Null mutations in dsx cause females and males to develop nogaster illustrates one way that the sex-chromosome
as intermediate intersexes, which have lost the distinct dif- genotype leads to different forms of regulatory proteins
ferences between male and female tissues. Although dsx being expressed in one sex and not the other. Interestingly,
function is required in both sexes, different gene products the genetic regulation of sex determination differs greatly
are produced from the locus in different sexes. In males, the between animal species, in that sexual genotype can lead
product is a specific, longer isoform, DsxM, that contains a to differential expression of regulatory genes through dis-
unique C-terminal region of 150 amino acids not found in tinctly different paths. However, proteins related to Dsx
the female-specific isoform DsxF, which instead contains a do play roles in sexual differentiation in a wide variety of
unique 30-amino-acid sequence at its carboxyl terminus. animals, including humans. Thus, although there are many
Each form of the Dsx protein is a DNA-binding transcription ways to generate differential expression of transcription
factor that apparently binds the same DNA sequences. How- factors, a family of similar proteins plays conserved roles in
ever, the activities of the two isoforms differ: DsxF activates sexual differentiation across a diversity of species.
certain target genes in females that DsxM represses in males.
The alternative forms of the Dsx protein are generated KEY CONCEPT The sex determination pathway in D. melano
gaster is an example of how the spatial and temporal expres-
by alternative splicing of the primary dsx RNA transcript.
sion of genes involved in developmental pathways can be
Thus, in this case, the choice of splice sites must be reg-
regulated by differential splicing.
ulated to produce mature mRNAs that encode different
Female-specific “Default”
Sex-lethal
splicing splicing
1 2 4 5 6 7 8 AAA 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 AAA
Male exon
Stop codon
Sex-lethal
transformer
1 3 4 AAA 1 2 3 4 1 2 3 4 AAA
Stop codon
Tra-2
Transformer
doublesex
1 2 3 4 AAA 1 2 3 4 5 6 1 2 3 5 6 AAA
FIGURE 13-22 Three pre-mRNAs of major Drosophila sex-determining genes are alternatively A N I M ATED A RT
spliced. The female-specific pathway is shown on the left and the male-specific pathway shown on
LOin Drosophila
Sex determination
the right. The pre-mRNAs are identical in both sexes and shown in the middle. In the male Sex-lethal
and transformer mRNAs, there are stop codons that terminate translation. These sequences are
removed by splicing to produce functional proteins in the female. The Transformer and Tra-2 proteins
then splice the female doublesex pre-mRNA to produce the female-specific isoform of the Dsx protein,
which differs from the male-specific isoform by the alternative splicing of several exons.
Regulation of mRNA translation and cell important in understanding the role of post-transcriptional
lineage in C. elegans regulation at the RNA level, and we will examine two mech-
anisms here: (1) control of translation by mRNA-binding pro-
In many animal species, the early development of the embryo teins, and (2) miRNA control of gene expression.
entails the partitioning of cells or groups of cells into discrete
lineages that will give rise to distinct tissues in the adult. This
process is best understood in the nematode worm C. elegans, in Translational control in the early embryo
which the adult animal is composed of just about 1000 somatic We first look at how a cell lineage begins. After two cell
cells (a third of which are nerve cells) and a similar number divisions, the C. elegans embryo contains four cells, called
of germ cells in the gonad. The simple construction, rapid life blastomeres. Each cell will begin a distinct lineage, and the
cycle, and transparency of C. elegans has made it a powerful descendants of the separate lineages will have different fates.
model for developmental analysis (see the Model Organism Already at this stage, differences are observed in the proteins
box on C. elegans on page 451). All of this animal’s cell line- present in the four blastomeres. However, the mRNAs encod-
ages were mapped out in a series of elegant studies led by John ing some worm toolkit proteins are present in all cells of the
Sulston at the Medical Research Council (MRC) Laboratory of early embryo, and post-transcriptional regulation determines
Molecular Biology in Cambridge, England. Systematic genetic which of these mRNAs will be translated into proteins. Thus,
screens for mutations that disrupt or extend cell lineages have in the C. elegans embryo, post-transcriptional regulation is
provided a bounty of information about the genetic control critical for the proper specification of early cell fates. During
of lineage decisions. C. elegans genetics has been especially the very first cell division, polarity within the zygote leads
to the partitioning of regulatory molecules to specific embry- On the basis of how we have seen transcription con-
onic cells. For example, the glp-1 gene encodes a transmem- trolled, we might guess that one or more proteins bind(s)
brane receptor protein (related to the Notch receptor of flies to the SCR to repress translation of the glp-1 mRNA. To
and other animals). Although the glp-1 mRNA is present in identify these repressor proteins, researchers isolated pro-
all cells at the four-cell stage, the GLP-1 protein is translated teins that bind to the SCR. One protein, GLD-1, binds spe-
only in the two anterior cells ABa and ABp (Figure 13-23a). cifically to a region of the SCR. Furthermore, the GLD-1
This localized expression of GLP-1 is critical for establish- protein is enriched in posterior blastomeres, just where
ing distinct fates. Mutations that abolish glp-1 function at the the expression of glp-1 is repressed. Finally, when GLD-1
four-cell stage alter the fates of ABp and ABa descendants. expression is inhibited by using RNA interference, the
GLP-1 is localized to the anterior cells by repressing its GLP-1 protein is expressed in posterior blastomeres (Figure
translation in the posterior cells. The repression of GLP-1 13-23d). This evidence suggests that GLD-1 is a transla-
translation requires sequences in the 3′ UTR of the glp-1 tional repressor protein controlling the expression of glp-1.
mRNA—specifically, a 61-nucleotide region called the spa- The spatial regulation of GLP-1 translation is but one
tial control region (SCR). The importance of the SCR has example of translational control in development. Transla-
been demonstrated by linking mRNA transcribed from tional control is also important in the establishment of the
reporter genes to different variants of the SCR. Deletion anteroposterior axis in Drosophila and the development of
of this region or mutation of key sites within it causes the sperm in mammals. Again, we see that genetic analysis in
reporter gene to be expressed in all four blastomeres of the model organisms can reveal deeply conserved mechanisms
early embryo (Figure 13-23c). for the regulation of gene expression.
Pharynx Ovary
Intestine Eggs Rectum
Anus
Oviduct Oocytes
Uterus
Vulva
An adult hermaphrodite Caenorhabditis elegans, showing various organs. [Republished with permission of Elsevier,
from J. E. Sulston and H. R. Horvitz, “Post-embryonic Cell Lineages of the Nematode, Caenorhabditis elegans,”
Developmental Biology, 1977, March; 56(1):110–56, Figure 1. Permission conveyed through Copyright Clearance
Center, Inc.]
Production of the vulval-cell lineages. Parts of the vulval anatomy are occupied by
Anterior Posterior so-called primary (1°), secondary (2°), and tertiary (3°) cells. The lineages or pedigrees
of the primary, secondary, and tertiary cells are distinguished by their cell-division
patterns and will give rise to different parts of the vulva in the adult worm, as shown
in the bottom image. [Republished with permission of Jennifer L. Green, Takao Inoue
and Paul W. Sternberg, Development and The Company of Biologists, The C. elegans
ROR receptor tyrosine kinase, CAM-1, nonautonomously inhibits the Wnt pathway, of
Jennifer L. Green, Takao Inoue and Paul W. Sternberg, Development 134, 4053–4062
(2007), and The Company of Biologists; permission conveyed through Copyright
Clearance Center, Inc.]
451
miRNA control of developmental timing in ( Figure 13-24a). Conversely, increased let-7 gene dosage
C. elegans and other species causes the precocious specification of adult fates in larval
stages.
Development is a temporally as well as spatially ordered The let-7 gene does not encode a protein. Instead,
process. When events take place is just as important as it encodes a temporally regulated mature 22-nucleotide
where. Mutations in the heterochronic genes of C. elegans RNA that is processed from an approximately
have been sources of insight into the control of develop- 70-nucleotide precursor. The mature RNA is complemen-
mental timing. Mutations in these genes alter the timing of tary to sequences in 3′ untranslated regions of a variety of
events in cell-fate specification, causing such events to be developmentally regulated genes, and the binding of the
either reiterated or omitted. Detailed investigation into the miRNA to these sequences hinders translation of these
products of heterochronic genes led to the discovery of an gene transcripts. One of these target genes, lin-41, also
entirely unexpected mechanism for regulating gene expres- affects the larval-to-adult transition. The lin-41 mutants
sion, through microRNAs (see Chapter 9). cause precocious specification of adult cell fates, suggest-
Among the first members of this class of regulatory ing that the effect of let-7 overexpression is due at least
molecules to be discovered in C. elegans is RNA produced in part to an effect on lin-41 expression. The let-7 mRNA
by the let-7 gene. The let-7 gene regulates the transition binds to lin-41 RNA in vitro at several imperfect comple-
from late-larval to adult cell fates. In let-7 mutants, for mentary sites (Figure 13-24b).
example, larval cell fates are reiterated in the adult stage
L1
L2
L3
L4
Adult
1 2
GUU A AUU
lin-41 3′ UTR 5′- UUAUACAACC CUACC UCA -3′ 5′- UUAUACAACC CUGCCUC -3′
let-7 RNA 3′- UGAUAUGUUGG GAUGG AGU -5′ 3′- UGAUAUGUUGG GAUGGAGU -5′
AU AU
FIGURE 13-24 Normally, C. elegans develops into an is complementary to sequences at two sites in the 3′ UTR of
adult after four larval stages, and hypodermal cell lineages lin-41 mRNA. [Data from A. E. Rougvie, “Intrinsic and Extrinsic
conclude their development at L4 (hatched lines at ends Regulators of Developmental Timing: From miRNAs to Nutritional
of V1–V4 lineages). (a) In let-7 mutants, the transition from Cues,” Development 132, 2005, 3787–3798; and from D. M.
the L4 larval stage to adult is delayed and the cell lineages Eisenmann, “Wnt signaling” (June 25, 2005), WormBook, ed.
of lateral hypodermal cells (V) are reiterated. Conversely, in The C. elegans Research Community, WormBook, doi/10.1895/
lin-41 mutants, there is precocious development of adult cell wormbook.1.7.1, http://www.wormbook.org.]
fates in the L3 larval stage. (b) let-7 encodes an miRNA that
C
U A
D. rerio lin-41 3′ UTR 5′- CUG AU ACC CCUACCUCA -3′
let-7 RNA UGAU UG UG GAUGGAGU -5′
3′- U A U GAU
G C GA
D. melanogaster lin-41 3′ UTR 5′- A UG UACAAC UUACCUCG -3′
let-7 RNA U AU AUGUUG GAUGGAGU -5′
3′- U G GAU
The role of miRNAs in C. elegans development extends that the Ultrabithorax protein represses limb formation
far beyond let-7. Several hundred miRNAs have been in the fly abdomen and promotes hind-wing development
identified, and many target genes have been shown to be in the fly thorax. Similarly, Sloppy-paired and Engrailed
miRNA regulated. Moreover, the discovery of this class participate in the generation of the basic segmental orga-
of regulatory RNAs prompted the search for such genes nization of the embryo and collaborate with Hox proteins
in other genomes, and, in general, hundreds of candidate to suppress limb formation. These roles are just a few of
miRNA genes have been detected in both plant and animal the many roles played by these toolkit genes in the entire
genomes, including those of humans. course of fly development. Most toolkit genes function at
Quite surprisingly, the let-7 miRNA gene is widely con- more than one time and place, and most may influence the
served and found in Drosophila, ascidian, mollusc, annelid, and formation or patterning of many different structures that
vertebrate (including human) genomes. The lin-41 gene also is are formed in different parts of the larval or adult body.
conserved, and evidence suggests that the let-7–lin-41 regulatory Those that regulate gene expression may directly regulate
interaction also controls the timing of events in the development scores to hundreds of different genes. The function of an
of other species, such as mouse and zebrafish (Figure 13-25). individual toolkit protein (or RNA) is almost always con-
The discoveries of miRNA regulation of developmental text dependent, which is why the toolkit analogy is perhaps
genes and of the scope of the miRNA repertoire are fairly so fitting. As with a carpenter’s toolkit, a common set of
recent. Geneticists and other biologists are quite excited about tools can be used to fashion many structures.
the roles of this class of regulatory molecules in normal devel- To illustrate this principle more vividly, we will look
opment, as well as in the pathology and treatment of disease, at the role of one toolkit protein in the development of
leading to a very vigorous, fast-paced area of new research. many vertebrate features, including features present in
humans. This toolkit protein is the vertebrate homolog of
KEY CONCEPT Sequence-specific RNA-binding proteins the Drosophila hedgehog gene. The hedgehog gene was
and micro RNAs act through cis-acting sequences in the 3′ first identified by Nüsslein-Volhard and Wieschaus as a seg-
untranslated regions of mRNAs to regulate the spatial and ment-polarity gene. It has been characterized as encoding a
temporal pattern of protein translation. signaling protein secreted from cells in Drosophila.
As the evidence grew that toolkit genes are common to
13.6 FROM FLIES TO FINGERS, different animal phyla, the discovery and characterization
of fly toolkit genes such as hedgehog became a common
FEATHERS, AND FLOOR springboard to the characterization of genes in other taxa,
PLATES: THE MANY ROLES OF particularly vertebrates. The identification of homologous
INDIVIDUAL TOOLKIT GENES genes based on sequence similarity was a fast track to the
identification of vertebrate toolkit genes. The application of
this strategy to the hedgehog gene illustrates the power and
LO 13.1 Outline experimental approaches to identify and
characterize members of the genetic toolkit for payoffs of using homology to discover important genes. Sev-
development in different animal phyla. eral distinct homologs of hedgehog were isolated from ver-
LO 13.5 Summarize the evidence that the genetic toolkit for tebrates including zebrafish, mice, chickens, and humans. In
development is conserved across animal phyla. the whimsical spirit of the Drosophila gene nomenclature,
the three vertebrate homologs were named Sonic hedge-
We have seen that toolkit proteins and regulatory RNAs hog (after the video-game character), Indian hedgehog, and
have multiple roles in development. For example, recall Desert hedgehog.
The Sonic hedgehog toolkit gene has multiple roles One of the first means of character-
izing the potential roles of these genes in
(a) (b) development was to examine where they
are expressed. Sonic hedgehog (Shh) was
found to be expressed in several parts
of the developing chicken, with similar
expression patterns in other vertebrates.
Most intriguing was its expression in the
posterior part of the developing limb bud
(Figure 13-26a). This part of the limb bud
was known for decades to be the zone of
polarizing activity (ZPA) because it is an
organizer responsible for establishing the
anteroposterior polarity of the limb and
its digits (see Figure 13-2). To test whether
(b) Shh might play a role in ZPA function,
Cliff Tabin and his colleagues at Harvard
Medical School caused the Shh protein
to be expressed in the anterior region
of developing chick limb buds. They
observed the same effect as transplanta-
FIGURE 13-26 The Shh gene is expressed in many different parts of the developing chick tion of the ZPA—the induction of extra
embryo (indicated by dark staining), including (a) the zone of polarizing activity in each of the digits with reversed polarity. Their results
two developing limb buds and the long neural tube, and (b) the developing feather buds. Shh were stunning evidence that Shh was the
mRNA is visualized by in situ hybridization. [Photomicrographs courtesy of (a) Cliff Tabin and long-sought morphogen produced by the
(b) Photomicrographs courtesy Dr. John Fallon, University of Wisconsin / Matthew Harris, Harvard
Medical School, Department of Genetics.]
ZPA (Figure 13-27).
IV
III
IV
II II IV
Shh is also expressed in other intriguing patterns in of toolkit-gene mutations have been identified that affect
the chicken and other vertebrates. For example, Shh is human development and health. We will focus here on just
expressed in developing feather buds, where it plays a role a few examples that illustrate how understanding gene
in establishing the pattern and polarity of feather forma- function and regulation in model animals has translated
tion (Figure 13-26b). Shh is also expressed in the develop- into better understanding of human biology.
ing neural tube of vertebrate embryos, in a region called
the floor plate (Figure 13-26a). Subsequent experiments
Polydactyly
have shown that Shh signaling from these floor-plate cells
is critical for the subdivision of the brain hemispheres and A fairly common syndrome in humans is the development
the subdivision of the developing eye into the left and right of extra partial or complete digits on the hands and feet.
sides. When the function of the Shh gene is eliminated by This condition, called polydactyly, arises in about 5 to 17
mutation in the mouse using techniques described in Chap- of every 10,000 live births. In the most dramatic cases, the
ters 10 and 14, these hemispheres and eye regions do not condition is present on both hands and feet (Figure 13-28).
separate, and the resulting embryo is cyclopic, with one cen- Polydactyly occurs widely throughout vertebrates—in cats,
tral eye and a single forebrain (it also lacks limb structures). chickens, mice, and other species.
Shh is just one striking example of the dramatic and The discovery of the role of Shh in digit patterning led
diverse roles played by toolkit genes at different places and geneticists to investigate whether the Shh gene was altered
times in development. The outcomes of Shh signaling are dif- in polydactylous humans and other species. In fact, some
ferent in each case: the Shh signaling pathway will induce the cases of polydactyly in humans (and also in cats) result
expression of one set of genes in the developing limb, a dif- from mutations of the Shh gene. Importantly, the muta-
ferent set in the feather bud, and yet another set in the floor tions are not in the coding region of the Shh gene; rather,
plate. How are different cell types and tissues able to respond they lie in a cis-acting regulatory element, far from the cod-
differently to the same signaling molecule? Just as we learned ing region, that controls Shh expression in the developing
from investigating the genetic control of patterning in the limb bud. The extra digits are induced by the expression
Drosophila embryo, the outcome of Shh signaling depends of Shh in a part of the limb where the gene is not normally
on the integration with the signals provided by other toolkit expressed. Mutations in cis-acting regulatory elements have
genes that are acting at the same time and in the same place. two important properties that are distinct from mutations
in coding regions. First, because they affect regulation in
KEY CONCEPT Most toolkit genes have multiple roles in cis, the phenotypes are often dominant. Second, because
different tissues and cell types. The specificity of their action is only one of several cis-acting regulatory elements may be
determined by the context provided by the other toolkit genes affected, other gene functions may be completely normal.
that act in combination with them. Polydactyly can occur without any collateral developmen-
tal problems that would be expected given the multiple
roles of Shh in development. For similar reasons, we will
see in Chapter 20 that mutations in cis-acting regulatory
13.7 DEVELOPMENT AND DISEASE elements of toolkit genes also play key roles during the evo-
lution of morphological differences among species. Coding
LO 13.5 Summarize the evidence that the genetic toolkit for mutations in Shh, however, tell a different story, as we will
development is conserved across animal phyla. see in the next section.
associated with a syndrome termed holoprosencephaly, in signaling proteins. In addition to causing inherited develop-
which abnormalities occur in brain size, in the formation of mental disorders such as polydactyly and holoprosenceph-
the nose, and in other midline structures. These abnormalities aly, mutations in the human patched gene are associated
appear to be less severe counterparts of the developmental with the formation of a variety of cancers. About 30 to 40
defects observed in homozygous Shh mutant mice. Indeed, percent of patients with a dominant genetic disorder called
the affected children seen in clinics are heterozygous. One basal cell nevus syndrome (BCNS) carry patched muta-
copy of a normal Shh gene appears to be insufficient for tions. These persons are strongly disposed to develop a
normal midline development (the gene is haploinsufficient). type of skin cancer called basal-cell carcinoma. They also
Human fetuses homozygous for loss-of-function Shh muta- have a greatly increased incidence of medulloblastoma, a
tions very likely die in gestation with more severe defects. very deadly form of brain tumor. A growing list of cancers
Holoprosencephaly is not caused exclusively by Shh are now associated with disruptions of signal-transduction
mutations. Shh is a ligand in a signal-transduction pathway. pathways—pathways that were first elucidated by these
As might be expected, mutations in genes encoding other early systematic genetic screens for patterning mutants in
components of the pathway affect the efficiency of Shh signal- fruit flies (Table 13-2).
ing and are also associated with holoprosencephaly. Several The discoveries of links between mutations of sig-
components of the human Shh pathway were first identified nal-transduction-pathway genes and human cancer have
as homologs of members of the fly pathway, demonstrating greatly facilitated the study of the biology of cancer and
once again both the conservation of the genetic toolkit and the development of new therapies. For example, about 30
the power of model systems for biomedical discovery. percent of mice heterozygous for a targeted mutation in the
patched gene develop medulloblastoma. These mice there-
fore serve as an excellent model for the biology of human
Cancer as a developmental disease disease and a testing platform for therapy.
In long-lived animals, such as ourselves and other mam- One promising avenue for the development of new can-
mals, development does not cease at birth or at the end of cer therapies is to identify drugs that can specifically target
adolescence. Tissues and various cell types are constantly and kill cancer cells without affecting normal cells. These
being replenished. The maintenance of many organ func- so-called targeted therapies are already employed today
tions depends on the controlled growth and differentiation for the treatment of some cancers. For example, Herceptin
of cells that replace those that are sloughed off or otherwise is a drug used to treat breast cancers with overexpression
die. Tissue and organ maintenance is generally controlled by of the human epidermal growth factor receptor 2 (HER2
signaling pathways. Inherited or spontaneous mutations in or ERBB2), a homolog of the Drosophila torpedo gene
genes encoding components of these pathways can disrupt (Table 13-2). Much current research is focused on identify-
tissue organization and contribute to the loss of control of ing additional drugs to specifically target the signal-trans-
cell proliferation. Because unchecked cell proliferation is a duction pathways that are disrupted in different types of
characteristic of cancer, the formation of cancers may be tumors and that were often first identified in genetic screens
a consequence. Cancer, then, is a developmental disease, a in flies and worms.
product of normal developmental processes gone awry. It is fair to say that even the most optimistic and far-
Some of the genes associated with types of human can- sighted researchers did not expect that the discovery of the
cers are shared members of the animal toolkit. For exam- genetic toolkit for building a fly would have such far-ranging
ple, the patched gene encodes a receptor for the Hedgehog effects on understanding human development and disease.
But such huge unforeseen dividends are familiar in the recent KEY CONCEPT Investigation into the genetic control of
history of basic genetic research. The advent of genetically development in model organisms such as Drosophila and
engineered medicines, monoclonal antibodies for diagnosis C. elegans has led to unexpected and far-reaching consequences
and therapy, and forensic DNA testing all had similar origins for the understanding and treatment of human disease.
in seemingly unrelated investigations.
SUMMARY
In Chapter 11, we mentioned the quip from Jacques Monod progression. Domains within the embryo are estab-
and François Jacob that “anything found to be true of E. lished by the expression of toolkit genes that mark out
coli must also be true of Elephants.”2 Now that we have progressively finer subdivisions along both embryonic
seen the regulatory processes that build worms, flies, chick- axes.
ens, humans, and elephants, would we say that they were 3. Spatially restricted patterns of gene expression are prod-
right? If Monod and Jacob were referring to the principle ucts of combinatorial regulation. Each pattern of gene
that gene transcription is controlled by sequence-specific expression has a preceding causal basis. New patterns are
regulatory proteins, we have seen that the bacterial Lac generated by the combined inputs of preceding patterns.
repressor and the fly Hox proteins do indeed act similarly. In the examples presented in this chapter, the positioning
Moreover, their DNA-binding proteins have the same type of pair-rule stripes and the restriction of appendage-reg-
of motif. The fundamental insights that Jacob and Monod ulatory-gene expression to individual segments requires
had concerning the central role of the control of gene tran- the integration of numerous positive and negative regula-
scription in bacterial physiology and that they expected tory inputs by cis-acting regulatory elements.
would apply to cell differentiation and development in com- Post-transcriptional regulation at the RNA level
plex multicellular organisms have been borne out in many adds another layer of specificity to the control of gene
respects in the genetic control of animal development. expression. Alternative RNA splicing and translational
Many features in single-celled and multicellular eukary- control by proteins and miRNAs also contribute to the
otes, however, are not found in bacteria and their viruses. spatial and temporal control of toolkit-gene expression.
Geneticists and molecular biologists have discovered the func- Combinatorial control is key to both the specificity
tions of introns, RNA splicing, distant and multiple cis-act- and the diversity of gene expression and toolkit-gene
ing regulatory elements, chromatin, alternative splicing, and, function. In regard to specificity, combinatorial mech-
more recently, miRNAs. Still, central to the genetic control of anisms provide the means to localize gene expression
development is the control of differential gene expression. to discrete cell populations by using inputs that are not
This chapter has presented an overview of the logic and specific to cell type or tissue type. The actions of tool-
mechanisms for the control of gene expression and develop- kit proteins can thus be quite specific in different con-
ment in fruit flies and a few other model species. We have texts. In regard to diversity, combinatorial mechanisms
concentrated on the toolkit of animal genes for developmen- provide the means to generate a virtually limitless vari-
tal processes and the mechanisms that control the organiza- ety of gene-expression patterns.
tion of major features of the body plan—the establishment of
4. The modularity of cis-acting regulatory elements allows
body axes, segmentation, and segment identity. Although we
for independent spatial and temporal control of tool-
explored only a modest number of regulatory mechanisms in
kit-gene expression and function. Just as the operators
depth, and just a few species, similarities in regulatory logic
and UAS elements of bacteria and simple eukaryotes act
and mechanisms allow us to identify some general themes
as switches in the physiological control of gene expression,
concerning the genetic control of development.
the cis-acting regulatory elements of toolkit genes act as
1. Despite vast differences in appearance and anatomy, ani- switches in the developmental control of gene expression.
mals have in common a conserved toolkit of genes that The distinguishing feature of toolkit genes is the typical
govern development. This toolkit is a small fraction of all presence of numerous independent cis-acting regulatory
genes in the genome, and most of these toolkit genes con- elements that govern gene expression in different spatial
trol transcription factors and components of signal-trans- domains and at different stages of development. The inde-
duction pathways. Individual toolkit genes typically have pendent spatial and temporal regulation of gene expres-
multiple functions and affect the development of different sion enables individual toolkit genes to have different but
structures at different stages. specific functions in different contexts. In this light, it is
2. The development of the growing embryo and its body not adequate or accurate to describe a given toolkit-gene
parts takes place in a spatially and temporally ordered function solely in relation to the protein (or miRNA) that
it encodes because the function of the gene product almost
2
F. Jacob and J. Monod, Cold Spring Harbor Quant. Symp. Biol. 26, 1963, always depends on the context in which it is expressed.
393.
KEY TERMS
enhancer (p. 444) homeodomain (p. 435) pair-rule gene (p. 439)
gain-of-function mutation (p. 433) homeotic transformation (p. 432) positional information (p. 442)
gap gene (p. 439) housekeeping gene (p. 432) segment-polarity gene (p. 439)
gene complex (p. 433) Hox gene (p. 433) serially reiterated structure (p. 432)
genetic screen (p. 432) loss-of-function mutation (p. 433) zygote (p. 438)
genetic toolkit (p. 431) maternal-effect gene (p. 438)
heterochronic gene (p. 452) morphogen (p. 429)
homeobox (p. 435) organizer (p. 429)
SOLVED PROBLEMS
SOLVED PROBLEM 1 mothers with three copies of bcd+, it is even more posterior.
The Bicoid gene (bcd) is a maternal-effect gene required As additional gene doses are added, the cephalic furrow
for the development of the Drosophila anterior region. moves more and more posteriorly, until, in the progeny of
A mother heterozygous for a bcd deletion has only one copy mothers with six copies of bcd+, it is midway along the A–P
of the bcd gene. With the use of P elements to insert copies axis of the embryo. Explain the gene-dosage effect of bcd+
of the cloned bcd+ gene into the genome by transformation, on the formation of the cephalic furrow in light of the con-
it is possible to produce mothers with extra copies of the tribution that bcd+ makes to A–P pattern formation.
gene. The early Drosophila embryo develops an indentation SOLUTION
called the cephalic furrow that is more or less perpendicular
The determination of anterior–posterior parts of the
to the longitudinal, anteroposterior (A–P) body axis. In the
embryo is governed by a concentration gradient of Bicoid
progeny of mothers with only a single copy of bcd+, this
protein, which is therefore a morphogen. The furrow devel-
furrow is very close to the anterior tip, lying at a position
ops at a critical concentration of bcd. As bcd+ gene dosage
one-sixth of the distance from the anterior to the posterior
(and, therefore, Bicoid protein concentration) decreases, the
tip. In the progeny of standard wild-type diploids (having
furrow shifts anteriorly; as the gene dosage increases, the
two copies of bcd+), the cephalic furrow arises more poste-
furrow shifts posteriorly.
riorly, at a position one-fifth of the distance from the ante-
rior to the posterior tip of the embryo. In the progeny of
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
WORKING WITH THE FIGURES the protein is expressed and the phenotype resulting
(The first 16 questions require inspection of text figures.) from the loss of its expression (shown in Figure 13-1)?
1. In Figure 13-2, the transplantation of certain regions 4. Why might there be more differences among the
of embryonic tissue induces the development of struc- sequences of all the Hox proteins within Drosoph-
tures in new places. What are these special regions ila (shown in Figure 13-8) than there are among the
called, and what are the substances they are proposed sequences of the Hox group 4 proteins in Drosophila
to produce? and different vertebrate species (shown in Figure
13-9)?
2. In Figure 13-5, two different methods are illustrated
for visualizing gene expression in developing animals. 5. In Figure 13-11, what is the evidence that vertebrate
Which method would allow one to detect where within Hox genes govern the identity of serially repeated
a cell a protein is localized? structures?
3. Figure 13-7 illustrates the expression of the Ultra- 6. As shown in Figure 13-14, what is the fundamental
bithorax (Ubx) Hox protein in developing flight distinction between a pair-rule gene and a segment-
appendages. What is the relationship between where polarity gene?
25. Suppose you isolate a mutation affecting A–P pattern- c. Summarize the advantages of your experimental
ing of the Drosophila embryo in which every other approaches.
segment of the developing mutant larva is missing. 32. The eyeless gene is required for eye formation in
a. Would you consider this mutation to be a mutation Drosophila. It encodes a homeodomain.
in a gap gene, a pair-rule gene, a segment-polarity a. What would you predict about the biochemical
gene, or a segment-identity gene? function of the Eyeless protein?
b. You have cloned a piece of DNA that contains four b. Where would you predict that the eyeless gene is
genes. How could you use the spatial-expression expressed in development? How would you test
pattern of their mRNA in a wild-type embryo to your prediction?
identify which represents a candidate gene for the
mutation described? c. The Small eye and Aniridia genes of mice and
humans, respectively, encode proteins with very
c. Assume that you have identified the candidate strong sequence similarity to the fly Eyeless pro-
gene. If you now examine the spatial-expression tein, and they are named for their effects on eye
pattern of its mRNA in an embryo that is homo- development. Devise one test to examine whether
zygous mutant for the gap gene Krüppel, would the mouse and human genes are functionally equiv-
you expect to see a normal expression pattern? alent to the fly eyeless gene.
Explain.
33. Gene X is expressed in the developing brain, heart, and
26. In an embryo from a homozygous Bicoid mutant lungs of mice. Mutations that selectively affect gene X
female, which class(es) of gene expression is (are) function in these three tissues map to three different
abnormal? regions (A, B, and C, respectively) 5′ of the X coding
a. Gap genes region.
b. Pair-rule genes a. Explain the nature of these mutations.
c. Segment-polarity genes b. Draw a map of the X locus consistent with the pre-
d. Hox genes ceding information.
e. All answer options are correct. c. How would you test the function of the A, B, and
C regions?
27. The Hunchback protein in normally expressed in the
anterior half of the Drosophila embryo. You find a 34. Why are regulatory mutations at the mouse Sonic
mutation in the 3′ untranslated region of the hunch- hedgehog gene dominant and viable? Why do coding
back gene that results in expression of Hunchback mutations cause more widespread defects?
protein throughout the entire embryo. Provide a 35. A mutation occurs in the Drosophila doublesex gene
molecular explanation for this result. that prevents Tra from binding to the dsx RNA tran-
28. During mouse development, a homolog of the Dro- script. What would be the consequences of this muta-
sophila wingless gene called Wnt7a is expressed in tion for Dsx protein expression in males? In females?
the developing limbs and the female reproductive 36. You isolate a glp-1 mutation of C. elegans and discover
tract. What phenotypes would be predicted to occur that the DNA region encoding the spatial control
in mice with a mutation in the coding region of region (SCR) has been deleted. What will the GLP-1
Wnt7a? protein expression pattern be in a four-cell embryo in
29. Mutations in the Wnt7a gene have been associated mutant heterozygotes? In mutant homozygotes?
with a human syndrome in which there are abnormal- 37. Assess the validity of Monod and Jacob’s remark that
ities of both the limbs and genitalia. Do you predict “anything found to be true of E. coli must also be true
that these mutations are in the coding sequence or in a of Elephants.”
cis-acting regulatory element of the Wnt7a gene? a. Compare the structures and mechanisms of action
of animal Hox proteins and the Lac repressor. In
CHALLENGING PROBLEMS what ways are they similar? In what ways are they
different?
30. Which of the proteins involved in Drosophila develop-
b. Compare the structure and function of the lac
ment can be classified as a morphogen?
operator with the even-skipped stripe 2 enhancer
31. You are interested in the genes that control the devel- (eve stripe 2). How is the control of these “genetic
opment of the eyes in Drosophila. switches” similar or different?
a. Outline the steps you would take to identify and
characterize these genes. GENETICS AND SOCIETY
b. Outline the steps you would take to determine Justify the genetic study of development in model organisms
whether the genes you find in Drosophila are such as Drosophila and C. elegans to understand human
found in other species. development and disease.
I
n the summer of 2009, Dr. Alan Mayer, a pediatrician at Nicholas Volker
Children’s Hospital of Wisconsin in Milwaukee, wrote
to a colleague about the heartbreaking and baffling case
of a four-year-old patient of his (Figure 14-1). For two years,
little Nicholas Volker had endured over 100 trips to the
operating room as doctors tried to manage a mysterious
disease that was destroying his intestines, leaving him vul-
nerable to dangerous infections, severely underweight, and
often unable to eat.
Neither Mayer nor any other doctors had ever seen a
disease like Nicholas’s; they were unable to diagnose it,
or to stem its ravages by any medical, surgical, or nutri-
tional treatment. It was difficult to treat a disease that
no one could identify. So, Dr. Mayer asked his colleague,
Dr. Howard Jacob at the Medical College of Wisconsin, “if
there is some way we can get his genome sequenced. There
is a good chance Nicholas has a genetic defect, and it is
likely to be a new disease. Furthermore, a diagnosis soon
could save his life and truly showcase personalized genomic
medicine.”1
FIGURE 14-1 DNA sequencing of all the exons of Nicholas Volker’s
Dr. Jacob knew that it would be a longshot. Finding genome revealed a single mutation responsible for his debilitating,
a single mutation responsible for a disease would require but previously unidentified, disease. [Gary Porter/Tribune News Service/
sifting through thousands of variations in Nicholas’s DNA. WAUWATOSA/WI/USA/Newscom.]
One key decision was to narrow the search to just the exon
sequences in Nicholas’s DNA. The rationale was that if the
causal mutation was a protein-coding change, then it could was previously known to have a role in the inflammatory
be identified by sequencing all of the exons, or Nicholas’s response, and mutations in the gene were associated with
exome, which comprise a little over 1 percent of the entire a very rare but potentially fatal immune disorder (although
human genome. Still, it would be an expensive search—the not Nicholas’s intestinal symptoms). Based on that knowl-
sequencing would cost about $75,000 with the technology edge, Nicholas’s doctors boosted his immune system with
available at the time. Nevertheless, the money was raised an infusion of umbilical-cord blood from a well-matched
from donors, and Jacob and a team of collaborators under- donor. Over the next several months, Nicholas’s health
took the task. improved to the point where he was able to eat steak and
As Jacob expected, they found more than 16,000 pos- other foods. And over the next two years, Nicholas did not
sible candidate variations in Nicholas’s DNA. They nar- require any further intestinal surgeries.
rowed this long list by focusing on those mutations that The diagnosis and treatment of Nicholas Volker illus-
had not been previously identified in humans, and that trate the dramatic advances in the technology and impact
caused amino acid replacements that were not found in of genomics—the study of genomes in their entirety. The
other species. Eventually, they identified a single base sub- long-awaited promise that genomics would shape clinical
stitution in a gene called the X-linked inhibitor of apoptosis medicine is now very much a reality. The technological and
(XIAP) that changed one amino acid at position 203 of the biological progress from what started as a trickle of data
protein—an amino acid that was invariant among mammals, in the 1990s has been astounding. In 1995, the 1.8-Mb
fish, and even the fruit-fly counterparts of the XIAP gene. (1.8-megabase) genome of the bacterium Haemophilus
Fortunately, the identification of Nicholas’s XIAP influenzae was the first genome of a free-living organism
mutation suggested a therapeutic approach. The XIAP gene to be sequenced. In 1996 came the 12-Mb genome of Sac-
charomyces cerevisiae; in 1998, the 100-Mb genome of
Caenorhabditis elegans; in 2000, the 180-Mb genome of
1
M. Johnson and K. Gallagher, “A Baffling Illness,” Milwaukee Journal Drosophila melanogaster; in 2001, the first draft of the
Sentinel. Published Dec. 10, 2010. Accessed Mar. 5, 2014. 3000-Mb human genome; and, in 2005, the first draft of
462
our closest living relative, the chimpanzee. These species of noncoding elements of the genome. New technologies
are just a small sample. By the end of 2017, over 130,000 for the global, genome-wide analysis of the physiological
bacterial genomes, and nearly 5500 eukaryotic genomes role of all gene products are driving the development of
(including protists, fungi, plants, and animals) had been the new field called systems biology. From an evolution-
sequenced. At the beginning of 2018, the Earth BioGenome ary perspective, genomics provides a detailed view of how
Project announced its bold intention to sequence all of the genomes and organisms have diverged and adapted over
approximately 1.5 million known species of eukaryotes in geological time.
the next 10 years. The DNA sequence of the genome is the starting point
It is no hyperbole to say that genomics has revolution- for a whole new set of analyses aimed at understanding the
ized how genetic analysis is performed and has opened structure, function, and evolution of the genome and its
avenues of inquiry that were not conceivable just a few components. In this chapter, we will focus on three major
years ago. Most of the genetic analyses that we have so far aspects of genomic analysis:
considered employ a forward genetics approach to ana- • Bioinformatics, the analysis of the information content
lyzing genetic and biological processes. That is, the anal- of entire genomes. This information includes the
ysis begins by first screening for mutants that affect some numbers and types of genes and gene products as well
observable phenotype, and the characterization of these as the location, number, and types of binding sites on
mutants eventually leads to the identification of the gene DNA and RNA that allow functional products to be
and the function of DNA, RNA, and protein sequences. In produced at the correct time and place.
contrast, having the entire DNA sequences of an organism’s
• Comparative genomics, which considers the genomes
genome allows geneticists to work in both directions—for-
of closely and distantly related species for evolution-
ward from phenotype to gene, and in reverse from gene
ary insight.
to phenotype ( Figure 14-2 ). Without exception, genome
sequences reveal many genes that were not detected from • Functional genomics, the use of an expanding variety
classical mutational analysis. Using so-called reverse of methods, including reverse genetics, to understand
genetics, geneticists can now systematically study the roles gene and protein function in biological processes.
of such formerly unidentified genes. Moreover, a lack of
prior classical genetic study is no longer an impediment
to the genetic investigation of organisms. The frontiers of
experimental analysis are growing far beyond the bounds 14.1 THE GENOMICS REVOLUTION
of the very modest number of long-explored model organ-
isms (for more, see the Beyond Model Organisms section After the development of recombinant DNA technology
of A Brief Guide to Model Organisms, at the back of this in the 1970s, research laboratories typically undertook
book). the cloning and sequencing of one gene at a time (see
Analyses of whole genomes now contribute to every Chapter 10), and then only after having had first found
corner of biological research. In human genetics, genom- out something interesting about that gene from a clas-
ics is providing new ways to locate genes that contribute sic mutational analysis. The steps in proceeding from the
to many genetic diseases, like Nicholas’s, which had pre- classical genetic map of a locus to isolating the DNA
viously eluded investigators. The day is soon approach- encoding a gene (cloning) to determining its sequence were
ing when a person’s genome sequence is a standard part often numerous and time consuming. In the 1980s, some
of his or her medical record. The availability of genome scientists realized that a large team of researchers mak-
sequences for long-studied model organisms and their ing a concerted effort could clone and sequence the entire
relatives has dramatically accelerated gene identification, genome of a selected organism. Such genome p rojects
the analysis of gene function, and the characterization would then make the clones and the sequence publicly
available resources. One appeal of having these resources
available is that, when researchers become interested in
a gene of a species whose genome has been sequenced,
Comparing forward and reverse genetic approaches
they need only find out where that gene is located on the
map of the genome to be able to zero in on its sequence
Forward genetics and potentially its function. By this means, a gene could
Phenotype Genotype be characterized much more rapidly than by cloning and
Reverse genetics sequencing it from scratch, a project that at the time could
take several years to carry out. This quicker approach is
now a reality for all model organisms.
FIGURE 14-2 Forward genetics is phenotype driven, and asks
Similarly, the Human Genome Project aimed to revo-
what genes underlie a particular phenotype, while reverse genetics
is genotype driven and asks what phenotypes are associated with a lutionize the field of human genetics. The availability of
particular gene. human genome sequences, and the ability to sequence the
genomes of patients and their relatives, has greatly aided advances in information technology aided the analysis of
the identification of disease-causing genes. Furthermore, the resulting data.
the ability to determine gene sequences in normal and dis- The first successes in genome sequencing set off waves
eased tissues (for example, cancers) has been a great cata- of innovation that led to faster and much less expensive
lyst to the understanding of disease processes, and pointed sequencing technologies. Now, individual machines can
the way to new therapies. produce as much sequence in a day as centers used to
From a broader perspective, the genome projects had accomplish in months. New technologies can now obtain
the appeal that they could provide some glimmer of the more than 1× 1012 bases of sequence in a working day
principles on which genomes are built. The human genome on a single instrument. This represents an approximately
contains 3 billion base pairs of DNA. Having the entire 1 million-fold increase in throughput over earlier instru-
sequence raised questions such as: How many genes does it ments used to obtain the first human genome sequences
contain? How are they distributed, and why? What fraction (Figure 14-3).
of the genome is coding sequence? What fraction is regula- Genomics, aided by the explosive growth in informa-
tory sequence? How is our genome similar to or different tion technology, has encouraged researchers to develop
from other animals? Although we might convince ourselves ways of experimenting on the genome as a whole rather
that we understand a single gene of interest, the major than simply one gene at a time. Genomics has also demon-
challenge of genomics today is genomic literacy: How do strated the value of collecting large-scale data sets in
we read the storehouse of information enciphered in the advance so that they can be used later to address specific
sequence of complete genomes? research problems. In the last sections of this chapter, we
The basic techniques needed for sequencing entire will explore some ways that genomics now drives basic
genomes were already available in the 1980s. But the scale and applied genetics research. In subsequent chapters, we
that was needed to sequence a complex genome was, as an will see how genomics is catalyzing advances in under-
engineering project, far beyond the capacity of the research standing the dynamics of mutation, recombination, and
community then. Genomics in the late 1980s and the 1990s evolution.
evolved out of large research centers that could integrate
these elemental technologies into an industrial-level pro- KEY CONCEPT Characterizing whole genomes is funda-
duction line. These centers developed robotics and auto- mental to understanding the entire body of genetic information
underlying the physiology, development, and evolution of living
mation to carry out the many thousands of cloning steps
organisms, and to the discovery of new genes such as those
and millions of sequencing reactions necessary to assemble
having roles in human genetic disease.
the sequence of a complex organism. Just as important,
1×1021 bp
Cumulative number of human genomes
Double every
1×109
12 months
Current 1×1018 bp
capacity
1×106
Recorded 1×1015 bp
growth
1×103 First human
genome
sequences
1×1012 bp
1
2000 2005 2010 2015 2020 2025
Year
FIGURE 14-3 This plot shows the increase in DNA sequencing capacity from the publication of the first
human genome sequences in 2001 until the end of 2015 (solid line), along with the projected growth if
capacity doubles every year (dashed line). The total number of human genomes sequenced is indicated
on the y-axis on the left, and the total worldwide annual sequencing capacity is indicated on the y-axis on
the right. [Data from Z. D. Stephens et al., “Big Data: Astronomical or Genomical?” PLoS Biology, 13(7), 2015,
e1002195, https://doi.org/10.1371/journal.pbio.1002195.]
14.2 OBTAINING THE SEQUENCE of chemistries and optical-detection methods. The meth-
ods now available vary in the length of DNA sequence
OF A GENOME obtained, the bases determined per second, and raw accu-
racy. For large-scale sequencing projects that seek to ana-
LO 14.1 Describe the combinations of strategies typically lyze large individual genomes or the genomes of many
necessary for obtaining and assembling the
complete DNA sequences of organisms.
different individuals or species, choosing a method requires
balancing speed, cost, and accuracy.
When people encounter new territory, one of their first Individual sequencing reactions (called sequencing reads)
activities is to create a map. This practice has been true provide letter strings that, depending on the sequencing tech-
for explorers, geographers, oceanographers, and astrono- nique employed, range on average from about 100 to 15,000
mers, and it is equally true for geneticists. Geneticists use bases long. Such lengths are tiny compared with the DNA
many kinds of maps to explore the terrain of a genome. of a single chromosome. For example, an individual read of
Examples are linkage maps based on inheritance patterns 300 bases is only 0.0001 percent of the longest human chro-
of gene alleles and cytogenetic maps based on the location mosome (about 3 × 108 bp of DNA) and only about 0.00001
of microscopically visible features such as rearrangement percent of the entire human genome. Thus, one major chal-
break points (see Chapters 4 and 17). lenge facing a genome project is sequence assembly—that
The highest-resolution map is the complete DNA is, building up all of the individual reads into a consensus
sequence of the genome—that is, the complete sequence of sequence, a sequence for which there is consensus (or agree-
nucleotides A, T, C, and G of each double helix in the genome. ment) that it is an authentic representation of the sequence
Because obtaining the complete sequence of a genome is such for each of the DNA molecules in that genome.
a massive undertaking of a sort not seen before in biology, Let’s look at these numbers in a somewhat different way
new strategies must be used, all based on automation. to understand the scale of the problem. As with any experi-
mental observation, automated sequencing machines do not
always give perfectly accurate sequence reads. Indeed, newer,
Turning sequence reads into an higher-throughput sequencing technologies generate a greater
assembled sequence frequency of errors than older methods; the error rate may
You’ve probably seen a magic act in which the magician cuts range from less than 1 percent to as much as 15 percent,
up a newspaper page into a great many pieces, mixes it in his depending upon the technology. Thus, to ensure accuracy,
hat, says a few magic words, and voila! an intact newspaper genome projects conventionally obtain many independent
page reappears. Basically, that’s how genomic sequences are sequence reads of each base pair in a genome. Many-fold
obtained. The approach is to (1) break the DNA molecules coverage ensures that chance errors in the reads do not give a
of a genome up into thousands to millions of more or less false reconstruction of the consensus sequence.
random, overlapping small segments; (2) read the sequence Given a sequence read of about 100 bases of DNA and
of each small segment; (3) computationally find the overlap a human genome of 3 billion base pairs, 300 million inde-
among the small segments where their sequences are identi- pendent reads are required to give 10-fold average coverage
cal; and (4) continue overlapping ever larger pieces until all of each base pair. However, not all sequences are represented
the small segments are linked (Figure 14-4). At that point, the equally, and so the number of reads required is even larger.
sequence of a genome is assembled. Typically, 30-fold average coverage is desired when sequenc-
Why does this process require automation? To under- ing a genome. The amount of information to be tracked is
stand why, let’s consider the human genome, which con- enormous. Thus, genome sequencing has required many
tains about 3 × 109 bp of DNA, or 3 billion base pairs advances in automation and information technology.
(3 gigabase pairs = 3 Gbp) . Suppose we could purify the What are the goals of sequencing a genome? First, we
DNA intact from each of the 24 human chromosomes (the strive to produce a consensus sequence that is a true and
22 autosomes, plus the X and the Y sex chromosomes), sep- accurate representation of the genome, starting with one
arately put each of these 24 DNA samples into a sequencing individual organism or standard strain from which the DNA
machine, and read their sequences directly from one telo- was obtained. This sequence will then serve as a reference
mere to the other. Obtaining a complete sequence would sequence for the species. We now know that there are many
be utterly straightforward, like reading a book with 24 differences in DNA sequence between different individuals
chapters—albeit a very, very long book with 3 billion char- within a species and even between the maternally and pater-
acters (about the length of 3000 novels). Unfortunately, nally contributed genomes within a single diploid individual.
such a sequencing machine does not yet exist. Thus, no single genome sequence truly represents the genome
Rather, automated sequencing is the current state of the of the entire species. Nonetheless, the genome sequence
art in DNA sequencing technology. Initially based on the serves as a standard or reference with which other sequences
pioneering dideoxy chain-termination sequencing method can be compared, and it can be analyzed to determine the
developed by Fred Sanger (discussed in Chapter 10; see Fig- information encoded within the DNA, such as the inventory
ure 10-18), automated sequencing now employs a variety of encoded RNAs and polypeptides.
Contigs
ATGTTCCGATTAGGAAACCTATCTGTAACTGTTTCATTCAGTAAAAGGAGG
Like written manuscripts, genome sequences can range generated by breaking the long chromosomes of DNA into
from draft quality (the general outline is there, but there many short segments. Two approaches to WGS sequencing are
are typographical errors, grammatical errors, gaps, sec- responsible for most genome sequences obtained to date. The
tions that need rearranging, and so forth), to finished qual- fundamental differences between them are in how the short
ity (a very low rate of typographical errors, some missing segments of DNA are obtained and prepared for sequencing
sections but everything that is currently possible has been and the sequencing chemistry employed. The first method,
done to fill in these sections), to truly complete (no typo- used to sequence the first human genome, relied on the cloning
graphical errors, every base pair absolutely correct from of DNA in microbial cells and employed the dideoxy sequenc-
telomere to telomere). Although complete assemblies have ing technique. We will refer to this approach as “traditional
been obtained for organisms with small genomes, such as WGS sequencing.” Methods in the second group are generally
bacteria and yeast, this is currently not possible for large cell-free methods that employ new techniques for sequencing
and complex eukaryotic genomes, including human. In the and are designed for very high throughput (referring to the
following sections, we will examine the strategy and some number of reads per machine per unit time). We will refer to
methods for producing draft and finished genome-sequence this group of methods as “next-generation WGS sequencing.”
assemblies. We will also encounter some of the features of
genomes that challenge genome-sequencing projects. Traditional WGS sequencing
The traditional WGS approach begins with the construction
Whole-genome sequencing of genomic libraries, which are collections of these short seg-
The current general strategy for obtaining and assembling ments of DNA, representing the entire genome. The short
the sequence of a genome is called whole-genome shotgun DNA segments in such a library have been inserted into one
(WGS) sequencing. This approach is based on determining the of a number of types of accessory chromosomes (nonessen-
sequence of many segments of genomic DNA that have been tial elements such as plasmids, modified bacterial viruses, or
artificial chromosomes) and propagated in microbes, usually introducing the molecules into bacterial cells. Each cell
bacteria or yeast. These accessory chromosomes carrying takes up one recombinant molecule. Then each recombi-
DNA inserts are called vectors (see Chapter 10). nant molecule is replicated in the normal growth and divi-
To generate a genomic library, a researcher first uses sion of its host so that many identical copies of the inserted
restriction enzymes, which cleave DNA at specific sequences, fragment are produced for use in analyzing the fragment’s
to cut up purified genomic DNA. Some enzymes cut the DNA DNA sequence. Because each recombinant molecule is
at many places, whereas others cut it at fewer places; so the amplified from an individual cell, each cell is a distinct
researcher can control whether the DNA is cut, on average, clone. The resulting library of clones is called a shotgun
into longer or shorter pieces. The resulting fragments have library because sequence reads are obtained from clones
short single strands of DNA at both ends. Each fragment randomly selected from the whole-genome library without
is then joined to the DNA molecule of the accessory chro- any information on where these clones map in the genome.
mosome, which also has been cut with a restriction enzyme Next, the genome fragments in clones from the shotgun
and which has ends that are complementary to those of the library are partially sequenced. The sequencing reaction must
genomic fragments. In order for the entire genome to be rep- start from a primer of known sequence. Because the sequence
resented, multiple copies of the genomic DNA are cut into of a cloned insert is not known (and is the goal of the exer-
fragments. By this means, thousands to millions of different cise), primers are based on the sequence of adjacent vector
fragment-vector recombinant molecules are generated. DNA. These primers are used to guide the sequencing reac-
As discussed in Chapter 10, the resulting pool of recom- tion into the insert. Hence, short regions at one or both ends
binant DNA molecules is then propagated, typically by of the genomic inserts can be sequenced (Figure 14-5). After
sequencing, the output is a large collection of random short the oligonucleotide on the flow cell. Then, the adaptor on
sequences, some of them overlapping. These sequence reads the other end of the DNA molecule will bind to its com-
are assembled into a consensus sequence covering the whole plementary oligonucleotide, which is called bridge forma-
genome by matching homologous sequences shared by reads tion. Once immobilized, each DNA molecule is amplified
from overlapping clones. The sequences of overlapping reads across this bridge by the polymerase chain reaction (PCR;
are assembled into units called sequence contigs, which are see Chapter 10). After one round of PCR bridge amplifi-
sequences that are contiguous, or touching. cation, there will be two DNA molecules with complemen-
tary sequence on the same location on the flow cell. One
KEY CONCEPT Whole genomes can be assembled from end of each of the two DNA molecules will be dissociated
sequencing many short segments of DNA. from the flow cell. This dissociation allows for another
round of bridge formation and PCR bridge amplification
to take place. Repeating this process many times will gen-
erate clusters. Each cluster contains thousands of copies
Next-generation WGS sequencing
of the same DNA fragment in a tiny spot (Figure 14-6b).
The goal of next-generation WGS is the same as that of Each channel of the flow cell contains millions to billions
traditional WGS—to obtain a large number of overlapping of these clusters.
sequence reads that can be assembled into contigs. How-
ever, the methodologies used differ in several substantial Stage 3. The sequencing of each cluster is performed using
ways from traditional WGS. A few different systems have a novel “sequencing-by-synthesis” approach (Figure 14-6c).
been developed that, while they differ in their sequencing DNA polymerase and a primer are added to the flow cell
chemistry and machine design, each employ three strategies to prime the synthesis of a complementary DNA strand.
that have dramatically increased throughput: Each of the four deoxyribonucleotide triphosphates, dATP,
dGTP, dTTP, and dCTP, is labeled with a different fluores-
1. DNA molecules are prepared for sequencing in cell-free cent dye that emits a signal at different wavelengths (and
reactions, without cloning in microbial hosts. therefore appears as a different color). In each sequencing
2. Millions of individual DNA fragments are isolated and cycle, a single nucleotide will be added that is comple-
sequenced in parallel during each machine run. mentary to the next base in the template strand in a given
3. Advanced fluid-handling technologies, cameras, and cluster. When the nucleotide is incorporated, the reaction
software make it possible to detect the products emits a unique wavelength depending upon which base was
of sequencing reactions in extremely small reaction vol- added. After each sequencing cycle, an image of the flow
umes. cell is taken. Each cluster will have added only one of the
four bases and will therefore appear as a spot of a single
Since the field of genomic technology is evolving rap- color in the image. The reaction is repeated for at least 100
idly, we will not describe every next-generation system. and up to 300 cycles, and the signals from each cluster over
Here, we will examine Illumina sequencing, which is cur- all of the cycles are integrated to generate the sequence
rently the most widely used approach that employs all of reads from each cluster.
these features. The Illumina approach illustrates the gains
that have been made in throughput and what such gains The pace of development of next-generation sequencing
enable geneticists to do. The approach can be considered to technologies has been astonishing and is continuing at a diz-
have three stages: zying rate. Recently, so-called “third-generation” sequenc-
ing technologies have been developed to enable sequencing
Stage 1. A DNA sequencing library of DNA molecules is of single molecules of DNA. Third-generation methods like
constructed. After genomic DNA is isolated from an organ- those developed by Pacific Biosciences (PacBio) and Oxford
ism of interest, it is fragmented into smaller pieces of a uni- Nanopore Technologies provide a number of advantages
form size. Then, short sequences called adaptors are added over second-generation sequencing methods such as Illumina.
to both ends of the DNA fragments. There are two adaptor These include the ability to generate very long sequence
sequences; one sequence is added to one end of the DNA reads, which greatly enables the assembly of whole genomes,
fragment, and the other sequence is added to the other end of as detailed in the next section. However, these newer sequenc-
the DNA fragment (Figure 14-6a). ing methods currently have a lower throughput and a higher
Stage 2. The DNA fragments are bound to a sequencing error rate. Thus, the method chosen by researchers depends a
flow cell. This is a glass slide with small channels that great deal on the application, and these choices will continue
are coated with oligonucleotides containing sequences to evolve in the coming years.
complementary to both adaptor sequences (Figure 14-6b,
KEY CONCEPT Next-generation WGS sequencing methods
inset). A single DNA molecule will bind to a unique loca-
have already enabled enormous gains in sequencing output
tion in the flow cell due to hybridization between the
and are continuing to evolve at a rapid rate.
adaptor sequence at one end of the DNA molecule and
A T A A G C C G C AT T G G G C C G G A A AT G G G C C A AT T G G
Isolated DNA
T A T T C G G C G TA A C C C G G C C T T T A C C C G G T TA A C C
Fragment
End repair
Adaptor ligation
(b) Stage 2: DNA fragment binding to flow cell and cluster formation
G C Cycle 1
T
A
A Cycle 2
Adaptors C
..
.
T Cycle 3
G Cycle 4
FIGURE 14-6 Illumina DNA sequencing consists of three stages: (a) DNA sequencing library A N I M ATED A RT
construction; (b) DNA fragment binding to flow cell and cluster formation; and (c) sequencing by
synthesis. See text for details. [Bainscou, Wikimedia Commons, Creative Commons Attribution 3.0 Unported Next-generation sequencing
license, https://commons.wikimedia.org/wiki/File:Next_generation_sequencing_slide.jpg#filehistory.]
Unique
sequence
contigs
Single, collapsed
sequence contig
Paired-end reads may be used to join two sequence contigs next-generation WGS researchers had to devise a way
to bridge these gaps without building genomic libraries
Sequenced contig A Sequenced contig B in vectors. One solution was to build a library of cir-
Paired-end cularized genomic DNA fragments of desired sizes. The
Sequence Sequence reads from circularization allows for short segments of previously
read 1 read 2 same cloned distant sequences located at the ends of each fragment to
insert
be juxtaposed. Shearing of these circular molecules and
amplification and sequencing of fragments containing the
Long-insert junction produces paired-end reads equivalent to those
vector obtained from sequencing of traditional genomic-library
inserts (Figure 14-10).
Scaffold A–B
Further, aligning the sequences of the two contigs by In both traditional and next-generation WGS sequenc-
using paired-end reads automatically determines the rela- ing, some gaps usually remain. Specific procedures targeted
tive orientation of the two contigs. In this manner, single- to individual gaps must be used to fill the missing data in
copy contigs could be joined together, albeit with gaps the sequence assemblies. If the gaps are short, missing frag-
where the repetitive elements reside. These gapped col- ments can be generated by using the known sequences at
lections of joined-together sequence contigs are called the ends of the assemblies as primers to amplify and ana-
scaffolds (sometimes also referred to as supercontigs). lyze the genomic sequence in between. If the gaps are lon-
Because most Drosophila repeats are large (3–8 kb) and ger, attempts can be made to isolate the missing sequences
widely spaced (one repeat approximately every 150 kb), as parts of larger inserts that have been cloned into a vector,
this technique was extremely effective at producing a cor- and then to sequence the inserts. In the future, the longer
rectly assembled draft sequence of the single-copy DNA. sequencing reads generated by third-generation sequencing
A summary of the logic of this approach is shown in methods will also contribute to filling the gaps in sequence
Figure 14-9. assemblies, particularly in regions of the genome that con-
Next-generation WGS sequencing does not circum- tain many repeat sequences.
vent the problem of repetitive sequences and gaps. Since Whether a genome is sequenced to “draft” or “finished”
this approach is intended to circumvent the construc- standards is a cost–benefit judgment. Currently, it is rel-
tion of libraries, which would otherwise facilitate the atively straightforward to create a draft but very hard to
bridging of gaps between contigs via paired-end reads, complete a finished sequence.
Scaffold
Sequenced Sequenced Sequenced
contig 1 GAP contig 2 GAP contig 3
B
2 DNA fragments are circularized so the purple and orange
sequences are located next to each other.
A1 A2
350–600 bp
150 bp
Flow cell
150 bp
Flow cell
14.3 BIOINFORMATICS: MEANING The genomic sequence is a highly encrypted code containing
FROM GENOMIC SEQUENCE the raw information for building and operation of organ-
isms. The study of the information content of genomes is
LO 14.2 Explain the role of various functional elements called bioinformatics. We are far from being able to read
within genomes, and differentiate between this information from beginning to end in the way that we
computational and experimental methods used to would read a book. Even though we know which triplets
identify these elements. encode which amino acids in the protein-coding segments,
much of the information contained in a genome is not deci- is to try to determine an inventory of all of the polypep-
pherable from mere inspection. tides encoded by an organism’s genome. This inventory is
referred to as the organism’s proteome. It can be considered
The nature of the information content a “parts list” for the cell. To determine the list of polypep-
of DNA tides, the sequence of each mRNA encoded by the genome
must be deduced. Because of intron splicing, this task is
DNA contains information, but in what way is it encoded? particularly challenging in multicellular eukaryotes, where
Conventionally, the information is thought of as the sum of introns are the norm. In humans, for example, an average
all the gene products, both proteins and RNAs. However, gene has about 10 exons. Furthermore, many genes encode
the information content of the genome is more complex alternative exons; that is, some exons are included in some
than that. The genome also contains binding sites for differ- versions of a processed mRNA but are not included in oth-
ent proteins and RNAs. Many proteins bind to sites located ers (see Chapter 8). The alternatively processed mRNAs
in the DNA itself, whereas other proteins and RNAs bind can encode polypeptides having much, but not all, of their
to sites located in mRNA (Figure 14-11). The sequence and amino acid sequences in common. Even though we have a
relative positions of those sites permit genes to be tran- great many examples of completely sequenced genes and
scribed, spliced, and translated properly, at the appropri- mRNAs, we cannot yet identify 5′ and 3′ splice sites merely
ate time in the appropriate tissue. For example, regulatory from DNA sequence with a high degree of accuracy. There-
protein-binding sites determine when, where, and at what fore, we cannot be certain which sequences are introns.
level a gene will be expressed. At the RNA level in eukary- Predictions of alternatively used exons are even more error
otes, the locations of binding sites for the RNAs and pro- prone. For such reasons, deducing the total polypeptide
teins of spliceosomes will determine the 5′ and 3′ splice parts list in higher eukaryotes is a large problem. Some
sites where introns are removed. Regardless of whether a approaches follow.
binding site actually functions as such in DNA or RNA, the
site must be encoded in the DNA. The information in the ORF detection The main approach to producing a
genome can be thought of as the sum of all the sequences polypeptide list is to use the computational analysis of
that encode proteins and RNAs, plus the binding sites that the genome sequence to predict mRNA and polypeptide
govern the time and place of their actions. As a genome sequences, an important part of bioinformatics. The proce-
draft continues to be improved, the principal objective is dure is to look for sequences that have the characteristics of
the identification of all of the functional elements of the genes. These sequences would be gene-size and composed
genome. This process is referred to as annotation. of sense codons after possible introns had been removed.
The appropriate 5′- and 3′-end sequences would be present,
KEY CONCEPT The functional elements of the genome
such as start and stop codons. Sequences with these char-
include the sequences that encode proteins and RNAs, as well
acteristics typical of genes are called open reading frames
as the binding sites for the proteins and RNAs that regulate
gene expression. (ORFs). To find candidate ORFs, computer programs scan
the DNA sequence on both strands in each reading frame.
Because there are three possible reading frames on each
Deducing the protein-encoding genes strand, there are six possible reading frames in all.
from genomic sequence
Because the proteins present in a cell largely determine its Direct evidence from cDNA sequences Another
morphology and physiological properties, one of the first means of identifying ORFs and exons is through the anal-
orders of business in genome analysis and annotation ysis of mRNA expression. This analysis can be done in
FIGURE 14-11 A gene within DNA may be viewed as a series of binding sites for proteins and RNAs.
FIGURE 14-12 Alignment of fully sequenced base coordinates of the cDNA sequence, where base
complementary DNAs (cDNAs) and expressed 1 is the 5′-most base and base 816 is the 3′-most
sequence tags (ESTs) with genomic DNA. The dashed base of the cDNA. For the ESTs, only a short sequence
lines indicate regions of alignment; for the cDNA, read is obtained from each end (5′ and 3′) of the
these regions are the exons of the gene. The dots corresponding cDNA. These sequence reads establish
between segments of cDNA or ESTs indicate regions the boundaries of the transcription unit, but they are not
in the genomic DNA that do not align with cDNA or informative about the internal structure of the transcript
EST sequences; these regions are the locations of the unless the EST sequences cross an intron (as is true for
introns. The numbers above the cDNA line indicate the the 3′ EST depicted here).
two ways. Both methods involve the synthesis of librar- Predictions of binding sites As already discussed, a
ies of DNA molecules that are complementary to mRNA gene consists of a segment of DNA that encodes a tran-
sequences, called cDNA (see Chapter 10). The longest estab- script as well as the regulatory signals that determine when,
lished method entails the cloning and amplification of these where, and how much of that transcript is made. In turn,
cDNA molecules in a vector. However, the next-generation that transcript has the signals necessary to determine its
sequencing technologies described in the previous section splicing into mRNA and the translation of that mRNA
also allow for the direct sequencing of short cDNA mole- into a polypeptide (Figure 14-13). There are now statistical
cules without the cloning step, called RNA sequencing or “gene-finding” computer programs that search for the pre-
“RNA-seq” for short (this technique will be described in dicted sequences of the various binding sites used for pro-
more detail later in Section 14.7). Whichever method is uti- moters, for transcription start sites, for 3′ and 5′ splice sites,
lized, complementary DNA sequences are extremely valu- and for translation initiation codons within genomic DNA.
able in two ways. First, they are direct evidence that a given These predictions are based on consensus motifs for such
segment of the genome is expressed and may thus encode a known sequences, but they are not perfect.
gene. Second, because the cDNA is complementary to the
mature mRNA, the introns of the primary transcript have Using polypeptide and DNA similarity Because organ-
been removed, which greatly facilitates the identification of isms have common ancestors, they also have many genes
the exons and introns of a gene (Figure 14-12). with similar sequences in common. Hence, a gene will likely
The alignment of cDNAs with their corresponding have relatives among the genes isolated and sequenced in
genomic sequence clearly delineates the exons, and hence other organisms, especially in the closely related ones. Candi-
introns are revealed as the regions falling between the date genes predicted by the preceding techniques can often be
exons. In the assembled cDNA sequence, the ORF should verified by comparing them with all the other gene sequences
be continuous from initiation codon through stop codon. that have ever been found. A candidate sequence is submit-
Thus, cDNA sequences can greatly assist in identifying ted as a “query sequence” to public databases containing a
the correct reading frame, including the initiation and record of all known gene sequences. This procedure is called
stop codons. Full-length cDNA evidence is taken as the a BLAST search (BLAST stands for Basic Local Alignment
gold-standard proof that one has identified the sequence of Search Tool). The sequence can be submitted as a nucleotide
a transcription unit, including its exons and its location in sequence (a BLASTn search) or as a translated amino acid
the genome. sequence (BLASTp). The computer scans the database and
In addition to full-length cDNA sequences, there are returns a list of full or partial “hits,” starting with the clos-
large data sets of cDNAs for which only the 5′ or the 3′ est matches. If the candidate sequence closely resembles that
ends or both have been sequenced. These short cDNA of a gene previously identified from another organism, then
sequence reads are called expressed sequence tags (ESTs). this resemblance provides a strong indication that the can-
Expressed sequence tags can be aligned with genomic didate gene is a real gene. Less-close matches are still useful.
DNA and thereby used to determine the 5′ and 3′ ends of For example, an amino acid identity of only 35 percent, but
transcripts—in other words, to determine the boundaries of at identical positions, is a strong indicator that two proteins
the transcript as shown in Figure 14-12. have a common three-dimensional structure.
Genomic DNA
Nontemplate strand
Splicing
Ribosome-binding site
mRNA
5′ UTR ORF 3′ UTR
Translation
Polypeptide
FIGURE 14-13 Eukaryotic information transfer from gene to polypeptide chain. Note the DNA and
RNA “binding sites” that are bound by protein complexes to initiate the events of transcription, splicing,
and translation.
BLAST searches are used in many other ways, but a diagnostic for Drosophila because, in other organisms,
always the goal is to find out more about some identified this “codon bias” pattern is quite different. Codon biases
sequence of interest. are thought to be due to the relative abundance of the
tRNAs complementary to these various codons in a given
Predictions based on codon bias Recall from Chapter species. If the codon usage of a predicted ORF matches
9 that the triplet code for amino acids is degenerate; that that species’ known pattern of codon usage, then this
is, most amino acids are encoded by two or more codons match is supporting evidence that the proposed ORF is
(see Figure 9-8). The multiple codons for a single amino genuine.
acid are termed synonymous codons. In a given species,
not all synonymous codons for an amino acid are used Putting it all together A summary of how differ-
with equal frequency. Rather, certain codons are present ent sources of information are combined to create the
much more frequently in mRNAs (and hence in the DNA best-possible mRNA and gene predictions is depicted in
that encodes them). For example, in D. melanogaster, of Figure 14-14 . These different kinds of evidence are com-
the two codons for cysteine, UGC is used 73 percent of plementary and can cross-validate one another. For exam-
the time, whereas UGU is used 27 percent. This usage is ple, the structure of a gene may be inferred from evidence
Predictions
BLAST similarity
from protein
Codon bias
Predictions Sequence
from mRNA motif
and its
properties EST
cDNA
Predictions
from binding- Promoter Splice Translation Splice sites Translation Polyadenylation
site analysis site sites start site termination site site
programs
Predicted gene
of protein similarity within a region of genomic DNA 14.4 THE STRUCTURE OF THE
bounded by 5′ and 3′ ESTs. Useful predictions are possible
even without a cDNA sequence or evidence of protein sim-
HUMAN GENOME
ilarities. A binding-site-prediction program can propose a
LO 14.2 Explain the role of various functional elements
hypothetical ORF, and proper codon bias would be sup-
within genomes, and differentiate between
porting evidence. computational and experimental methods used to
identify these elements.
KEY CONCEPT Predictions of mRNA and polypeptide struc- LO 14.5 Outline reverse genetic approaches to analyze
ture from genomic DNA sequence depend on the integration the function of genes and genetic elements
of information from cDNA and EST sequence, binding-site pre- identified by genome sequencing and comparative
dictions, polypeptide similarities, and codon bias. genomics.
ancient transposable elements that are now immobile and have been reverse-transcribed from RNA and randomly
have accumulated random mutations, causing them to inserted into the genome. Seventy percent or so of human
diverge in sequence from the ancestral transposable ele- pseudogenes appear to be of this type. Most of the other
ments. Thus, much of the human genome appears to be pseudogenes in the human genome appear to have arisen
composed of genetic “hitchhikers.” from gene duplication events in which one of the dupli-
Only a small part of the human genome encodes poly- cates has acquired one or more ORF-disrupting mutations
peptides; that is, somewhat less than 3 percent of it encodes in the course of evolution. As the challenges in annotation
exons of mRNAs. Exons are typically small (about 150 have been overcome, the estimated number of genes in the
bases), whereas introns are large, many extending more human genome has dropped steadily. A recent estimate is
than 1000 bases and some extending more than 100,000 that there are about 20,000 protein-coding genes.
bases. Transcripts are composed of an average of 10 exons, The annotation of the human genome progressed as the
although many have substantially more. Finally, introns sequences of each chromosome were finished one by one.
may be spliced out of the same gene in locations that These sequences then became the searching ground in the
vary. This variation in the location of splice sites generates hunt for candidate genes for human diseases. An example
considerable added diversity in mRNA and polypeptide of gene predictions for a chromosome from the human
sequence. On the basis of current cDNA and EST data, at genome is shown in Figure 14-15. Such predictions are being
least 60 percent of human protein-coding genes are likely revised continually as new data become available. The cur-
to have two or more splice variants. On average, there are rent state of the predictions can be viewed at many Web
several splice variants per gene. Hence, the number of dis- sites, most notably at the public DNA databases in the
tinct proteins encoded by the human genome is several-fold United States and Europe (see Appendix B). These predic-
greater than the number of recognized genes. tions are the current best inferences of the protein-coding
genes present in the sequenced species and, as such, are
KEY CONCEPT Only a small proportion of the human works in progress.
genome consists of protein-coding genes.
Chromosome 20
FIGURE 14-15 Numerous genes have been identified on human panel. [Reprinted with permission from Macmillan Publishers Ltd: from Jim
chromosome 20. The recombinational and cytogenetic map coordinates Kent, Ewan Birney, Darryl Leja, and Francis Collins, After the International
are shown in the top lines of the figure. Various graphics depicting gene Human Genome Sequencing Consortium, “Initial Sequencing and Analysis of
density and different DNA properties are shown in the middle sections. the Human Genome,” Nature, 2001, February 15; 409, 860–921, Figure 9,
The identifiers of the predicted genes are shown at the bottom of the pdf 1. Permission conveyed through Copyright Clearance Center, Inc.]
little over 1 percent of the total genome DNA, encode pro- approximately 500,000 potential enhancers and promot-
tein sequences. So, nearly 99 percent of our genome does ers associated with known genes. The project also detected
not encode proteins. How do we identify other functional transcripts emanating from nearly 80 percent of the human
parts of the genome? genome.
Introns and 5′ and 3′ untranslated sequences are read- This is a much larger fraction of the genome than was
ily annotated by analysis of gene transcripts, while gene expected. After all, as stated earlier, only a little over 1 per-
promoters are usually identified by their proximity to tran- cent of the genome is protein-coding sequence. However,
scription units and signature DNA sequences. However, the production of a transcript does not necessarily mean
other regulatory sequences such as enhancers are not iden- that the transcript contributes to human biology. It is pos-
tifiable by mere inspection of DNA sequences, and other sible that some proportion of these transcripts represent
sequences that encode various kinds of RNA transcripts “noise” in the cell—transcripts that have no biological
(microRNAs, small interfering RNAs, piwi-interacting function, but also do no harm. It is not sound to ascribe
RNAs, long noncoding RNAs; see Chapter 8) require function to a sequence without some form of additional
detection and annotation of their transcripts. While many data, so what kinds of additional data can be used to
such noncoding elements have been identified in the course resolve questions of function?
of the study of human molecular genetics, the potentially Evolutionary conservation of sequences has proven to
vast number of such elements warrants a more systematic be a good indicator of biological function. Sequences will
approach. The Encyclopedia of DNA Elements (ENCODE) not be preserved over evolutionary time unless mutations
project was thus launched with the ambitious goal of iden- that alter them are weeded out by natural selection. One
tifying all functional elements within the human genome. way to locate potentially functional noncoding elements
This large-scale collaborative endeavor has employed a then is to look for conserved sequences, which have not
diverse array of techniques to detect sequences potentially changed much over millions of years of evolution.
involved in the control of gene transcription, as well as all For example, one can search for very highly conserved
transcribed regions. Because such sequences are expected sequences of modest length among a few species or for
to be active in only individual or subsets of cell types, less perfectly conserved sequences of greater length among
researchers studied 147 human cell types. By searching for a larger number of species. Comparisons of the human,
regions that were associated with the binding of transcrip- rat, and mouse genomes have led to the identification of
tion factors, the ENCODE project estimated that there are so-called ultraconserved elements, which are sequences that
are perfectly conserved among the three
species. Searches of these genomes have
Testing the role of a conserved element in gene regulation found more than 5000 sequences of more
than 100 bp and 481 sequences of more
(a) (b) than 200 bp that are absolutely conserved.
Nearly all of these elements were highly
conserved in the chicken genome, and
about two-thirds were also conserved in a
fish genome. Although many of these ele-
ments are found in gene-poor regions, they
are most richly concentrated near regula-
tory genes important for development. The
majority of highly conserved noncoding
elements may largely take part in regulat-
ing the expression of the genetic toolkit for
the development of mammals and other
vertebrates (see Chapter 13).
How can we verify that such conserved
elements play a role in gene regulation?
These elements can be tested in the same
FIGURE 14-16 A transcriptional cis-acting regulatory element is identified in an manner as the transcriptional cis-acting
ultraconserved element of the human genome. An ultraconserved element lying near the
human ISL1 gene was coupled to a reporter gene and injected into fertilized mouse oocytes.
regulatory elements examined in earlier
The regions where the gene is expressed are stained dark blue or black. (a) The reporter gene chapters, with the use of reporter genes
is expressed in the head and spinal cord of a transgenic mouse, as seen here on day 11.5 of (see Figure 13-18). A researcher places
gestation. This expression pattern corresponds to (b) the native pattern of expression of the candidate regulatory regions adjacent to
mouse ISL1 gene on day 11.5 of gestation. This experiment demonstrates how functional a promoter and reporter gene and intro-
noncoding elements can be identified by comparative genomics and tested in a model
organism. [Reprinted with permission from Macmillan Publishers Ltd: from G. Bejerono et al. “A
duces the reporter gene into a host species.
distal enhancer and an ultraconserved exon are derived from a novel retroposon” Nature, 2006, One such example is shown in Figure 14-16.
April 16; 441: 87–90. Figure 3. Permission conveyed through Copyright Clearance Center, Inc.] An element that is highly conserved among
mammalian, chicken, and a frog species lies 488 kb from we will examine how differences are identified among indi-
the 3′ end of the human ISL1 gene, which encodes a protein vidual humans.
required for motor-neuron differentiation. This element
was placed upstream of a promoter and the β-galactosidase Phylogenetic inference
(lacZ) reporter gene, and the construct was injected into
The first step in comparing species’ genomes is to decide
the pronuclei of fertilized mouse oocytes (see Figure 10-25).
which species to compare. In order for comparisons to be
The reporter protein is expressed along the spinal cord and
informative, it is crucial to understand the evolutionary
in the head, as one would expect for the location of future
relationships among the species to be compared. The evo-
motor neurons (see Figure 14-16). Most significantly, the
lutionary history of a group is called an evolutionary tree,
expression pattern corresponds to part of the expression
or a phylogeny. Phylogenies are useful because they allow
pattern of the native mouse ISL1 gene (presumably other
us to infer how species’ genomes have changed over time.
noncoding elements control the other features of ISL1
The second step in comparing genomes is the identifi-
expression). The expression pattern strongly suggests that
cation of the most closely related genes, called homologous
the conserved element is a regulatory region for the ISL1
genes ( Figure 14-17 ). These genes can be recognized by
gene in each species. The success of this approach sug-
similarities in their DNA sequences and in the amino acid
gests that many additional human noncoding regulatory
sequences of the proteins they encode. It is important to
elements will likely be identified on the basis of sequence
distinguish here two classes of homologous genes. Some
conservation and the activity of those elements in reporter
homologs are genes at the same genetic locus in different
assays.
species. These genes would have been inherited from a
common ancestor and are referred to as orthologs. How-
KEY CONCEPT Noncoding regulatory elements can be iden-
tified through a combination of computational approaches and ever, many homologous genes belong to families that have
reporter gene assays. expanded (and contracted) in number in the course of evo-
lution. These homologous genes are at different genetic
loci in the same organism. They arose when genes within
a genome were duplicated. Genes that are related by
14.5 THE COMPARATIVE gene-duplication events in a genome are called paralogs.
GENOMICS OF HUMANS The history of gene families can be quite revealing about
WITH OTHER SPECIES the evolutionary history of a group.
For example, suppose we would like to know how
LO 14.3 Infer the evolutionary direction of genomic changes the mammalian genome has evolved over the history of
among species based on their phylogenetic the group. We would like to know whether mammals as
relationships.
LO 14.5 Outline reverse genetic approaches to analyze
Relationships between homologs, orthologs, and paralogs
the function of genes and genetic elements
identified by genome sequencing and comparative Ancestral gene
genomics.
a group might have acquired some unique genes, whether that it contains one egg-yolk gene called vitellogenin. Anal-
mammals with different lifestyles might possess different yses of marsupial and eutherian genomes revealed no such
sets of genes, and what the fate was of genes that existed in functional yolk genes. The presence of vitellogenin in the
mammalian ancestors. platypus and its absence from other mammals could be
Fortunately, we now have a large and rapidly expand- explained in one of two ways: (1) vitellogenin is a novel
ing set of mammal genome sequences to compare that invention of the platypus, or (2) vitellogenin existed in a
includes representatives of the three main branches of common ancestor of monotremes, marsupials, and eutheri-
mammals—monotremes (for example, platypus), marsupi- ans but was subsequently lost from marsupials and euthe-
als (for example, wallaby, opossum), and eutherian mam- rians. The direction of evolutionary change is opposite in
mals (for example, human, chimpanzee, dog, mouse). The these two alternatives.
relationships between these groups, some members within A simple pair-wise comparison between the platypus
these groups, and other amniote vertebrates (amniotes are and another mammal does not distinguish between these
mostly land-dwelling vertebrates that have a terrestrially alternatives. To do that, first we have to infer whether
adapted egg) are shown in Figure 14-18. vitellogenin was likely to be present in the last common
To illustrate the importance of understanding phylog- ancestor of the platypus, marsupials, and eutherians. We
enies and how to utilize them, we consider the platypus make this phylogenetic inference by examining whether
genome. Monotremes differ from other mammals in that vitellogenin is found in taxa outside of this entire group of
they lay eggs. Inspection of the platypus genome revealed mammals, what is referred to as an evolutionary outgroup.
Turtles
Sauropsids Lepidosaurs
Tuatara
280
MYA 252 MYA
Lizards,
snakes
Crocodilians
237 MYA
Amniotes Archosaurs Birds
312
MYA
Monotremes
159 MYA
Humans
Mammals
96 90 6.7
MYA MYA MYA
Chimpanzees
Eutherians
Mice
Dogs
FIGURE 14-18 The phylogenetic tree depicts the evolutionary relationships among the three major I N TER A C TI VE
groups of mammals (monotremes, marsupials, and eutherians) and other amniotes, including birds R ES O U R C E
and various reptiles. By mapping the presence or absence of genes in particular groups onto known
phylogenies, one can infer the direction of evolutionary change (gain or loss) in particular lineages.
Understanding evolutionary
trees
Indeed, three homologous vitellogenin genes exist in the this section, the identities of those pseudogenes reflect how
chicken. Next, we consider the relationship of the chicken human biology has diverged from that of our ancestors.
to mammals. Chickens belong to another major branch Of course, evolution is also about the acquisition of new
of the amniotes. Looking at the evolutionary tree in traits. For example, milk production is a shared trait among
Figure 14-18, we can explain the presence of vitellogenins all mammals. A family of genes encoding the casein milk pro-
in chickens and the platypus as the result of two indepen- teins are unique to mammals and tightly clustered together in
dent acquisitions (in the platypus lineage and the chicken their genomes, including that of the platypus. Just this brief
lineage, respectively) or as the result of just one acquisition glance at a few mammalian genomes informs us that, indeed,
in a common ancestor of the platypus and chicken (which, some mammals have genes that others do not, some genes
based on the tree, would be a common ancestor of all amni- are shared by all mammals, and the presence or absence of
otes) followed by the loss of vitellogenin genes in marsupi- certain genes correlates with mammals’ lifestyle. The latter is
als and eutherians. a pervasive finding in comparative genomics.
How do we decide between these alternatives? When
studying infrequent events such as the invention of a gene, KEY CONCEPT Determining which genomic elements have
evolutionary biologists prefer to rely on the principle of been gained or lost during evolution requires knowledge of the
parsimony, that is, to favor the simplest explanation involv- phylogeny of the species being compared. The presence or
absence of genes often correlates with organism lifestyles.
ing the smallest number of evolutionary changes. Therefore,
the preferred explanation for the pattern of vitellogenin
evolution in mammals is that this egg-yolk protein and cor- Let’s look at a few more examples that illuminate the
responding gene were present in some egg-laying amniote evolutionary history of our genome and how we are differ-
ancestor and were retained in the egg-laying platypus and ent from, and similar to, other mammals.
lost from non-egg-laying mammals.
As it turns out, there is one additional and very com- Of mice and humans
pelling piece of evidence that supports this inference. The sequence of the mouse genome has been particularly
While inspection of eutherian genomes does not reveal informative for understanding the human genome because
any intact, functional vitellogenin genes, traces of vitello- of the mouse’s long-standing role as a model genetic species,
genin gene sequences are detectable in the human and dog the vast knowledge of its classical genetics, and the mouse’s
genomes at positions that are in the same position as (syn- evolutionary relationship to humans. The mouse and
tenic to) the vitellogenin genes of the platypus and chicken human lineages diverged approximately 90 million years
(Figure 14-19). These sequences are molecular relics of our ago, which is sufficient time for mutations to cause their
egg-laying ancestors. As our mammalian ancestors shifted genomes to differ, on average, at about one of every two
away from yolky eggs, natural selection was relaxed on the nucleotides. Thus, sequences common to the mouse and
vitellogenin gene sequences such that they have been nearly human genomes are likely to indicate common functions.
eroded away by mutations over tens of millions of years. Homologs are identified because they have similar DNA
Our genome contains numerous relics of genes that once sequences. Analysis of the mouse genome indicates that the
functioned in our ancestors, and as we will see again in number of protein-coding genes that it contains is similar to
Chicken
Human
Monotreme
FIGURE 14-19 Strings of genes along chicken chromosome 8 and human chromosome 1 and in
the platypus are in the same relative order (boxes). Whereas the chicken genome has three genes that
encode egg-yolk proteins, the egg-laying platypus has one functional gene and two pseudogenes, and
humans have fragmented, very short remnants of the yolk genes.
that of the human genome. Further inspection of the mouse The mouse and human genomes have large
genes reveals that at least 99 percent of all mouse genes have syntenic blocks of genes in common
some homolog in the human genome and that at least 99 per-
cent of all human genes have some homolog in the mouse
genome. Thus, the kinds of proteins encoded in each genome
are essentially the same. Furthermore, about 80 percent of
all mouse and human genes are clearly identifiable orthologs.
The similarities between the genomes extend well beyond
the inventory of protein-coding genes to overall genome
organization. More than 90 percent of the mouse and human
genomes can be partitioned into corresponding regions of
conserved synteny, where the order of genes within variously Scale (Mb)
sized blocks is the same as their order in the most recent com- 0 20 40
mon ancestor of the two species. This synteny is very help-
ful in relating the maps of the two genomes. For example,
human chromosome 17 is orthologous to a single mouse
chromosome (chromosome 11). Although there have been
extensive intrachromosomal rearrangements in the human
chromosome, there are 26 segments of collinear sequences
more than 100 kb in size (Figure 14-20).
lineage. Genome sequencing has revealed that there are about explain the dramatic differences in morphology, behavior,
35 million single-nucleotide differences between chimpanzees and physiology between chimpanzees and humans? In 1975,
and humans, corresponding to about a 1.06 percent degree of well before the advent of whole-genome sequencing, Mary-
divergence. In addition, about 5 million insertions and dele- Claire King and Allan Wilson boldly proposed that most of
tions, ranging in length from just a single nucleotide to more the phenotypic differences between humans and chimpanzees
than 15 kb, contribute a total of about 90 Mb of divergent result from mutations that affect gene regulation. Compara-
DNA sequence (about 3 percent of the overall genome). Most tive genomics has now provided a tool to identify regulatory
of these insertions or deletions lie outside of coding regions. mutations that might be responsible for the phenotypic dif-
Overall, the proteins encoded by the human and chim- ferences between chimpanzees and humans.
panzee genomes are extremely similar. Twenty-nine percent
of all orthologous proteins are identical in sequence. Most KEY CONCEPT Great phenotypic differences can evolve
proteins that differ do so by only about two amino acid from genomes containing similar sets of genes. Many of the
replacements. There are a few detectable differences between phenotypic differences between species are likely due to
chimpanzees and humans in the sets of functional genes. genetic changes that affect gene regulation.
About 80 or so genes that were functional in their common
ancestor are no longer functional in humans, owing to their Here, we will discuss just one example of the many
deletion or to the accumulation of mutations. Some of these approaches used to identify putative cis-acting regulatory ele-
changes may contribute to differences in physiology. ments that differ between chimpanzees and humans. In this
In addition to changes in particular genes, duplications case, researchers searched for noncoding sequences that were
of chromosome segments in a single lineage have contributed highly conserved in the genomes of chimpanzee, macaque, and
to genome divergence. More than 170 genes in the human other mammals but missing in the human genome. There were
genome and more than 90 genes in the chimpanzee genome 510 such deletions in the human genome, and these deletions
are present in large duplicated segments. These duplications were enriched near genes with neural function as well as steroid
are responsible for a greater amount of the total genome diver- hormone signaling. One of these deletions is near the androgen
gence than all single-nucleotide mutations combined. Intrigu- receptor gene, which encodes a protein necessary for responses
ingly, duplications unique to the human genome are enriched to circulating androgens such as testosterone. Using the previ-
for genes that are predicted to play a role in brain develop- ously introduced reporter gene assays in transgenic mice (see
ment. It has been suggested that at least some of these gene Chapter 10), the researchers showed that both the mouse and
duplications were involved in the expansion of the neocortex chimpanzee sequence drove expression in the developing sen-
in humans relative to other primates. However, whether these sory vibrissae (or whiskers) as well as the penile spines, which
duplicated genes contribute to major phenotypic differences are both androgen-responsive structures that are present in
between humans and our closest relatives is not yet clear. most mammals but have been lost in humans (Figure 14-21).
Despite the existence of these few differences in gene con- Testing the functions of other putative cis-acting regulatory ele-
tent between chimpanzees and humans, we have seen that ments missing in the human genome will likely uncover addi-
the vast majority of genes are highly conserved, with very tional insight into the genetic changes that underlie differences
few changes in protein-coding regions. How, then, can we between humans and our closest relatives.
Testing the role of a conserved enhancer that has been deleted in humans
Androgen Vibrissae Penile Chimp enhancer Mouse enhancer
receptor spines
Human − −
Chimp + +
Macaque + +
Mouse + +
FIGURE 14-21 The androgen receptor coding sequence (pink) is present assay in mice shows that this sequence is a cis-acting enhancer that drives
in the human, chimp, macaque, and mouse genomes. Although some expression in the sensory vibrissae (arrows) and penile spines (not shown).
conserved, noncoding sequences near the androgen receptor gene are [Reprinted by permission from Macmillan Publishers Ltd. from McLean et al.,
present in the genomes of all four species (yellow, green, and light pink “Human-specific loss of regulatory DNA and the evolution of human-specific
rectangles), one conserved, noncoding sequence is present in the chimp, traits,” Nature, 2011, March 10; 471, 216–219, Figure 2. Permission conveyed
macaque, and mouse genomes but absent from the human genome (red through Copyright Clearance Center, Inc.]
rectangle). Using either the chimp or mouse sequence in a reporter gene
KEY CONCEPT Genetic changes that underlie phenotypic increased copy number appears to have been adaptive con-
differences between humans and our closest relatives can be cerns diet. People with high-starch diets have, on average,
identified using a combination of computational approaches more copies of a salivary amylase (an enzyme that breaks
and reporter gene assays. down starch) gene than people with traditionally low-
starch diets. In other cases, copy number variations have
been associated with human diseases. For example, it now
Of course, all genetic differences between species orig- appears that at least 15 percent of human neurodevelop-
inate as variations within species. The sequencing of the mental diseases are due to changes in copy number that
human genome and the advent of faster and less expensive are found at a very low frequency in human populations.
high-throughput sequencing methods have opened the door Copy number polymorphisms that are relatively common
to the detailed analysis of human genetic variation. in human populations have also been associated with
immune-related diseases such as Crohn’s disease, psoriasis,
and lupus.
14.6 COMPARATIVE GENOMICS
AND HUMAN MEDICINE The evolutionary history of human
disease genes
LO 14.4 Compare genomic methods used to identify
mutations that have been associated with human One might ask when and where the mutations that cause
disease thus far. human disease originated, and why some of these disease
alleles are maintained at a relatively high frequency in
The human species, Homo sapiens, originated in Africa human populations. Although we are still a long way from
approximately 200,000 years ago. Sometime between answering these questions, some insight has come from ana-
50,000 and 100,000 years ago, populations left Africa and lyzing the genome sequences of ancient hominins, including
migrated across the world, eventually populating five addi- those of our own species Homo sapiens as well as now-
tional continents. These migrating populations encountered extinct, archaic hominin lineages like Neanderthals.
different climates, adopted different diets, and combated Advances in sequencing and other technologies have made
different pathogens in different parts of the world. Much of it possible to extract and sequence whole genomes from
the recent evolutionary history of our species is recorded in ancient DNA samples, even when very small amounts of
our genomes, as are the genetic differences that make indi- tissue are found. For example, whole genome sequencing
viduals or populations more or less susceptible to disease. of ancient DNA from a single finger bone and three teeth
Overall, any two unrelated humans’ genomes are 99.9 per found in the Denisova Cave in Siberia revealed the exis-
cent identical. That difference of just 0.1 percent still cor- tence of an archaic hominin lineage, now called Deniso-
responds to roughly 3 million bases. The challenge today is vans, which are genetically very distinct from Neanderthals,
to decipher which of those base differences are meaningful diverging approximately 640,000 years ago.
with respect to physiology, development, or disease. Analyses of these archaic genomes has revealed that
Once the sequence of the first human genome was as anatomically modern humans left Africa and spread
advanced, that accomplishment opened the door to much around the globe, they interbred with other hominin species
more rapid and less costly analysis of other individuals. The that had already been living in Eurasia for over 200,000
reason is that with a known genome assembly as a refer- years. Traces of these hybridization events can be seen in
ence, it is much easier to align the raw sequence reads of the genomes of humans living today. Reflecting the migra-
additional individuals, and to design approaches to study- tory paths of modern humans, all non-African individuals
ing and comparing parts of the genome. sequenced to date have between 1 percent and 4 percent
One of the first and greatest surprises that has emerged Neanderthal ancestry, while indigenous Australians and
from comparing individual human genomes is that humans Melanesians also have up to 6 percent Denisovan ancestry
differ not merely at one base in a thousand, but also in the (Figure 14-22). Many direct-to-consumer genetic testing ser-
number of copies of parts of individual genes, entire genes, vices will now report what percentage of a person’s DNA
or sets of genes. These copy number variations (CNVs) has been inherited from their archaic human ancestors.
include repeats and duplications that increase copy num- Box 14-1 discusses the various types of direct-to-consumer
ber and deletions that reduce copy number. Between any genetic testing options available today, as well as some
two unrelated individuals, there may be hundreds of seg- important ethical and social implications of such services.
ments of DNA greater than 1000 bp in length that differ in Remarkably, some of the Neanderthal and Deniso-
copy number. Some CNVs can be quite large and span up van alleles present in modern humans have effects on phys-
to 5 million base pairs. Together, CNVs account for more iology. For example, gene variants that cause lighter skin in
sequence variation among humans than all the 3 million northern Eurasians were present in Neanderthals, and one
single base pair changes combined. of the gene variants that has enabled Tibetans to live at high
How such copy numbers may play a role in human altitudes (see Chapter 1) is Denisovan in origin. However,
evolution and disease is of intense interest. One case where Neanderthal-derived alleles of some genes are associated with
Sequencing ancient DNA of archaic hominins reveals interbreeding with the ancestors of modern humans
(a) (b) (c)
Denisovan
Denisovan
range European
Modern
humans East Asian
Melanesian
African
~800 ~600 ~100
Time
(thousand years ago)
FIGURE 14-22 (a) Side view of a molar found in the Denisova Cave in Siberia. (b) Anatomically
modern humans migrated out of Africa into Eurasia through the ranges of both Neanderthals and
Denisovans. (c) Genome sequencing provides evidence for early hybridization between Neanderthals
and the ancestors of modern Melanesians, East Asians, and Europeans, as well as later hybridization
between Neanderthals and the ancestors of modern East Asians (blue arrows). There is also evidence
for hybridization between Denisovans and the ancestors of modern Melanesians (green arrow).
[Courtesy of the Max Planck Institute for Evolutionary Anthropology.]
the risk of diseases in modern humans. For example, modern exon sequences ( Figure 14-23). The DNA is prepared by
humans with a Neanderthal allele at a gene involved in blood (1) shearing genomic DNA into short, single-stranded pieces,
coagulation have a higher risk of blood clots and stroke. Rapid (2) hybridizing the single-stranded pieces to biotin-labeled
clotting might have been an advantage in early hominids, who probes complementary to the known exonic regions of the
were hunting dangerous animals and at risk of excessive bleed- human genome and purifying the biotin-labeled duplexes,
ing during childbirth. However, in our modern times, these (3) amplifying the exon-rich duplexes, and (4) sequencing
risks are lessened, and humans live much longer. Thus, fast the exon-rich duplexes. In this manner, 30–60 megabases
clotting is no longer an advantage and leads to an increased of the human genome is targeted for sequencing, as opposed
risk of stroke and blood clots. Genetic variants of Neander- to the 3200 megabases of total sequence.
thal origin are also linked to increased risk of neurological, As of late 2017, the exomes of more than 120,000
immunological, and skin diseases in modern humans. These individuals have been sequenced, at the current cost of
examples serve as a reminder that our genetic susceptibility to only a few hundred dollars per exome. One particularly
disease has been shaped by our evolutionary history. Genomics important power of exome sequencing is to identify de
has provided a tool that allows us to explore this evolutionary novo mutations in individuals (mutations that are not
history in ways that could not be imagined before. present in either parent). Such mutations are responsi-
ble for many spontaneously appearing genetic diseases
KEY CONCEPT The ability to sequence whole genomes of whose origins would not be revealed by traditional
modern and archaic humans provides a tool to uncover the pedigree-based studies. As such, whole-exome sequencing
evolutionary history of humans and to identify mutations asso- is now a rapidly spreading clinical diagnostic tool, par-
ciated with disease. ticularly for neurodevelopmental and other disorders in
pediatric populations.
And just as exome sequencing can be used to identify
The exome and personalized genomics genetic differences between individuals, it can also be used
Advances in sequencing technologies have reduced the cost to identify differences between normal and abnormal cells,
of sequencing individual genomes from about $300 mil- such as cancer cells. Cancer is a suite of genetic diseases in
lion in 2000, to $1 million in 2008, to about $1000 in which combinations of gene mutations typically contribute
2017. But for many large-scale studies, that figure is still to the loss of growth control and metastasis. Understanding
prohibitive. For some applications, it is more practical and what genetic changes are common to particular cancers, or
cost effective, and can be just as informative, to sequence to subsets of cancers, will not only further our understand-
only part of the genome. For example, since many disease- ing of cancer, but also promises to impact diagnosis and
causing mutations occur in coding sequences, strategies have treatment in powerful ways. Researchers across the world
been designed to sequence all of the exons, or the exome, of have recently completed an “atlas” of cancer genomes that
individuals, as was done in the case of Nicholas Volker. has uncovered the extraordinary genetic heterogeneity
The strategy for exome sequencing involves gen- present in cancer cells and provided a framework for clas-
erating a library of genomic DNA that is enriched for sifying tumor subtypes based on the underlying genomic
alterations. This knowledge opens up new opportunities to The ability to rapidly analyze organisms’ genomes is
develop therapies that specifically target the genetic changes also impacting other dimensions of medicine. We will look
found in a particular tumor rather than treating cancer as at one such case next.
a homogeneous disease. (See http://cancergenome.nih.gov/
for further information.) Comparative genomics of nonpathogenic
KEY CONCEPT Exome sequencing is a powerful approach and pathogenic E. coli
to cheaply and rapidly identify mutations associated with
Escherichia coli are found in our mouths and intestinal
human disease.
tracts in vast numbers, and this species is generally a benign
A B C D
a
d
b
c
are interspersed with islands of genes specific to either K-12 14.7 FUNCTIONAL GENOMICS
or O157 (Figure 14-24).
Among the 1387 genes specific to E. coli O157 are
AND REVERSE GENETICS
many genes that are suspected to encode virulence factors,
including toxins, cell-invasion proteins, adherence proteins, LO 14.2 Explain the role of various functional elements
within genomes, and differentiate between
and secretion systems for toxins, as well as possible met- computational and experimental methods used to
abolic genes that may be required for nutrient transport, identify these elements.
antibiotic resistance, and other activities that may con- LO 14.5 Outline reverse genetic approaches to analyze
fer the ability to survive in different hosts. Most of these the function of genes and genetic elements
genes were not known before sequencing and would not be identified by genome sequencing and comparative
known today had researchers relied solely on E. coli K-12 genomics.
as a guide to all E. coli.
The surprising level of diversity between two members Geneticists have been studying the expression and inter-
of the same species shows how dynamic genome evolu- actions of individual gene products for the past several
tion can be. Most new genes in E. coli strains are thought decades. With the advent of genomics, we have an oppor-
to have been introduced by horizontal transfer from the tunity to expand these studies to a global level by using
genomes of viruses and other bacteria (see Chapter 6). genome-wide approaches to study most or all gene prod-
Differences can also evolve owing to gene deletion. Other ucts systematically and simultaneously, and in species that
pathogenic E. coli and bacterial species also exhibit many are not previously established experimental models (see
differences in gene content from their nonpathogenic cous- the Beyond Model Organisms section of A Brief Guide to
ins. The identification of genes that may contribute directly Model Organisms, at the back of this book). This global
to pathogenicity opens new avenues to the understanding, approach to the study of the function, expression, and
prevention, and treatment of infectious disease. interaction of gene products is termed functional genomics.
“ ’Omics” present in the cell can tell us what genes are active. Here, the
application of next-generation sequencing technologies has
In addition to the genome, other global data sets are of
been extremely powerful by permitting the assay of RNA
interest. Following the example of the term genome, for
transcripts for all genes simultaneously in a single experi-
which “gene” plus “-ome” becomes a word for “all genes,”
ment. Let’s see how this process works in more detail.
genomics researchers have coined a number of terms to
The first step is to isolate the total set of RNA molecules
describe other global data sets on which they are working.
from cells of interest. For example, one set might be extracted
This ’ome wish list includes
from a particular cell type grown under typical conditions.
The transcriptome. The sequence and expression pat- A second set might be made from RNA extracted from cells
terns of all RNA transcripts (which kinds, where in tis- grown under some experimental condition. Although methods
sues, when, how much). exist for capturing and sequencing different types of RNAs in
The proteome. The sequence and expression patterns of the cell, we will focus here on the sequencing of mRNA, which
all proteins (where, when, how much). is the fraction of the RNA that encodes proteins. The mRNA
can be captured from total RNA using an oligo-dT primer,
The interactome . The complete set of physical inter-
which is complementary to the 3′ poly(A) tail of the mRNA.
actions between proteins and DNA segments, between
Then, the mRNA is subjected to reverse transcription to trans-
proteins and RNA segments, and between proteins.
form it into cDNA (see Chapter 10), which can then be used as
We will not consider all of these ’omes in this section but a substrate for next-generation sequencing libraries just as for
will focus on some of the global techniques that are begin- genomic DNA (see Figure 14-6). The sequencing reads are then
ning to be exploited to obtain these data sets. mapped to the genome, where they align to the transcribed
regions of genes. The number of reads present for a particular
Using RNA-seq to study the transcriptome Suppose we transcript should reflect its levels of expression in the cell; genes
want to answer the question, what genes are active in a par- expressed at a low level in a particular cell type will have few
ticular cell under certain conditions? Those conditions could reads, and genes expressed at a high level in a particular cell
be one or more stages in development, or they could be the type will have many reads (Figure 14-25). In this manner, genes
presence or absence of a pathogen or a hormone. Active genes whose levels of expression are increased or decreased under
are transcribed into RNA, and so the set of RNA transcripts the given experimental condition are identified. Similarly, genes
Fragmentation
Mapping to genes
that are active in a given cell type or at a given stage of develop- start site and (2) an activation domain that will activate
ment can be identified. transcription but cannot itself bind to DNA. Thus, the
With an understanding of which genes are active or two domains must be in close proximity in order for tran-
inactive at a given developmental stage, in a particular cell scriptional activation to take place. Suppose that you are
type, or in various environmental conditions, the sets of investigating whether two proteins interact. The strategy
genes that may respond to similar regulatory inputs can be of the two-hybrid system is to separate the two domains
identified. Furthermore, gene-expression profiles can paint of the activator encoded by GAL4, making activation of
a picture of the differences between normal and diseased a reporter gene impossible. Each domain is connected to a
cells. By identifying genes whose expression is altered by different protein. If the two proteins interact, they will join
mutations, in cancer cells, or by a pathogen, researchers the two domains together. The activator will become active
may be able to devise new therapeutic strategies. and start transcription of the reporter gene.
How is this scheme implemented in practice? The GAL4
Using the two-hybrid test to study the protein– gene is divided between two plasmids so that one plasmid
protein interactome One of the most important activ- contains the part encoding the DNA-binding domain and
ities of proteins is their interaction with other proteins. the other plasmid contains the part encoding the activation
Because of the large number of proteins in any cell, biol- domain. On one plasmid, a gene for one protein under inves-
ogists have sought ways of systematically studying all of tigation is spliced next to the DNA-binding domain, and this
the interactions of individual proteins in a cell. One of fusion protein acts as “bait.” On the other plasmid, a gene
the most common ways of studying the interactome uses for another protein under investigation is spliced next to the
an engineered system in yeast cells called the two-hybrid activation domain, and this fusion protein is said to be the
test , which detects physical interactions between two pro- “target” (Figure 14-26). The two hybrid plasmids are then
teins. The basis for the test is the transcriptional activator introduced into the same yeast cell—perhaps by mating hap-
encoded by the yeast GAL4 gene (see Chapter 12). loid cells containing bait and target plasmids. The final step is
Recall that this protein has two domains: (1) a to look for activation of transcription by a GAL4-regulated
DNA-binding domain that binds to the transcriptional reporter gene construct, which would be proof that bait and
target bind to each other. The two-hybrid system
Studying protein interactions with the use of the can be automated to make it possible to hunt for
yeast two-hybrid system protein interactions throughout the proteome.
Yeast two-hybrid vectors
Studying the protein–DNA interactome
Gal4-binding Gal4-activation
2 ori domain 2 ori domain using chromatin immunoprecipitation assay
Cam R (BD) ampR (AD) (ChIP) The sequence-specific binding of proteins
to DNA is critical for correct gene expression.
For example, regulatory proteins bind to promot-
“Bait” “ Target ” ers and activate or repress transcription in both
protein protein
bacteria and eukaryotes (see Chapters 11, 12, and
13). In the case of eukaryotes, chromosomes are
organized into chromatin, in which the fundamen-
Trp 1+ Leu 2+ tal unit, the nucleosome, contains DNA wrapped
Unite around histones. Post-translational modification
Interaction of histones often dictates what proteins bind and
Target Bait where (see Chapter 12 ). A variety of technolo-
Gal4 AD
gies have been developed that allow researchers
to isolate specific regions of chromatin so that
Gal4 BD DNA and its associated proteins can be analyzed
together. The most widely used method is called
ChIP (for chromatin immunoprecipitation), and
Transcription
its application is described below (Figure 14-27).
Let’s say that you have isolated a gene from
GAL Reporter yeast and suspect that it encodes a protein that
promoter lacZ
binds to DNA when yeast is grown at high tem-
FIGURE 14-26 The system uses the binding of A N I M ATED perature. You want to know whether this protein
two proteins, a “bait” protein and a “target” protein, A RT binds to DNA and, if so, to what yeast sequence.
to restore the function of the Gal4 protein, which One way to address this question is first to treat
activates a reporter gene. Cam, Trp, and Leu are
Yeast two-hybrid systems
components of the selection systems for moving
yeast cells that have been grown at high tempera-
the plasmids around between cells. The reporter gene is lacZ, which resides on a yeast ture with a chemical that will cross-link proteins
chromosome (shown in blue). to the DNA. In this way proteins bound to the
A N I M ATED A RT
ChIP
Antibody
Amplify and
sequence
DNA at the time of chromatin isolation will remain bound Reverse genetics
through subsequent treatments. The next step is to break
The kinds of data obtained from RNA-seq, ChIP-seq, and
the chromatin into small pieces. To separate the fragment
protein-interaction screens are suggestive of interactions
containing your protein–DNA complex from others, you
within the genome and proteome, but they do not allow one
use an antibody that reacts specifically with the encoded
to draw firm conclusions about gene functions and interac-
protein. You add your antibody to the mixture so that it
tions in vivo. For example, finding out that the expression
forms an immune complex that can be purified. The DNA
of certain genes is lost in some cancers is not proof of cause
bound in the immune complex can be analyzed after
and effect. The gold standard for establishing the function
cross-linking is reversed. DNA bound by the protein may
of a gene or genetic element is to disrupt its function and to
be amplified into many copies by PCR to prepare for DNA
understand phenotypes in native conditions. Starting from
sequencing, or the DNA may be sequenced directly.
available gene sequences, researchers can now use a variety
As we saw in Chapter 12, regulatory proteins often acti-
of methods to disrupt the function of a specific gene. These
vate transcription of many genes simultaneously by binding
methods are referred to as reverse genetics. Reverse-genetic
to several promoter regions. A variation of the ChIP proce-
analysis starts with a known molecule—a DNA sequence,
dure, called ChIP-seq , has been devised to identify all the
an mRNA, or a protein—and then attempts to disrupt this
binding sites of a protein in a sequenced genome. Proteins
molecule to assess the role of the normal gene product in
that bind to many genomic regions are immunoprecipitated
the biology of the organism (see Figure 14-2).
as described previously. Then, after cross-linking is reversed,
There are several approaches to reverse genetics, and
the DNA fragments are subjected to DNA sequencing using
new technologies are constantly being developed and
a next-generation method such as Illumina sequencing. The
refined. One approach is to introduce random mutations
sequencing reads are mapped to the genome to reveal the
into the genome, but then to hone in on the gene of inter-
locations where the regulatory protein binds in a particular
est by molecular identification of mutations in the gene. A
cell type, environmental condition, or disease state.
second approach is to conduct a targeted mutagenesis that
produces mutations directly in the gene of interest. A third
KEY CONCEPT Advances in genomic technologies have
made it possible to catalog the transcripts and proteins as well
approach is to create phenocopies—effects comparable to
and protein–DNA and protein–protein interactions found in nor- mutant phenotypes—usually by treatment with agents that
mal and diseased cells. interfere with the mRNA transcript of the gene.
Each approach has its advantages. Random mutagene- revolutionizing the ability to disrupt and manipulate genes
sis is well established, but it requires that one sift through in both model and nonmodel species.
all the mutations to find those that include the gene of Gene-specific mutagenesis usually requires the replace-
interest. Targeted mutagenesis can also be labor intensive, ment of a resident wild-type copy of an entire gene by a
but, after the targeted mutation has been obtained, its mutated version of that gene. The mutated gene inserts
characterization is more straightforward. Creating pheno- into the chromosome by a mechanism resembling homol-
copies can be very efficient, especially as libraries of tools ogous recombination, replacing the normal sequence with
have been developed for particular model species. The tech- the mutant (Figure 14-28). This approach can be used for
nical details of these methods are covered in Chapters 8 targeted gene knockout, in which a null allele replaces the
and 10, so we will here consider examples of each of these wild-type copy. Some techniques are so efficient that in
approaches. E. coli and S. cerevisiae, for example, it has been possible
to mutate every gene in the genome to try to ascertain its
Reverse genetics through random mutagenesis biological function.
Random mutagenesis for reverse genetics employs the same
kinds of general mutagens that are used for forward genet- KEY CONCEPT Targeted mutagenesis is the most precise
ics: chemical agents, radiation, or transposable genetic ele- means of obtaining mutations in a specific gene and can now
ments (see Figure 6-38). However, instead of screening the be practiced in a variety of model systems, including mice
genome at large for mutations that exert a particular phe- and flies.
notypic effect, reverse genetics focuses on the gene in ques-
tion, which can be done in one of two general ways. Reverse genetics by phenocopying The advantage of
One approach is to focus on the map location of the inactivating a gene itself is that mutations will be passed on
gene. Only mutations falling in the region of the genome from one generation to the next, and so, once obtained, a
where the gene is located are retained for further detailed line of mutants is always available for future study. On the
molecular analysis. Thus, in this approach, the recovered other hand, phenocopying can be applied to a great many
mutations must be mapped. One straightforward way is to organisms regardless of how well developed the genetic
cross a new mutant with a mutant containing a known dele- technology is for a given species.
tion or mutation of the gene of interest (see Figure 17-21). One of the most exciting discoveries of the past decade or
Only the pairings that result in progeny with a mutant phe- so has been the discovery of a widespread mechanism whose
notype (showing lack of complementation) are saved for natural function seems to be to protect a cell from foreign
study. DNA. This mechanism is called RNA interference (RNAi),
In another approach, the gene of interest is identified described in Chapter 8. Researchers have capitalized on this
in the mutagenized genome and checked for the presence
of mutations. For example, if the mutagen causes small Disrupting gene function with the use of targeted
deletions, then, after PCR amplification of gene fragments, mutagenesis
genes from the parental and mutagenized genomes can be
compared, looking for a mutagenized genome in which the
gene of interest is reduced in size. Similarly, transposable-
element insertions into the gene of interest can be readily
detected because they increase its size. As the ability to rap- Gene A with
mutant Mutant
idly and cheaply sequence whole genomes improves, it is segment sequence
becoming feasible to identify mutations in genes of interest, 5′ 3′
including single–base-pair substitutions, by simply sequenc-
ing the parental and mutagenized genomes. In these ways, a Chromosome
set of genomes containing random mutations can be effec- 5′ Gene A 3′
tively screened to identify the small fraction of mutations in
Recombination between
a gene of interest to a researcher. mutant transgene
and chromosomal gene
Reverse genetics by targeted mutagenesis For most
of the twentieth century, researchers viewed the ability
Mutant gene A
to direct mutations to a specific gene as the unattainable
Chromosome
“holy grail” of genetics. However, now several such tech- 5′ 3′
niques are available. After a gene has been inactivated in
an individual, geneticists can evaluate the phenotype exhib- FIGURE 14-28 The basic molecular event in targeted gene
ited for clues to the gene’s function. While the tools for replacement. A transgene containing sequences from two ends of a
gene but with a selectable segment of DNA in between is introduced
targeted gene mutations were first developed using genetic into a cell. Double recombination between the transgene and a normal
techniques for model organisms, new technologies, partic- chromosomal gene produces a recombinant chromosomal gene that has
ularly those that are CRISPR-based (see Chapter 10), are incorporated the abnormal segment.
cellular mechanism to make a powerful method for inactivat- Functional genomics with nonmodel organisms
ing specific genes. The inactivation is achieved as follows. A Much of our consideration of mutational dissection and
double-stranded RNA is made with sequences homologous phenocopying has focused on genetic model organisms.
to part of the gene under study and is introduced into a cell One current focus of many geneticists is the broader appli-
(Figure 14-29). The RNA-induced silencing complex, or RISC, cation of these techniques to other species, including those
then degrades native mRNA that is complementary to the that have negative effects on human society, such as parasites,
double-stranded RNA. The net result is a complete or consid- disease carriers, or agricultural pests, or those species that
erable reduction of mRNA levels that lasts for hours or days, are of interest to evolutionary biologists and ecologists (see
thereby nullifying expression of that gene. Because the RISC Chapter 20). Classical genetic techniques are not read-
complex is found in most eukaryotes, the technique has been ily applicable to most of these species, but whole-genome
widely applied in model systems such as C. elegans, Drosoph- sequencing and functional genomics can now be conducted in
ila, zebrafish, and several plant species. any species for which DNA and tissue can be obtained. Fur-
But what makes RNAi especially powerful is that it can thermore, the roles of specific genes can be assessed through
be applied to nonmodel organisms. First, target genes of the generation of phenocopies by RNAi and targeted muta-
interest can be identified by comparative genomics. Then genesis. In particular, the recently developed CRISPR-based
RNAi sequences are produced to target the inhibition of methods for genome engineering are already being used in a
the specific target genes. This technique has been applied, number of nonmodel systems and promise to enable reverse
for example, to a mosquito that carries malaria (Anopheles genetic approaches in nearly any species (see Chapter 10).
gambiae). Using these techniques, scientists can better
understand the biological mechanisms relating to the med- KEY CONCEPT Reverse genetic methods are the gold stan-
ical or economic effect of such species. The genes that con- dard to test the functions of genes and genetic elements dis-
trol the complicated life cycle of the malaria parasite, partly covered through genomic approaches. Recent technological
inside a mosquito host and partly inside the human body, advances mean that these methods can now be practiced in a
can be better understood, revealing new ways to control the variety of model and nonmodel systems.
single most common infectious disease in the world.
SUMMARY
Genomic analysis takes the approaches of genetic analysis The key problem in compiling an accurate sequence of
and applies them to the collection of global data sets to a genome is to take short sequence reads and relate them
fulfill goals such as the mapping and sequencing of whole to one another by sequence identity to build up a consen-
genomes and the characterization of all transcripts and pro- sus sequence of an entire genome. This can be done in a
teins. Genomic techniques require the rapid processing of straightforward way for bacterial or archaeal genomes by
large sets of experimental material, all dependent on exten- aligning overlapping sequences from different sequence
sive automation. reads to compile the entire genome, because few or no DNA
segments are present in more than one copy in such organ- complex genomes of many animals and plants. Comparative
isms. The problem is that complex genomes of plants and genomics can also reveal how genomes have changed in the
animals are replete with such repetitive sequences. These course of evolution and how these changes may relate to dif-
repetitive sequences interfere with accurate sequence-contig ferences in physiology, anatomy, or behavior among species.
production. The problem is resolved in whole-genome shot- Comparisons of modern and archaic human genomes are
gun (WGS) sequencing with the use of paired-end reads. accelerating the discovery of rare disease mutations. In bacte-
Having a genomic sequence map provides the raw, rial genomics, comparisons of pathogenic and nonpathogenic
encrypted text of the genome. The job of bioinformatics is strains have revealed many differences in gene content that
to interpret this encrypted information. For the analysis of contribute to pathogenicity.
gene products, computational techniques are used to iden- Functional genomics attempts to understand the work-
tify ORFs and noncoding RNAs, then to integrate these ing of the genome as a whole system. Two key elements are
results with available experimental evidence for mRNA the transcriptome, the set of all transcripts produced, and
transcript structures (cDNA sequences), protein similarities, the interactome, the set of interacting gene products and
and knowledge of characteristic sequence motifs. other molecules that together enable a cell to be produced
One of the most powerful means to advance the analysis and to function. The function of individual genes and gene
and annotation of genomes is by comparing with the genomes products for which classical mutations are not available can
of related species. Conservation of sequences among species be tested through reverse genetics—by targeted mutation or
is a reliable guide to identifying functional sequences in the phenocopying.
KEY TERMS
annotation (p. 473) genome project (p. 463) pseudogene (p. 477)
bioinformatics (p. 472) genomics (p. 462) reverse genetics (p. 463)
ChIP (chromatin immunoprecipita- homologous gene (p. 479) RNA interference (RNAi) (p. 492)
tion) (p. 490) interactome (p. 489) RNA sequencing (RNA-seq)
ChIP-seq (p. 491) open reading frame (ORF) (p. 473) (p. 474)
comparative genomics (p. 479) ortholog (p. 479) scaffold (p. 471)
consensus sequence (p. 465) outgroup (p. 480) sequence assembly (p. 465)
copy number variation (CNV) paired-end read (p. 470) sequence contig (p. 468)
(p. 484) paralog (p. 479) supercontig (p. 471)
DNA sequencing library (p. 468) parsimony (p. 481) synteny (p. 482)
exome (p. 485) personal genomics (p. 486) transcriptome (p. 489)
expressed sequence tag (EST) phylogeny (p. 479) two-hybrid test (p. 490)
(p. 474) phylogenetic inference (p. 480) whole-genome shotgun (WGS)
forward genetics (p. 463) processed pseudogene (p. 477) sequencing (p. 466)
functional genomics (p. 488) proteome (p. 489)
SOLVED PROBLEMS
SOLVED PROBLEM 1 passage-lining cells and using this RNA for an RNA-seq
You want to study the development of the olfactory experiment. For example, you may choose to first examine
(smell-reception) system in the mouse. You know that mRNAs that are expressed in the nasal-passage lining but
the cells that sense specific chemical odors (odorants) are nowhere else in the mouse as important candidates for a
located in the lining of the nasal passages of the mouse. specific role in olfaction. (Many of the important molecules
Describe some approaches for using functional genomics may also have other jobs elsewhere in the body, but you
and reverse genetics to study olfaction. have to start somewhere.) Alternatively, you may choose to
start with those genes whose protein products are candidate
SOLUTION proteins for binding the odorants themselves. Regardless of
Many approaches can be imagined. For reverse genetics, your choice, the next step would be to engineer a targeted
you would want to first identify candidate genes that are knockout of the gene that encodes each mRNA or protein
expressed in the lining of the nasal passages. Given the of interest or to use RNA interference to attempt to pheno-
techniques of functional genomics, this identification could copy the loss-of-function phenotype of each of the candi-
be accomplished by purifying RNA from isolated nasal- date genes.
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
WORKING WITH THE FIGURES 13. Figure 14-20 shows syntenic regions of mouse chro-
mosome 11 and human chromosome 17. What do
(The first 18 questions require inspection of text figures.)
these syntenic regions reveal about the genome of the
1. You have identified a noncoding sequence that is last common ancestor of mice and humans?
conserved across all mammals, except for primates.
14. Based on Figure 14-22, what percent of Denisovan
You decided to engineer a targeted knockout of this
ancestry do you predict would be found in modern
sequence in mice. Based on Figure 14-2, is this a for-
Western Europeans?
ward or reverse genetic experiment?
15. In Figure 14-23 , what key step enables exome
2. Based on the projection shown in Figure 14-3, what is
sequencing and distinguishes it from whole-genome
the approximate number of human genomes that will
sequencing?
be sequenced by 2025? How many basepairs will this
represent? 16. The genomes of two E. coli strains are compared in
Figure 14-24. Would you expect any third strain to
3. Based on Figure 14-4 , why must the DNA frag-
contain more of the blue, tan, or red regions shown in
ments sequenced overlap in order to obtain a genome
Figure 14-24? Explain.
sequence?
17. In Figure 14-25, why do the mRNA-sequencing reads
4. In Figure 14-6, the color pink indicates the base T, the
map only to parts of the genome? Which gene is more
color orange indicates the base A, the color yellow
highly expressed in sample 1?
indicates the base G, and the color purple indicates the
base C. What is the scanned sequence of the middle 18. Figure 14-26 depicts the Gal4-based two-hybrid system.
cluster in this figure? What is the scanned sequence of Why do the “bait” proteins fused to the Gal4 DNA-bind-
the cluster on the left? ing protein not activate reporter-gene expression?
5. Filling gaps in draft genome sequences is a major chal-
lenge. Based on Figures 14-8 and 14-9, can paired-end BASIC PROBLEMS
reads from a library of 2-kb fragments fill a 10-kb gap? 19. Explain the approach that you would apply to
6. In Figure 14-11 , how are the positions of codons sequencing the genome of a newly discovered bacterial
determined? species.
7. In Figure 14-11, how are the positions of transcrip- 20. Terminal-sequencing reads of clone inserts are a rou-
tional regulatory elements determined? tine part of genome sequencing. How is the central
8. In Figure 14-12, expressed sequence tags (ESTs) are part of the clone insert ever obtained?
aligned with genomic sequence. How are ESTs helpful 21. What is the difference between a contig and a scaffold?
in genome annotation? 22. Two particular contigs are suspected to be adjacent,
9. In Figure 14-12, cDNA sequences are aligned with possibly separated by repetitive DNA. In an attempt
genomic sequence. How are cDNA sequences helpful to link them, end sequences are used as primers to try
in genome annotation? Are cDNAs more important to bridge the gap. Is this approach reasonable? In what
for bacterial or eukaryotic genome annotations? situation will it not work?
10. Based on Figure 14-16 and the features of ultracon- 23. In a genomic analysis looking for a specific dis-
served elements, what would you predict you would ease gene, one candidate gene was found to have a
observe if you injected a reporter-gene construct of the single–base-pair substitution resulting in a nonsynon-
rat ortholog of the ISL1 ultraconserved element into ymous amino acid change. What would you have to
fertilized mouse oocytes and examined reporter gene check before concluding that you had identified the
expression in the developing embryo? disease-causing gene?
11. Based on Figure 14-17, did the duplication that cre- 24. Is a bacterial operator a binding site?
ated the A and B genes occur before or after speciation 25. A sequenced fragment of DNA in Drosophila was used
of the common ancestor of frogs, humans, and mice? in a BLAST search. The best (closest) match was to a
12. Based on Figure 14-18 , are humans more closely kinase gene from Neurospora. Does this match mean
related to mice or to dogs? that the Drosophila sequence contains a kinase gene?
26. In a two-hybrid test, a certain gene A gave positive a. What are two alternative explanations for this
results with two clones, M and N. When M was used, observation?
it gave positives with three clones, A, S, and Q. Clone b. How could you distinguish between these two
N gave only one positive (with A). Develop a tentative possibilities?
interpretation of these results.
31. The platypus is one of the few venomous mammals.
27. You have the following sequence reads from a genomic The male platypus has a spur on the hind foot through
clone of the Drosophila melanogaster genome: which it can deliver a mixture of venom proteins.
Read 1: TGGCCGTGATGGGCAGTTCCGGTG Looking at the phylogeny in Figure 14-18, how would
Read 2: TTCCGGTGCCGGAAAGA you go about determining whether these venom pro-
teins are unique to the platypus?
Read 3: CTATCCGGGCGAACTTTTGGCCG
32. You have sequenced the genome of the bacterium Sal-
Read 4: CGTGATGGGCAGTTCCGGTG monella typhimurium, and you are using BLAST anal-
Read 5: TTGGCCGTGATGGGCAGTT ysis to identify similarities within the S. typhimurium
Read 6: CGAACTTTTGGCCGTGATGGGCAGTTCC genome to known proteins. You find a protein that
is 100 percent identical in the bacterium Escherichia
Use these six sequence reads to create a sequence con- coli. When you compare nucleotide sequences of the
tig of this part of the D. melanogaster genome. S. typhimurium and E. coli genes, you find that their
28. Sometimes, cDNAs turn out to be “chimeras”; that is, nucleotide sequences are only 87 percent identical.
fusions of DNA copies of two different mRNAs acci- a. Explain this observation.
dentally inserted adjacently to each other in the same
clone. You suspect that a cDNA clone from the nema- b. What do these observations tell you about the mer-
tode Caenorhabditis elegans is such a chimera because its of nucleotide- versus protein-similarity searches
the sequence of the cDNA insert predicts a protein in identifying related genes?
with two structural domains not normally observed in 33. If you sequenced the genomes of any two unrelated
the same protein. How would you use the availability humans, what types of sequence changes would you
of the entire genomic sequence to assess if this cDNA expect to find, and how many total base pairs would
clone is a chimera or not? be affected by each type of sequence change?
29. In browsing through the human genome sequence, 34. You have access to both normal cells and cancerous
you identify a gene that has an apparently long coding cells taken from a biopsy from a patient with liver can-
region, but there is a two–base-pair deletion that dis- cer. Describe the genomic approaches you would use
rupts the reading frame. to characterize the differences between these cells.
a. How would you determine whether the deletion 35. To inactivate a gene by RNAi, what information do
was correct or an error in the sequencing? you need? Do you need the map position of the target
b. You find that the exact same deletion exists in gene?
the chimpanzee homolog of the gene but that the 36. What is the purpose of generating a phenocopy?
gorilla gene reading frame is intact. Given the phy- 37. What is the difference between forward and reverse
logeny of great apes in the figure below, what can genetics?
you conclude about when in ape evolution the
mutation occurred? 38. Why might exome sequencing fail to identify a
disease-causing mutation in an affected person?
Human
39. You have identified a noncoding sequence that is con-
Chimp served in all mammals. Can you conclude that it is
Bonobo functional?
Gorilla
CHALLENGING PROBLEMS
Orangutan
40. You have the following sequence reads from a genomic
Siamang
clone of the Homo sapiens genome:
Green monkey
Read 1: ATGCGATCTGTGAGCCGAGTCTTTA
Owl monkey
Read 2: AACAAAAATGTTGTTATTTTTATTTCAGATG
b. Translate the sequence contig in all possible read- b. What additional functional genomics experiments
ing frames. would you do to identify differences between these
two species?
c. Go to the BLAST page of the National Center for
Biotechnology Information, or NCBI (http://www c. How would you show that the genetic differences
.ncbi.nlm.nih.gov/BLAST/, Appendix B) and see if you identify actually contribute to behavioral dif-
you can identify the gene of which this sequence ferences between the species?
is a part by using each of the reading frames as a 46. Different strains of E. coli are responsible for entero-
query for protein–protein comparison (BLASTp). hemorrhagic and urinary tract infections. Based on the
41. Some sizable regions of different chromosomes of the differences between the benign K-12 strain and the
human genome are more than 99 percent nucleotide enterohemorrhagic O157:H7 strain, would you pre-
identical with one another. These regions were over- dict that there are obvious genomic differences
looked in the production of the draft genome sequence a. between K-12 and uropathogenic strains?
of the human genome because of their high level of b. between O157:H7 and uropathogenic strains?
similarity. Of the techniques discussed in this chapter,
which would allow genome researchers to identify the c. What might explain the observed pair-by-pair dif-
existence of such duplicate regions? ferences in genome content?
42. Some exons in the human genome are quite small (less d. How might the function of strain-specific genes be
than 75 bp long). Identification of such “microexons” tested?
is difficult because these distances are too short to reli- K-12 CFT073
193
ably use ORF identification or codon bias to determine Nonpathogenic 2.5% Uropathogenic
if small genomic sequences are truly part of an mRNA 585 1623
7.6% 21.2%
and a polypeptide. What techniques of “gene finding”
can be used to try to assess if a given region of 75 bp 2996
constitutes an exon? 39.2%
43. A certain cDNA of size 2 kb hybridized to eight 514 204
6.7% 2.6%
genomic fragments of total size 30 kb and contained
two short ESTs. The ESTs were also found in two of
the genomic fragments each of size 2 kb. Sketch a pos- 1346
Total proteins = 7638 17.6%
sible explanation for these results. 2996 (39.2%) in all 3
911 (11.9%) in 2 out of 3
44. You are studying proteins having roles in translation O157:H7
3554 (46.5%) in 1 out of 3
in the mouse. By BLAST analysis of the predicted pro- Enterohemorrhagic
teins of the mouse genome, you identify a set of mouse
genes that encode proteins with sequences similar to GENETICS AND SOCIETY
those of known eukaryotic translation-initiation fac-
1. You decide to submit samples to two different “direct-
tors. You are interested in determining the phenotypes
to-consumer genetic testing” companies to learn more
associated with loss-of-function mutations of these
about your genetic ancestry. However, the results pro-
genes.
vided by the two companies give you very different
a. Would you use forward- or reverse-genetics estimates of your genetic origins. What might explain
approaches to identify these mutations? these differences?
b. Briefly outline two different approaches that you 2. What advice would you give to a friend who was con-
might use to look for loss-of-function phenotypes sidering doing “direct-to-consumer genetic testing” to
in one of these genes. learn more about their genealogy?
45. You are interested in identifying genetic changes that
might contribute to behavioral differences between
two species of mice: one is promiscuous and the other
is monogamous.
I
CHAPTER 15 magine for a moment that the first living cell contained within it a single DNA molecule
DNA DAMAGE, REPAIR, AND MUTATION or chromosome and that this molecule was replicated without error each time the cell
CHAPTER 16 and its descendants divided. Had life on Earth been conceived this way, our biological
THE DYNAMIC GENOME: TRANSPOSABLE world would be monochromatic. There would be a single species within which all individ-
ELEMENTS uals are exactly the same. What makes life as we know it remarkable is not the fidelity of
CHAPTER 17 DNA replication, but rather its imprecision. If mutation generates imperfections, then life
LARGE-SCALE CHROMOSOMAL CHANGES flaunts its imperfections.
CHAPTER 18 Part 1 of this text introduced the roles of constancy and variation in heredity, and
POPULATION GENETICS Part 2 elucidated how that constancy is maintained and regulated through the transfer and
CHAPTER 19 expression of genetic information. In this final section of the text, you will learn how the
THE INHERITANCE OF COMPLEX TRAITS process of mutation generates heritable variation within populations that provides the raw
CHAPTER 20 material for the evolution of the diversity of life on Earth. In the pipeline that runs from
EVOLUTION OF GENES, TRAITS, AND mutation to variation to evolution, there are checkpoints. A mutation created by an error
SPECIES in DNA replication or cellular damage can be lost or retained by chance, or affected by
natural selection if it alters the fitness of the cell or individual that carries it.
These six core principles can be seen as life’s essential ingredients:
1. Mutation generates variation in populations.
2. Variation occurs at the level of DNA and phenotype.
3. Chance influences the fate of variation within populations.
4. Fitness is measured by the number of offspring contributed to the next generation.
5. Natural selection occurs when individuals with a particular variation have higher
fitness in a specific environment.
6. Over time, natural selection results in Darwinian evolution.
Interactions between mutation, variation, chance, fitness, and natural selection result in the evolution of populations
Selection Selection
Mutation
Chance Chance
Time
As an example, a population of peppers may evolve over time from a population of round and green peppers to a population of long and red peppers due to chance and/
or selection acting on variation produced by mutation.
We are often accustomed to thinking about the deleterious effects that mutations
can have on the individuals that carry them. Indeed, a mutation can have negative
impacts on the health, fertility, or viability of an individual. However, many, and per-
haps most, mutations have no effect on the phenotype of the individual. Furthermore,
mutations can sometimes result in phenotypic changes that are beneficial. Whether a
particular mutation is detrimental, neutral, or beneficial can depend greatly upon the
environment, and this will affect whether the mutation is lost, maintained, or fixed
within a population or a species.
In Part 1 of this book, you learned of the discoveries of Mendel and others that estab-
lished the basic rules of inheritance, which are the foundation for the science of genetics.
In Part 2, you learned of the discoveries of many other geneticists and biochemists that
elucidated the details of how information encoded in DNA controls the metabolism of
cells and the growth and development of whole organisms. The final part of the book
builds on these breakthroughs and covers: (1) the generation of genetic variation through
mutation; (2) the rules governing the transmission of genetic variation from one genera-
tion to the next in whole populations; (3) the theory of how genetic variation in many
genes working together can give rise to continuous trait variation among individuals, such
as difference in height among people; and (4) the synthesis of Mendel’s theory of inheri-
tance with Darwin’s theory of evolution. When combined, these two theories provide a
powerful paradigm for how the diversity of life on Earth evolved.
500
501
I
ndividuals are phenotypically different because of varia- Overview of DNA damage, repair, and mutation
tion in genotype, the sequence of genomic DNA. Preced-
ing chapters described the inheritance of variants. This
chapter addresses the origin of variants. Two major pro- Wild-type DNA sequence
cesses are responsible for genetic variation: mutation and
recombination. Mutations are changes in the sequence of
DNA that cannot be repaired. Therefore, mutations are Accurate
DNA repair
transmitted during DNA replication to succeeding genera-
tions. Mutations are significant because they are the source
of evolutionary change; new alleles arise in all organisms,
DNA damage DNA repair
some occur spontaneously, and others are induced by
exposure to environmental factors such as radiation and • Base loss, alteration, • Direct reversal
chemicals. New alleles produced by mutations become the mismatch, or crosslink • Multi-step repair
raw material for a second level of variation, carried out by • Strand break pathways
recombination. As its name suggests, recombination is the
outcome of cellular processes that cause alleles of different Inaccurate
genes to become grouped in new combinations (Chapter 4). DNA repair
To use an analogy, mutations change the identities of indi-
vidual playing cards, and recombination shuffles the cards
and deals them out as different hands. Mutant DNA sequence
In the cellular environment, DNA is not completely sta-
• Beneficial
ble: each base pair in a DNA double helix has a certain
• Harmful
probability of mutating. The term mutation covers a broad
• Neutral
array of different kinds of changes that range from the sim-
ple swapping of one base pair for another, to the elimina-
FIGURE 15-1 DNA damage (also called a lesion) occurs through
tion of an entire chromosome (Figure 15-1). Mutations arise spontaneous and induced mechanisms that alter bases and the
from DNA damage (also called a lesion), which is a physi- phosphodiester backbone. A variety of DNA repair mechanisms can
cal or chemical abnormality in the structure of DNA. Types detect and correct DNA damage. However, DNA damage that escapes
of DNA damage include abasic sites, base mismatches, repair, or is caused during the process of repair, leads to mutations
(i.e., DNA sequence changes) that can have beneficial, harmful, or
modified bases, interstrand and intrastrand crosslinks, and
neutral effects on organisms.
strand breaks. Chapter 17 considers mutational changes
that affect entire chromosomes or large pieces of chromo-
somes, while this chapter focuses on mutations that occur
within individual genes. daughter cells when cells divide by mitosis. In contrast,
Cells have evolved sophisticated systems to detect and in multicellular eukaryotic organisms, there are two gen-
repair damaged DNA, thereby preventing the occurrence of eral types of mutational inheritance because there are
most, but not all, mutations. DNA can be viewed as being two general types of cells, somatic and germ-line. Somatic
subjected to a dynamic tug-of-war between chemical pro- mutations arise in single cells such as human skin or liver
cesses that damage DNA and lead to new mutations and cells during an organism’s life and are passed on to daugh-
cellular repair processes that constantly monitor DNA for ter cells but are not inherited by offspring. In contrast,
such damage and correct it. However, this tug-of-war is not germ-line mutations in gametes such as human sperm and
straightforward. As already mentioned, mutations provide eggs are inherited by offspring and are present in all of their
the raw material for evolution, and, thus, the introduction cells (i.e., both somatic and germ-line cells). In this chap-
of a low level of mutation must be tolerated. In fact, DNA ter, we discuss both somatic and germ-line mutations in the
replication and DNA repair systems can actually introduce context of human diseases. However, keep in mind that not
mutations. all mutations are detrimental. Some mutations are benefi-
In unicellular organisms such as E. coli (bacteria) and cial to an organism, while others are neutral and do not
S. cerevisiae (yeast), mutations are passed from parent to help or harm the organism.
502
KEY CONCEPT DNA damage and mutation do not mean code is degenerate (multiple codons can encode the same
the same thing. DNA damage is a physical or chemical defect amino acid), codons for amino acids with similar chemi-
in the structure of DNA that can often be fixed by repair cal properties and size often differ by one nucleotide, and 3
mechanisms in cells. In contrast, mutations are changes in of the 64 codons signal translation termination (i.e., stop)
the base sequence of both strands of DNA that cannot be (the genetic code is shown in Figure 9-8). New mRNA
repaired. Therefore, mutations are transmitted during DNA codons produced by base substitutions in DNA may code
replication to succeeding generations.
for the same amino acid (synonymous mutations), a differ-
ent amino acid (missense mutations), or a translation stop
KEY CONCEPT In multicellular organisms, there are two (nonsense mutations) (Figure 15-2a).
types of mutations. Germ-line mutations occur in gametes and
are passed on to offspring, whereas somatic mutations occur •• Synonymous mutations (also called silent mutations)
in all other cell types and are not passed on to offspring. change the sequence of a codon but not the encoded
amino acid. As an example, for 32 of the 61 codons
that encode an amino acid, mutating the third position
of the codon to any other base does not change the
15.1 MOLECULAR CONSEQUENCES encoded amino acid. For instance, GUA, GUC, GUG,
and GUU all encode valine (see Figure 9-8, bottom left).
OF POINT MUTATIONS
•• Missense mutations (also called nonsynonymous muta-
LO 15.1 Explain the effects of point mutations on the tions) change the sequence of a codon to one that codes
sequence and expression of RNAs and proteins. for a different amino acid. Missense mutations can result
LO 15.5 Using examples, describe how different types in an amino acid being replaced with a chemically sim-
of mutations in different genes lead to particular ilar amino acid. This is called a conservative mutation
human genetic diseases. because it may not significantly affect the protein’s
structure and function. A lysine-to-arginine change is
The term point mutation refers to a single–base-pair an example of a conservative mutation because both
change in DNA sequence. In this section, we will consider amino acids are positively charged, and they are similar
the effects on gene expression of point mutations in the in size (see Figure 9-2 to compare their chemical struc-
protein-coding regions and the noncoding regions of genes. tures). Alternatively, an amino acid may be replaced
by a chemically different amino acid. This is called a
The types of point mutations nonconservative mutation because it is likely to pro-
There are three types of point mutations in DNA: base duce a change in the protein’s structure and function. A
substitutions, base insertions, and base deletions. Base lysine-to-threonine change is an example of a noncon-
s ubstitutions are mutations in which one base pair is servative mutation because lysine is positively charged
replaced by another. They can be divided into two subtypes: and has a long hydrocarbon chain, whereas threonine is
transitions and transversions. A transition replaces a purine uncharged and has a short hydrocarbon chain.
with a purine (from A to G or G to A) or a pyrimidine •• Nonsense mutations change the sequence of a codon
with a pyrimidine (from C to T or T to C). A t ransversion that codes for an amino acid into one that stops trans-
replaces a pyrimidine with a purine (from C to A, C to G, lation (i.e., UAA, UAG, or UGA). The closer a non-
T to A, or T to G) or a purine with a pyrimidine (from A to sense mutation is to the 3′ end of the ORF, the more
C, A to T, G to C, or G to T). Thus, there are four possible likely that the resulting protein will retain its biological
transitions and eight possible transversions. In describing activity. However, nonsense mutations often produce
the changes in double-stranded DNA, an example of a tran- proteins that are completely inactive. In addition, in
sition is G ⋅ C → A ⋅ T and a transversion is G ⋅ C → T ⋅ A. eukaryotes, nonsense mutations can completely block
A base insertion is the addition of one base pair, and a base protein production by triggering decay of the mRNA by
deletion is the removal of one base pair. Collectively, base nonsense-mediated decay (NMD) (Chapter 8).
insertions and deletions are termed indel mutations (for
insertion-deletion). KEY CONCEPT A point mutation in the open reading frame
of a gene changes the sequence of a single codon and has
three potential consequences on the translated amino acid:
The molecular consequences of a point (1) no change, (2) a change to another amino acid, or (3) a
mutation in an open reading frame change to a translation stop.
Figure 15-2 shows the three types of point mutations and
their effects when they occur within the open reading Because the genetic code consists of triplet nucleotides,
frame (ORF, the protein-coding region) of a gene. The the other two types of point mutations—single base pair
variety of outcomes is a direct consequence of features of insertions and base pair deletions in the DNA sequence—
the genetic code: mRNA codons are read as triplets, the lead to a frameshift, that is, a change in the translation
(a)
Single base substitution
CAT AAG CAG AGT ACT CAT AAA CAG AGT ACT CAT AGG CAG AGT ACT CAT ACG CAG AGT ACT CAT TAG CAG AGT ACT
DNA
GTA TTC GTC TCA TGA GTA TTT GTC TCA TGA GTA TCC GTC TCA TGA GTA TGC GTC TCA TGA GTA ATC GTC TCA TGA
mRNA CAU AAG CAG AGU ACU CAU AAA CAG AGU ACU CAU AGG CAG AGU ACU CAU ACG CAG AGU ACU CAU UAG CAG AGU ACU
Protein His Lys Gln Ser Thr His Lys Gln Ser Thr His Arg Gln Ser Thr His Thr Gln Ser Thr His Stop
(b)
Single base insertion or base deletion (indel)
mRNA CAU UGC GAC AAG GAU AGU ACU CCU CAU GUG CGA CAA GGA UAG UAC UCC U CAU GCG ACA AGG AUA GUA CUC CU
Protein His Cys Asp Lys Val Ser Thr Pro His Val Arg Gln Gly Stop His Ala Thr Arg Ile Val Leu
FIGURE 15-2 (a) Mutations that change a single base pair in an open reading frame can be silent,
missense, or nonsense, with regard to encoded amino acids. (b) Mutations that insert or delete a single
base pair change the codon reading frame downstream of the mutation, which often generates a stop
codon, as shown in the insertion example.
reading frame for all codons downstream of the muta- Point mutations in ORFs can have major phenotypic
tion (Figure 15-2b). Hence, single base insertions and base consequences, as illustrated by a missense mutation in the
deletions are also called frameshift mutations. The string human ras gene. The mutation in question changes a gly-
of wrong amino acids that is encoded after the frameshift cine at position 12 of the Ras protein to a valine (G12V)
typically does not continue to the end of the original open ( Figure 15-3a). Glycine is a unique amino acid in that it
reading frame in the mRNA because there is a high prob- contains a hydrogen as its side chain, which gives it much
ability of encountering a stop codon. Thus, the effect of a more conformational flexibility relative to other amino
frameshift on a protein’s normal structure and function can acids, which all have carbon side chains. This means that
vary depending on where the insertion or deletion occurs changing a glycine to any other amino acid can have severe
in the ORF and what sequence is appended downstream effects on protein function. The Ras protein is a GTPase
of the frameshift. As described in Figure 9-6, insertions that cycles between active and inactive states when bound
or deletions of two base pairs also cause frameshifts, but to GTP and GDP, respectively (Figure 15-3b). In response
insertions or deletions of three base pairs or any multiple to signals received by cells, GTP exchange factors (GEFs)
of three do not. activate Ras by exchanging its GDP for GTP, and GTPase
activating proteins (GAPs) inactivate Ras by activating its
KEY CONCEPT A single–base-pair insertion or deletion in intrinsic GTPase activity that converts GTP to GDP. The
the open reading frame of a gene shifts the reading frame of all
G12V mutation blocks the GTPase activity of Ras, inap-
of the downstream codons, which changes the encoded amino
propriately locking it into its active form. This matters
acids and often leads to a translation stop.
because when Ras is active, it affects cells in a number of
FIGURE 15-3 (a) A G-to-T transversion in codon 12 of the ras gene changes a glycine (gly, G) to a
valine (val, V) in the Ras protein. (b) The G12V mutation blocks conversion of Ras from the active to the
inactive state. Abbreviations: GAP, GTPase activating protein; GEF, GTP exchange factor.
ways, including promoting cell proliferation. The uncon- elements for the regulation of RNA processing, stability,
trolled cell proliferation resulting from the mutant Ras pro- localization, translation, and function. In general, DNA and
tein can lead to cancer. In fact, approximately 30 percent of RNA regulatory elements are short and serve as binding
human cancers are driven by mutations in Ras. sites for proteins and RNAs that scaffold or catalyze molec-
ular processes. Thus, point mutations in regulatory ele-
ments can block binding and disrupt molecular processes.
The molecular consequences of a point As shown in Figure 15-4, point mutations in transcrip-
mutation in a noncoding region tion enhancer or promoter elements that block binding
Point mutations do not need to occur within a protein-coding of transcription factors or general transcription factors,
region to affect the phenotype of cells and organisms. Pre- respectively, will affect transcription activation of the
vious chapters have drawn attention to the importance of associated gene (Chapters 8 and 12). Point mutations in
noncoding DNA sequence elements for the regulation of eukaryotic pre-mRNA splice sites that block complemen-
DNA replication and transcription. In addition, some DNA tary base pairing of small nuclear RNAs (snRNAs) will
sequences that are copied into RNA function as sequence affect the removal of introns (Chapter 8). Point mutations
DNA
Enhancer Promoter
Transcription
5′ SS 3′ SS
mRNA m7G AAAAAA
5′ UTR Exon Intron Exon 3′ UTR
FIGURE 15-4 Point mutations in noncoding regions affect gene expression by altering the binding of
proteins and RNAs to DNA and RNA regulatory elements. Abbreviations: TF, transcription factor; GTF,
general transcription factor; snRNP, small nuclear ribonucleoprotein; snRNA, small nuclear RNA; RBP,
RNA-binding protein; RISC, RNA-induced silencing complex; miRNA, microRNA, SS, splice site; UTR,
untranslated region; m7G, 7-methylguanosine cap; and AAAAAA, poly(A) tail.
in eukaryotic mRNA untranslated regions (UTRs) that be due to genuine mutations. However, it was not known
block binding of proteins or microRNAs (miRNAs) will whether these mutants were produced spontaneously or
affect translation (Chapter 9), and point mutations that were induced by the phage.
block intramolecular base pairing will affect the function of Luria reasoned that if mutations were spontaneous,
RNAs (Chapter 8). In addition, point mutations can create they would occur at different times in different E. coli
new regulatory elements such as enhancers that affect tran- cultures. Mutations that occurred early in the growth of
scription and splice sites that affect RNA processing. a culture would give rise to a higher number of resistant
It is important to keep in mind the distinction between cells than mutations that occurred later because the mutant
the occurrence of a point mutation—that is, a change in cells had more time to produce many resistant descendants
DNA sequence—and the detection of such an event at the (Figure 15-5). Thus, if mutations occurred spontaneously,
phenotypic level. Many point mutations within noncoding the numbers of resistant colonies per culture should show
sequences elicit little or no phenotypic change; these muta- high variation (or “fluctuation,” in his words).
tions are often located between binding sites for regulatory Luria and Delbrück designed their fluctuation test as
factors. Such sites may be functionally irrelevant, or other follows. They inoculated 20 small cultures, each with a
sites may duplicate their function. few cells, and incubated them until there were 10 8 cells
per milliliter. At the same time, a larger culture was inocu-
KEY CONCEPT A point mutation in a noncoding region of lated and incubated until there were 108 cells per milliliter.
a gene can affect expression of the encoded RNA by altering The 20 individual cultures and 20 aliquots (samples) from
binding sites for proteins and RNAs that regulate transcription, the large culture were plated in the presence of phage. The
post-transcription, and translation events. 20 individual cultures showed high variation in the num-
ber of phage-resistant colonies: 11 plates had no resistant
KEY CONCEPT Point mutations in coding or noncoding colonies, and the remainder had 1, 1, 3, 5, 5, 6, 35, 64, and
regions of genes can have phenotypic consequences. 107 per plate. In contrast, the 20 aliquots from the large
culture showed much less variation from plate to plate, all
in the range of 14 to 26. If the phage were inducing muta-
tions, there was no reason why fluctuation should be higher
15.2 MOLECULAR BASIS OF for the individual cultures than the aliquots because they
SPONTANEOUS MUTATIONS were all similarly exposed to phage. This result led to the
reigning “paradigm” of mutation; that is, whether in bac-
LO 15.2 Summarize the causes of spontaneous DNA teria or eukaryotes, mutations can occur in any cell at any
damage that lead to mutations following DNA time and their occurrence is random. For this and other
replication. work, Luria and Delbrück were awarded the Nobel Prize
LO 15.5 Using examples, describe how different types in Physiology or Medicine in 1969. Interestingly, this was
of mutations in different genes lead to particular after Luria’s first graduate student, James Watson, was
human genetic diseases. awarded his Nobel Prize (with Francis Crick in 1964) for
the discovery of the DNA double-helix structure.
Mutations can arise spontaneously, or they can be induced.
Spontaneous mutations occur naturally and arise in all KEY CONCEPT Mutations can occur spontaneously, that is,
cells. Induced mutations arise through the action of exter- independently of external agents.
nal agents called mutagens that increase the rate at which
mutations occur. In this section, we consider the nature of
spontaneous mutations. Induced mutations are the subject Mechanisms of spontaneous mutations
of the next section. Spontaneous mutations arise from a variety of sources.
One source is DNA replication. Although DNA replica-
Evidence for spontaneous mutations:
tion is a remarkably accurate process, mistakes are made
the Luria and Delbrück fluctuation test in the copying of the millions to billions of base pairs in a
The causes of mutations are of considerable interest because genome. Spontaneous mutations also arise because of dam-
they are the basis of evolution and disease. In 1943, Salva- age to DNA by the cellular environment. Lastly, as we will
dor Luria and Max Delbrück developed an experimental see in Chapter 16, spontaneous mutations can be caused by
approach to test whether mutations occur spontaneously insertion of a transposable element.
as a result of cellular processes that act on DNA. It was
known at the time that if E. coli are spread on a plate of Errors in DNA replication cause spontaneous muta-
nutrient medium in the presence of phage T1, the phages tions An error in DNA replication can result when a mis-
infect and kill the bacteria. Rarely, but regularly, bacterial matched base pair forms, leading to a base substitution that
colonies were seen that were resistant to phage attack; the may be either a transition or a transversion. The genera-
resistance phenotype was heritable, and so it appeared to tion of a transition by a DNA replication error involves the
Time
Late
3 1 4 1
Late
2 4 10 0
FIGURE 15-5 Cell pedigrees illustrate the expected outcomes of induced versus spontaneous
mutation as the cause of resistant cells in the Luria and Delbrück fluctuation experiment. Dark green
cells are resistant to phage infection.
pairing of a pyrimidine with the wrong purine (e.g., C ⋅ A, Base insertions and deletions (indels) are also caused by
instead of T ⋅ A, where A is the template) or a purine with DNA replication errors. The prevailing model (Figure 15-7)
the wrong pyrimidine (e.g., G ⋅ T, instead of A ⋅ T, where T is proposes that indels arise when loops in single-stranded
the template). Other errors may insert or delete base pairs. regions of DNA are stabilized by the “slipped mispairing”
Mismatching occurs because of tautomerization and of repeated sequences in the course of DNA replication.
ionization of bases (Figure 15-6). Each of the bases in DNA This mechanism is sometimes called replication slippage or
can reside in one of several forms, called tautomers, which slipped-strand mispairing.
are isomers that differ in the positions of their atoms and in
the bonds between the atoms. The forms are in equilibrium. KEY CONCEPT Spontaneous mutations are generated by
The keto form of each base is normally present in DNA, but errors in DNA replication.
in rare instances a base may shift to the imino or enol form.
The imino and enol forms may pair with the wrong base,
forming a mispair. For example, when a C shifts to its rare DNA replication slippage and other mechanisms lead
imino form, the DNA polymerase incorporates an A rather to expansion of three base-pair repeats that are responsi-
than a G to pair with it (Figure 15-6b). Similarly, ioniza- ble for more than 40 human neurodegenerative diseases,
tion of bases brought about by proton exchange between collectively called trinucleotide-repeat diseases. Fragile X
water and hydrogen bonds can allow DNA polymerase syndrome is an example of a trinucleotide-repeat disease.
to insert a mismatch (Figure 15-6c) that resembles a wob- It is the most common form of inherited intellectual dis-
ble base pair (Figure 15-6d). Fortunately, mismatch errors ability, affecting about 1 of 4000 males and 1 of 8000
due to tautomerization and ionization of bases are usually females. Fragile X syndrome results from an increase in the
detected and removed by the 3′-to-5′ exonuclease activity number of CGG repeats in the 5′ UTR of the FMR1 (frag-
(proofreading activity) of DNA polymerase (Figure 7-18). ile X mental retardation) gene. Normal individuals have
If proofreading does not occur, the mismatches lead to tran- fewer than 45 CGG repeats, but Fragile X syndrome occurs
sition mutations, in which a purine substitutes for a purine when the number of CGG repeats expands to at least 200
or a pyrimidine for a pyrimidine. Repair systems (described (Figure 15-8). Sometimes, unaffected parents and grandpar-
later in this chapter) correct many of the mismatched bases ents contain increased numbers of repeats, but only ranging
that escape correction by DNA polymerases. from 55 to 200. For this reason, these ancestors are said to
CH3 O O– N
(b) Tautomeric base pairs
H CH3 N H N N
N O N N
H H H O H N
N O Ionized form
N N N N of guanine
Thymine H
H N H N
N N
O O
Imino form H
N N Enol form N N N
of cytosine (d) Wobble base pair
of thymine Guanine
Adenine H
H CH3 CH3 O
N O
H H H N H
N O O N
N N N N N
H N H N
N N O
H N N
O O
Cytosine H Thymine
N Thymine N N
N N N
H N
Guanine
Imino form H Enol form H
of adenine
of guanine
FIGURE 15-6 Normal base pairing compared with mismatched bases. (a) Pairing between the
normal (keto) forms of bases. (b) Rare tautomeric forms of bases (imino and enol forms) result in
mismatches. (c) Ionization of bases results in mismatches that resemble wobble base pairs (d).
carry premutations. The repeats in these premutation alleles It aggregates and causes disease of the central nervous sys-
are not sufficient to cause the disease phenotype, but they tem at some point during a normal human life span. The
are much more prone to expansion than normal alleles, and age of onset of the disease correlates with the number of
so they lead to even greater expansion in their offspring. repeat copies, that is, more repeats lead to earlier onset.
Expansion over generations to the full mutation allele with Because the Huntingtin disease allele is autosomal domi-
more than 200 repeats results in silencing of FMR1 tran- nant, it means that an individual needs to inherit only one
scription by DNA methylation of CpGs in an island (Chap- copy of the expanded gene to develop the disorder.
ter 12) at the FMR1 promoter, in surrounding regulatory
regions, and in the CGG repeats (the CpG is underlined). KEY CONCEPT Trinucleotide-repeat diseases arise through
Because no FMR1 mRNA is produced, no FMR1 protein is expansion of the number of copies of a three–base-pair
produced, which leads to impaired cognitive development. sequence normally present in several copies within coding or
Other trinucleotide-repeat diseases are associated with noncoding regions of a gene.
expansion of trinucleotide CAG codons that code for glu-
tamine (Q in the single letter amino acid nomenclature). The cellular environment causes spontaneous
Huntington’s disease (Chapter 2) is one of nine diseases mutations An important source of point mutations is
caused by polyglutamine (polyQ) expansion of a protein. In damage to DNA by water and reactive oxygen species that
normal individuals, the N-terminus of the Huntingtin pro- naturally reside in cells. Chemical reactions of DNA with
tein has about 20 glutamines encoded by 20 CAG repeats, water (i.e., hydrolysis) lead to depurination and deamina-
but when the number of glutamines is expanded to more tion, and reactions with reactive molecules that contain
than about 40, the protein becomes abnormally folded. oxygen lead to several types of DNA damage.
Insertion Deletion
Direction of DNA replication Direction of DNA replication
5′ — CGTTT T 5′ — C TGAGAGA
3′ — GCAAAAACGTAC — 3′ — GACTCTCTCTCTGCA —
T
5′ — CG T T T 5′ — CT GAGAGA
3′ — GC AAAAACGTAC — 3′ — GA CTCTC TCTGCA —
CT
Loop stabilized by Loop stabilized by
repetitive sequences repetitive sequences
T
5′ — CG T T T T TGCATG 5′ — C T GAGAGAGACGT
3′ — GC AAAAACGTAC — 3′ — GA CTCTCTCTGCA —
CT
Next round of Next round of
replication replication
FIGURE 15-7 In the course of DNA replication, base insertions and deletions (indel mutations) are A N I M ATED A RT
formed through the slipped mispairing of repeated sequences.
Molecular mechanism of mutation
Premutation
ATG Unstable,
Largely
prone to No Yes
normal
CpG CGG 55-200 expansion
Full mutation
ATG Unstable,
Affected prone to Yes No
CpG CGG >200 expansion
FIGURE 15-8 The number of trinucleotide CGG repeats in the 5′ UTR of FMR1 affects phenotype,
the probability of repeat expansion, CpG methylation, and transcription. Red and white lollipops
indicate methylated and unmethylated cytosines, respectively, CpG indicates a CpG island, the arrow
indicates the transcription start site, and ATG indicates the translation start site.
Depurination is the loss of a purine base, either guanine because apurinic (AP) sites cannot specify a complementary
or adenine (Figure 15-9a). It is brought about by hydrolysis base. However, as described later in this chapter, efficient
of the glycosidic bond between the base and deoxyribose repair systems remove apurinic sites. Additionally, under cer-
sugar. Depurination results in loss of the base from DNA, tain conditions, a base can be inserted across from an apu-
but the phosphodiester backbone remains intact. Approxi- rinic site, but this insertion frequently results in a mutation.
mately 2000–10,000 depurination events occur every day in Deamination , the hydrolytic removal of an amine
each human cell. If these lesions were to persist, they would group, alters the three DNA bases that contain an amino
result in significant genetic damage during DNA replication group (i.e., cytosine, adenine, and guanine). Deamination
(a) Depurination O P O
P
N NH N NH
O O Potentially
H
N C N mutagenic
C N NH2 O O + N NH2
O Guanine Guanine None
Blocked DNA
H2O replication and
transcription
O O
Apurinic site
(b) Deamination H
NH2 O
H O H N N
N N C•G T•A
N H N N
Transition
N O H2O NH3 N O N N
O
Cytosine Uracil Uracil Adenine
H
NH2 O N O H N
H
N N
N N A•T G•C
N N H N
Transition
N N N N
N H2O NH3 N
O
Adenine Hypoxanthine Hypoxanthine Cytosine
H
O O N O H N
N N
HN HN G•C G•C
N N H N
No change
H2 N N H2O NH3 O N HN N
N N
H O O
Guanine Xanthine Xanthine Cytosine
H
NH2 O H3C O H N N
H
H3C H3C
5
4
3N N C•G T•A
N H N N
6 2 Transition
1
H2O NH3 N N
N O N O
O
Adenine
FIGURE 15-9 Examples of the types of spontaneous mutations, (a) depurination, (b) deamination,
and (c) oxidative damage. Atoms shown in red are altered by the chemical reaction.
converts cytosine to uracil, adenine to hypoxanthine, and Base alkylation can be mutagenic by preventing base pair-
guanine to xanthine (Figure 15-9b). Uracil base pairs with ing. Certain alkylating agents, including ethylmethanesul-
adenine in replication, converting a C ⋅ G base pair into fonate (EMS) and methylnitronitrosoguanidine (MNNG),
a T ⋅ A base pair, and hypoxanthine base pairs with cyto- operate by adding alkyl groups to many positions on all
sine, converting an A ⋅ T base pair into a G ⋅ C base pair. On four bases. As an example, EMS adds an ethyl group to the
the other hand, conversion of guanine to xanthine is less oxygen at position 6 of guanine to create O-6-ethylguanine
harmful because xanthine still base pairs with cytosine but ( Figure 15-10a). This addition leads to base pairing with
with only two hydrogen bonds rather than three. Lastly, as thymine and results in G ⋅ C → A ⋅ T transitions at the next
discussed in Chapter 12, deamination of 5-methylcytosine round of DNA replication. Similarly, MNNG adds a methyl
(5mC) produces thymine, which converts a C ⋅ G base pair group to the oxygen at position 4 of thymine to pro-
into a T ⋅ A base pair. Although T ⋅ G mispairs that result duce O-4-methylthymine, which base pairs with guanine,
from deamination of 5mC can be repaired (discussed later resulting in T ⋅ A → C ⋅ G transitions. One of the primary
in this chapter), CpG dinucleotides, the major sites of cyto- experimental strategies used by geneticists to understand
sine methylation in eukaryotic genomes, remain hotspots the relationship between genotype and phenotype is to
for mutations. use exogenous agents such as EMS to induce mutations in
Oxidative damage represents a third type of spontane- genes and observe the phenotypic consequences.
ous lesion that generates mutations. Reactive oxygen spe-
cies (ROS) such as superoxide radicals (⋅O2−), hydrogen Base damage by bulky adducts Aflatoxin B1 causes
peroxide (H2O2), and hydroxyl radicals (⋅OH) are byprod- mutations by attaching to guanine at the N-7 position
ucts of normal aerobic metabolism of molecular oxygen by (Figure 15-10b). Formation of this addition product leads
mitochondria. Over 100 different types of oxidative DNA to breakage of the glycosidic bond between the base and
modifications have been identified in mammalian genomes. the sugar, thereby liberating the base and generating an
For example, thymine is converted to thymine glycol, which apurinic site. When covalently bound to DNA, aflatoxin
cannot base pair with any nucleotide and blocks DNA B1 is called a DNA adduct. Other compounds that form
replication, and guanine is converted to 8-oxoguanine DNA adducts include the diol epoxides of benzo(a)pyrene,
(8-oxoG), which mispairs with A, resulting in G ⋅ C to T ⋅ A a compound produced by internal combustion engines. All
transversions (Figure 15-9c). compounds of this class induce mutations, although the
mechanisms are not always clear.
KEY CONCEPT Spontaneous mutations are generated by
chemical reactions of DNA with water and reactive oxygen Incorporation of base analogs Some chemical com-
species in cells. pounds are sufficiently similar to the normal nitrogenous
bases of DNA that they are occasionally incorporated
into DNA in place of normal bases; such compounds are
15.3 MOLECULAR BASIS OF called base analogs. To be mutagenic, a base analog must
mispair more often than the normal base it replaces. The
INDUCED MUTATIONS base analog exists in only a single strand, but it can cause
a base-pair substitution that is replicated in all DNA copies
LO 15.3 Tabulate the causes of induced DNA damage that descended from the original strand. A base analog widely
result in mutations following DNA replication.
used in research is 2-aminopurine (2-AP). This analog of
adenine base pairs with thymine but also mispairs with
While some mutations are spontaneously produced by
cytosine when protonated, as shown in Figure 15-10c.
reactive molecules within cells, other mutations are induced
Therefore, when 2-AP is incorporated into DNA by base
by agents present in the external environment. These exog-
pairing with thymine, it can generate A ⋅ T → G ⋅ C transi-
enous agents, called mutagens, can be present in air, food,
tions by mispairing with cytosine in subsequent rounds of
or water. Mutagens can be chemical agents such as reac-
DNA replication. Alternatively, if 2-AP is incorporated by
tive oxygen species (ROS), alkylating agents, DNA adducts,
mispairing with cytosine, then G ⋅ C → A ⋅ T transitions will
base analogs, and intercalating agents, or they can be phys-
result when it base pairs with thymine. Genetic studies have
ical agents such as ultraviolet (UV) light and ionizing radi-
shown that 2-AP almost exclusively causes transitions.
ation (IR). Mutagens induce mutations by at least three
different mechanisms. They can replace a base in DNA,
Binding of intercalating agents A group of compounds
alter a base so that it mispairs with another base, or dam-
called intercalating agents are planar molecules that mimic
age a base so that it can no longer base pair with any base.
base pairs and are able to slip themselves in (intercalate)
between the stacked nitrogenous bases inside the DNA dou-
Mechanisms of induced mutagenesis ble helix (Figure 15-10d). Intercalating compounds such as
Base modification by alkylating agents Alkylation proflavin and acridine orange differ from other mutagenic
is the addition of an alkyl group (CnH2n+1, e.g., a methyl compounds in that they distort the DNA duplex “fooling”
group (CH3) or an ethyl group (C2H5)) to a nucleotide base. DNA polymerase into inserting extra bases or skipping
CH3
CH3
O N N H
C N H3C O H3C O O N
NO2
N
4 MNNG 4
T•A C•G
3 NH 3N H N N
1 1 Transition
H N N N
MNNG O O H N
H
Thymine O -4-methylthymine Guanine
H H
Guanine Protonated 2-AP Cytosine
H2N N NH2
FIGURE 15-10 Examples of induced mutations caused by analogs, 2-AP (2-aminopurine); and (d) intercalating agents, proflavin
(a) alkylating agents, EMS (ethylmethanesulfonate) and MNNG and acridine orange. Atoms shown in red are altered by the chemical
(methylnitronitrosoguanidine); (b) bulky adducts, aflatoxin B1; (c) base reaction.
templated bases, leading to base insertions and deletions, bonded adjacent pyrimidines on the same DNA strand. Two
respectively, rather than base substitutions. common types of pyrimidine dimers are cyclobutane pyrim-
idine dimers, which are characterized by a cyclobutane ring
KEY CONCEPT Chemical agents cause damage to DNA by involving carbons 5 and 6 of adjacent pyrimidines, and 6-4
adding chemical groups to bases, mimicking bases, or altering photoproducts, which contain linked carbons 6 and 4 of
the structure of DNA. In doing so, these agents increase the adjacent pyrimidines. Figure 15-11a illustrates these two types
frequency of mutations due to DNA replication errors. of UV-induced thymine dimers. During DNA replication,
pyrimidine dimers stall DNA polymerases because the bases
Base damage by ultraviolet light Ultraviolet light (UV) cannot specify a complementary partner by hydrogen bond-
of wavelengths around 300 nanometers can form various ing. Mechanisms that repair pyrimidine dimers frequently
types of pyrimidine dimers in DNA, which are covalently introduce mutations such as T ⋅ T → T ⋅ C transitions (see
(a) UV light 3′ 3′
5′ 5′
O O– O O–
O O P O O P
P Thymine O O P Thymine O O
–O –O
O O
O O
N N
CH2 N1
2 3
4 O CH2 N1
2 3
4 O
6 5 6 5
O O O O
C C C C
CH2 CH2
CH3 UV light CH3
O O– O O–
O O P O O Cyclobutane ring P
P O O P O O
–O O –O O
O O
N N
CH2 N1
2 3
4 O CH2 N1
2 3
4 O
6 5 6 5
O O
O C C O C C
CH2 CH2
CH3 CH3
Thymine O O– Thymine O O–
O O P O O P
O O O O
–O
P
–O
P Pyrimidine
O O
5′ 5′ transitions
3′ 3′ Cyclobutane pyrimidine dimer
Blocked DNA
replication and
5′ 5′ transcription
O O Thymine O O
P O H P O H
–O –O
O 2 3
N O N
CH2
N1 6 5
4 O CH2
N O
O O H
CH3 OH CH3
UV light H3C
N
O O O O
P H P
–O
O –O
O N O
2 3
CH2 N1 4 O CH2 O
6 5 N
O O
CH3 T (6-4) T
Thymine
O O O O
P P 6-4 photoproduct
–O –O
O O
3′ 3′
(b) Ionizing
radiation
Double-
strand
break
Various mutations
Gamma rays
due to strand
X-rays break repair and
oxidative damage
Single-
strand
break
FIGURE 15-11 Examples of mutations caused by (a) UV light and (b) ionizing radiation. Atoms shown
in red are altered by the chemical reaction.
the section on base excision repair). In addition, unrepaired KEY CONCEPT Physical agents such as UV light and ionizing
pyrimidine dimers can cause transcription errors. radiation cause damage to DNA by inducing intra-strand
crosslinks or strand breaks and in doing so increase the
Base damage and modification and DNA strand frequency of mutations due to DNA replication errors.
breaks by ionizing radiation Ionizing radiation (IR) from
gamma rays and X rays causes mutations by generating reac-
tive oxygen species (ROS) through radiolysis of water and
disruption of mitochondrial functions. IR produces many
Identifying mutagens in the environment:
different types of ROS, but the most damaging to DNA the Ames test
bases are superoxide radicals (⋅O2−), hydrogen peroxide In the 1970s, Bruce Ames recognized that there is a strong
(H2O2), and hydroxyl radicals (⋅OH). As illustrated in Fig- correlation between the ability of chemical compounds
ure 15-8c, ROS leads to the generation of thymine glycol and to cause cancer and their ability to cause mutations. He
8-oxoguanine, both of which can lead to mutations. IR can surmised that measurement of mutation rates in bacteria
also directly damage DNA. It breaks glycosidic bonds, lead- would be an effective way to evaluate the mutagenicity of
ing to the formation of apurinic (AP) sites or apyrimidinic compounds as a first level of detection of potential carcin-
(AP) sites (more generally called abasic sites), and it produces ogens (cancer-causing agents). The Ames test that he devel-
single-strand and double-strand breaks in DNA by severing oped uses strains of the bacterium Salmonella typhimurium
the phosphodiester backbone (Figure 15-11b). In fact, DNA that can grow only in medium that includes the amino
strand breaks are responsible for most of the lethal effects of acid histidine because the strain contains mutant alleles of
ionizing radiation because if unrepaired, strand breaks can a gene responsible for histidine synthesis (Figure 15-12a).
direct a cell to undergo cell death, and if incorrectly repaired, These mutants are known as auxotrophs because they
they can lead to chromosomal translocations (Chapter 17). require nutrients for growth that are not needed by a
(a) (b)
Culture of Ames test of aflatoxin B1 mutagenicity
his– Salmonella
typhimurium 2000
TA100 Base
substitutions
Uniformly cover plate that
lacks histidine with bacteria
1500
Revertant colonies per plate
Potential
Incubate to promote mutagenic
O O
colony growth compound 1000
O
Control Experiment
O O OCH3
Aflatoxin B1
500
Concentration
gradient Base insertions
High Low
and deletions
TA1538
TA1535
20 40 60 80 100 120 140
Spontaneous his+ Induced his+ Aflatoxin B1 dose on filter (ng)
revertant colony revertant colonies
FIGURE 15-12 (a) Summary of the procedure used for the Ames test insertions or deletions. [Republished with permission of Elsevier, Ames BN,
for mutagenic compounds. (b) Results of Ames tests of aflatoxin B1 with McCann J, Yamasaki E., “Methods for detecting carcinogens and mutagens
three strains of S. typhimurium (TA100, TA1538, and TA1535) containing with the Salmonella/mammalian-microsome mutagenicity test,” Mutation
different histidine auxotrophic mutations. The data show that aflatoxin Research, 1975, December; 31 (6):347–64, Figure 2. Permission conveyed
B1 is a potent mutation that causes base substitutions but not base through Copyright Clearance Center, Inc.]
wild-type strain. Additionally, the S. typhimurium mutants that yielded elevated levels of reversion relative to the com-
were known to “revert”—that is, grow in histidine-free pound or liver extract alone would be defined as mutagenic
medium—only by certain kinds of additional mutations. and possibly carcinogenic. Therefore, the Ames test provides
For example, an allele called TA100 could be reverted to an important way of screening thousands of compounds and
wild type only by a base substitution mutation, whereas evaluating one aspect of their risk to health and the environ-
TA1538 and TA1535 could be reverted only by base inser- ment. The Ames test is still in use today as an important tool
tion or base deletion mutations that result in a protein for evaluating the safety of chemical compounds.
frameshift. Thus, as illustrated by a study of the mutagen
aflatoxin B1, these strains can be used in the Ames test not KEY CONCEPT The Ames test is used to determine the
only to identify mutagens but also to determine the types of mutagenic activity of chemicals by testing whether they
mutations they induce (Figure 15-12b). increase the frequency of mutations in bacteria.
The Ames test works as follows. A potential muta-
genic compound such as aflatoxin B 1 is absorbed onto a
filter disc, which is placed in the center of a plate that has 15.4 DNA REPAIR MECHANISMS
S. typhimurium uniformly spread over medium that lacks
histidine (Figure 15-12a). The compound diffuses from the
LO 15.4 Illustrate the molecular mechanisms that repair
disc into the surrounding medium, creating a gradient of distinct types of DNA damage.
concentrations, with the highest concentration closest to LO 15.5 Using examples, describe how different types
the disc. The absence of histidine ensures that only rever- of mutations in different genes lead to particular
tant bacteria containing the appropriate base substitu- human genetic diseases.
tion or frameshift mutation will grow and form colonies.
After incubation, the number of colonies on each plate After surveying the numerous ways that DNA can be
and the total number of bacteria tested are determined, damaged—from sources both inside the cell (DNA rep-
allowing a determination of the frequency of reversion lication, water, and reactive oxygen species) and outside
(Figure 15-12b). (chemical and physical agents)—you might be wondering
While this works well, it became clear that not all com- how life has managed to survive and thrive for billions
pounds were directly mutagenic; rather, the actual mutagenic of years. The fact is that organisms ranging from bacte-
agent is sometimes a metabolite of the compound that is ria to plants to humans can efficiently repair their DNA.
produced in the body. Typically, these metabolites are pro- All organisms make use of a variety of repair mechanisms.
duced in the liver, and the enzymatic reactions that convert The major DNA repair pathways are base excision repair
the compound into the bioactive metabolites does not take (BER), nucleotide excision repair (NER), mismatch repair
place in bacteria. Ames realized that he could overcome this (MMR), translesion synthesis (TLS), homologous recom-
problem by pre-treating compounds with extracts of rat liv- bination (HR), and nonhomologous end joining (NHEJ)
ers containing the metabolic enzymes. Treated compounds (Figure 15-13). Each of these pathways involves numerous
Non-bulky damage
Pyrimidine Bulky Base Loop Double-strand
Alkylation Oxidation Deamination Depurination dimer adduct mismatch break
G G U T T G O O C
CH3 O
O
O O OCH3
C C G C A A C A
FIGURE 15-13 DNA repair mechanisms are paired with the types of DNA damage they act on. DNA
damage is indicated in red. DNA repair mechanisms highlighted in tan function during G1 phase of
the cell cycle, and those highlighted in blue function during S phase of the cell cycle. Homologous
recombination also functions in G2 phase of the cell cycle.
G G
O O
C C
C C
AP endonuclease AP endonuclease
C C
G
C C C
G G G
C C C
FIGURE 15-15 In base excision repair in both bacteria (a) and eukaryotes (b), damaged bases are A N I M ATED A RT
removed and repaired in a stepwise manner through the sequential action of enzymes. DNA damage is
DNA repair mechanisms: BER, NER,
indicted in red, abasic sites are indicated by dotted circles, and replacement DNA is indicated in green.
and mismatch repair
purposes. 5-Methylcytosines are hotspots for mutations light (Figure 15-11a), or damage to more than one base. If
because deamination of 5-methylcytosine generates thy- left unrepaired, these lesions can severely affect cellular phys-
mine (5-methyluracil) (Figure 15-9b), which is not recog- iology because they obstruct the progress of DNA and RNA
nized by uracil-DNA glycosylase and thus is not repaired. A polymerases, resulting in DNA replication and transcription
consequence of the frequent mutation of 5-methylcytosine blocks, respectively. Nucleotide excision repair (NER) path-
to thymine is that methylated regions of the genome are ways relieve replication and transcription blocks by repairing
converted, over evolutionary time, to AT-rich regions. the damaged DNA. In bacteria and eukaryotes, NER path-
ways are comprised of a common set of steps: damage detec-
KEY CONCEPT In base excision repair, non-bulky damage tion, strand separation, incision (i.e., cleavage), excision (i.e.,
to DNA is detected by one of several DNA glycosylases that removal), polymerization, and ligation.
cleave the base–sugar bond, releasing the damaged base. In E. coli , bulky adducts and pyrimidine dimers are
Repair consists of removal of the abasic site and insertion of either detected by UvrA acting along with UvrB or by RNA
the correct nucleotide as guided by the complementary nucle-
polymerase, which stalls during transcription at the dam-
otide in the undamaged strand.
aged site and recruits UvrA-UvrB. The pathways are respec-
tively called global genome nucleotide excision repair
(GG-NER) and transcription-coupled nucleotide excision
repair (TC-NER) (Figure 15-16a). In both cases, the DNA
Nucleotide excision repair damage site is handed from UvrA to UvrB, which is a heli-
Although the vast majority of DNA damage sustained by an case that separates the DNA strands and promotes release
organism is minor base damage that can be handled by BER, of UvrA. UvrB then serves as a scaffold for recruitment of
this mechanism cannot repair bulky adducts that distort the UvrC, which uses separate endonuclease domains to cleave
DNA helix (Figure 15-10b), pyrimidine dimers caused by UV phosphodiester bonds 8 nucleotides 5′ and 4–5 nucleotides 3′
Incision Incision
TFII
UvrD H
T T XPG
T T
XPF RPA
DNA polymerase I
A A A A
ERCC1
Synthesis Synthesis
PCNA
T T DNA polymerase I RFCT T DNA polymerase δ/ε
A A A A
Ligation Ligation
of the DNA damage site. Next, the DNA helicase UvrD encoding components of the GG-NER pathway (e.g., XPC and
and DNA polymerase I (Pol I) work together to excise the XPE) and shared components of the GG-NER and TC-NER
12- to 13-nucleotide region between the two incision sites pathways (e.g., XPB, XPD, XPF, and XPG). Lastly, patients
and fill in the gap with complementary nucleotides. Lastly, with TTD also carry mutations in genes encoding components
DNA ligase seals the nick left by Pol I. of the shared pathway (e.g., TTD-A, XPB, and XPA). Because
In eukaryotes, NER pathways follow the same steps but they have reduced capacity to repair DNA damage caused by
involve a distinct set of enzymes and excise a larger, about exposure to sunlight (i.e., UV light), XP patients are thousands
27-nucleotide, region spanning the site of DNA damage. of times more likely to develop certain types of skin cancer. In
Much of what is known about NER in humans has come contrast, CS is characterized by developmental and neurolog-
from studies of human diseases, including Cockayne syndrome ical symptoms, and TTD patients have brittle hair and ichyth-
(CS), xeroderma pigmentosum (XP), and trichothiodystrophy iosis (a dry skin disorder). Furthermore, some individuals
(TTD) that are caused by autosomal recessive mutations in present with symptoms of XP and CS or XP and TTD.
genes that encode protein components of the NER machinery. Similar to the process in bacteria, in eukaryotes, GG-NER
Patients with CS have mutations in genes encoding the CSA is initiated when XPC and XPE complexes detect damaged
and CSB proteins, which recognize stalled transcription com- DNA, whereas TC-NER is initiated when RNA polymerase II
plexes in TC-NER. Patients with XP carry mutations in genes stalls at the site of DNA damage and recruits CSA and CSB
(Figure 15-16b). After lesion detection, the GG-NER and Mismatch repair of DNA replication errors
TC-NER pathways mostly use the same proteins to remove
CH3
and repair the damaged DNA because both XPC-XPE and 3′
C GA TC
CSA-CSB recruit the multiprotein TFIIH complex, which also G CT AG
5′
functions in RNA polymerase II transcription (Chapter 8). 5′
Base
Two of the 11 TFIIH subunits, XPB and XPD, are helicases 3′
mismatch Hemi-methylation
(3′-to-5′ and 5′-to-3′, respectively) that separate the strands T 3′
GA TC
of DNA around the site of DNA damage. XPA along with CT AG
G 5′
RPA (replication protein A) further expand the DNA bub- CH3
FIGURE 15-17 A model for mismatch repair in E. coli. During DNA 3′ Ligase
3′
replication, mismatch errors are incorporated into the newly synthesized C GA TC
G CT AG
strand. To initiate repair, MutS detects a base mismatch and MutH 5′
determines which of the bases is incorrect by binding a nearby GATC CH3
sequence and distinguishing the parental from the newly synthesized
strand by the presence of a methylated adenosine. The base mismatch A N I M ATED A RT
is then repaired through a series of enzymatic steps that removes a DNA repair mechanisms: BER, NER,
region of DNA containing the incorrect nucleotide and synthesizing the and mismatch repair
correct sequence using the parental strand as a template.
MMR mechanism comes from decades of genetic and bio- they accumulate DNA damage that can lead to uncontrolled
chemical analysis in E. coli. Especially noteworthy is recon- cell growth and cancer. NHPCC is frequent because an
stitution of the MMR system in vitro in the laboratory of important target of the human MMR system is short repeat
Paul Modrich. Conservation of many of the MMR proteins sequences that can be expanded or deleted during DNA rep-
from bacteria to humans indicates that this pathway is both lication by the slipped-mispairing mechanism described pre-
ancient and important in all living organisms. We will focus viously (Figure 15-7). There are thousands of short repeats
on the very well-characterized E. coli system and briefly located throughout the human genome (Chapter 4). Although
highlight similarities and differences in eukaryotes. most are located in noncoding regions because most of the
The first step in MMR is detection of mismatches in genome is noncoding, a few are located in genes that are crit-
newly replicated DNA by the MutS protein. Binding of ical for normal growth and development.
MutS to distortions in the DNA double helix recruits MutL
and MutH. The key protein is MutH, which cuts the newly KEY CONCEPT The mismatch repair system corrects errors
synthesized strand containing the incorrect base. Without in DNA replication that are not corrected by the proofreading
the ability to discriminate between the parental and newly function of replicative DNA polymerases. Repair is restricted
synthesized strands, the MMR system could not deter- to the newly synthesized strand, which is identified in bacteria
by the lack of DNA methylation and in eukaryotes by a DNA
mine which base to excise. Strand recognition by MutH
replication factor.
is directed by adenine methylation at GATC sequences.
Because adenine methylation occurs after DNA synthesis,
newly synthesized DNA is temporarily unmodified, and KEY CONCEPT Individuals that carry mutations in genes that
this temporary absence of methylation directs repair to encode components of MMR pathway are at increased risk of
the new strand. The MutH endonuclease cuts the unmeth- cancer because dividing cells accumulate DNA damage.
ylated strand at a hemimethylated GATC sequence, and
this activity is dramatically stimulated by interactions with
MutS-MutL. MutH is targeted by hemimethylated GATC Translesion synthesis
sequences that can be either 5′ or 3′ to the mismatch and Despite the fact that mechanisms exist to repair lesions
several hundred base pairs away. MutS-MutL also activates (i.e., DNA damage) that stall DNA replication forks, some
excision, which involves the DNA helicase UvrD and four lesions persist and, by blocking DNA replication, they can
single-strand exonucleases. After incision, UvrD is loaded cause severe consequences to cells, including death. To
onto the DNA in such a way that it unwinds the DNA in avoid these consequences, both bacterial and eukaryotic
the direction of the mismatch. To finish, DNA polymerase cells use a variety of DNA polymerases to replicate past
III repairs the single-stranded gap, and DNA ligase gener- lesions and permit the completion of genome duplication.
ates a continuous covalent DNA backbone. This process, called translesion synthesis (TLS), provides
In eukaryotes, MMR is initiated when a MutS-like com- additional time for other mechanisms to repair the lesion
plex of Msh2 and Msh6 proteins binds to a mismatch. Msh2- before the replicative DNA polymerase returns to finish
Msh6 then interacts with a MutL-like complex of Mlh1 and synthesizing the genome.
Pms1 proteins, which is activated to incise the newly syn- The TLS mechanism is conserved from E. coli to
thesized strand by an interaction with the DNA replication humans (Figure 15-18). It is initiated by stalled DNA poly
β-clamp PCNA (proliferating cell nuclear antigen). After the merase, which triggers recruitment of a translesion (TLS)
replication error is removed, DNA polymerases δ and ε syn- polymerase that synthesizes past the lesion. In some cases,
thesize the correct sequence, and ligation follows. So, while the TLS polymerase extends beyond the lesion and in other
many of the steps in MMR are similar between bacteria and cases extension is carried out by another DNA polymerase.
eukaryotes, the mechanism of strand discrimination is differ- Once extension passes the lesion, the TLS polymerase is
ent. In eukaryotes, the strand-specificity of the endonuclease replaced by the replicative DNA polymerase.
is directed by PCNA, rather than by DNA methylation. More specifically, replicative DNA polymerases, E. coli
Mutations in components of the MMR pathway are Pol III and eukaryotic Pol ε, stall at sites of depurination
responsible for human diseases, especially cancers. A case in (Figure 15-9a), bulky adducts (Figure 15-10b), and pyrim-
point is Lynch Syndrome, often called hereditary nonpolypo- idine dimers (Figure 15-11a). TLS polymerases are then
sis colorectal cancer (HNPCC), which, despite its name, is not recruited by an interaction with the replicative β-clamp in
a cancer itself but increases cancer risk. The disease affects E. coli and PCNA (the β-clamp equivalent) in eukaryotes.
as many as 1 in 200 people in the Western world, making it E. coli has three TLS polymerases (Pol II, Pol IV, and Pol V)
one of the most common inherited predispositions to cancer. and humans have at least five (Pol η, Pol ι, Pol κ, Rev1, and
HNPCC results from mutations in the Msh2, Msh6, Mlh1, Pol ζ). TLS polymerases differ from replicative polymerases
and Pms1 genes. Inheritance of HNPCC is autosomal dom- in three important respects. First, whereas replicative poly
inant. Cells with one functional copy of MMR genes have merases stall because damaged bases do not fit into their
normal MMR activity, but tumor cells arise from cells that active site, TLS polymerases have much larger active sites
have lost the one functional copy and are thus deficient for that can accommodate damaged bases. Second, depending
MMR. As the mutant cells replicate their DNA and divide, on the type of lesion, TLS polymerases can be error-prone,
Translesion synthesis bypass of lesions at stalled replication forks that allows the cell to trade death for a certain level of
mutagenesis.
3′
T T
DNA lesion
KEY CONCEPT In translesion synthesis, TLS polymerases
5′ Replicative 5′ are recruited to replication forks that have stalled because of
polymerase β-clamp damage in the template strand. TLS polymerases may
3′ introduce errors in the course of synthesis that either persist
3′
as mutations or are corrected by other mechanisms such as
5′ mismatch repair.
Replicative polymerase
stalls at lesion
3′
Repair of double-strand breaks
T T
Many of the DNA damage repair systems covered thus far
5′ 5′
exploit DNA complementarity to make error-free repairs.
3′ Such error-free repair is characterized by two events: (1)
3′ removal of the damaged bases, perhaps along with nearby
5′ DNA, from one strand of the double helix; and (2) use of
TLS polymerase replaces the other strand as a template for DNA synthesis to fill the
replicative polymerase and single-strand gap. However, DNA complementarity can-
synthesizes past lesion not be exploited to repair some types of DNA damage. For
3′ example, exposure to X rays often causes both strands of
T T
G A the double helix to break at sites that are close together.
5′ 5′
TLS β-clamp This type of damage is called a double-strand break (DSB).
3′
polymerase If left unrepaired, DSBs can cause a variety of chromosomal
3′ aberrations resulting in cancer or cell death. Interestingly,
5′ the generation of DSBs is an integral feature of normal
Replicative polymerase cellular processes that require DNA rearrangements. One
replaces TLS polymerase example is meiotic recombination (Chapter 3). Cells use
and continues synthesis many of the same proteins and pathways to repair DSBs as
β-clamp
Replicative 3′ they do to carry out meiotic recombination.
T T
polymerase G A DSBs can arise spontaneously (for example, in response
5′
5′ to reactive oxygen species produced as a by-product of cel-
3′
lular metabolism) or they can be induced (for example, by
3′ ionizing radiation). There are two primary pathways used
5′ to repair DSBs in higher eukaryotes such as mammals:
nonhomologous end joining (NHEJ) and homologous
FIGURE 15-18 A generalized model for translesion synthesis in recombination (HR). NHEJ joins DNA ends independent
E. coli and humans. In the course of DNA replication, replicative of sequence complementarity, whereas HR uses comple-
polymerase stalls at lesions and is temporarily replaced by a mentary sequence on a homologous chromosome as a tem-
TLS polymerase that can continue replicating past the lesion. TLS
polymerases are error prone, as shown by the G ⋅ T mismatch. Newly
plate to extend DNA ends past a break point. Because of
synthesized DNA is indicated in green. Specific E. coli and human mechanistic differences between NHEJ and HR, NHEJ is
factors are described in the text. more prone to incorporating errors at the break-point junc-
tion than HR. Furthermore, NHEJ can function in both
dividing and non-dividing cells, whereas HR occurs primar-
in part because they lack 3′-to-5′ proofreading activity. ily in S (DNA replication) and G2 phases of the cell cycle
Third, TLS polymerases have low processivity; that is, they and is thus restricted to cells that are actively dividing. As
can add only a few nucleotides before falling off the DNA described in Chapter 10, NHEJ and HR pathways are used
template. Thus, TLS polymerases can unblock the replica- to repair DSBs that are introduced by CRISPR-Cas9 for the
tion fork but cannot synthesize long stretches of DNA. purpose of genome engineering. Repair of the DNA break
In E. coli, TLS is activated by the SOS response. Anal- by NHEJ often results in inactivation of the target gene,
ogous to its common use, Save Our Ship, the name SOS whereas repair by HR introduces a precise change in the
comes from the idea that this system is induced as an sequence of the target gene.
emergency response to prevent cell death in the presence
of significant DNA damage. The SOS pathway regulates Nonhomologous end joining One way that higher
the production of DNA repair and damage tolerance pro- eukaryotes repair DSBs is by NHEJ, which processes and
teins, including TLS polymerases. As such, SOS induction then rejoins DSB ends, frequently generating small inser-
is a mechanism of last resort, a form of damage tolerance tions and deletions at the break site (Figure 15-19). Like in
Nonhomologous end joining repair of double-strand breaks As might be expected, cells deficient for NHEJ proteins are
more sensitive to ionizing radiation.
Double-strand break It is important to note that NHEJ does not join tel-
omeres, the double-stranded ends of eukaryotic chromo-
somes. This is because multiple mechanisms cooperate to
inhibit NHEJ at telomeres, including many proteins that
specifically bind telomere sequences (Figure 7-25a).
Ku70-Ku80 binds DNA ends
Ku70-Ku80 Ku70-Ku80
KEY CONCEPT Nonhomologous end joining is an
error-prone pathway that repairs double-strand breaks in
higher eukaryotes by ligating the free DNA ends back together,
independently of extensive sequence homology.
(a)
Double-strand break
3′
5′
5′
3′
End resection
3′ 3′
5′ 3′
5′
3′
Strand invasion
DNA synthesis
D-loop
3′ 3′
5′
5′
3′
(b) Double-strand break repair (DSBR) (c) Synthesis-dependent strand annealing (SDSA)
3′ 3′
5′ 5′
HJ HJ
5′ 5′
3′ 3′
3′ 3′
5′ 5′
Noncrossover Non-
5′ 5′ crossover
3′ 3′
or
3′
5′
Crossover
5′
3′
FIGURE 15-20 A model for homologous recombination repair of DSBs. (a) Both the DSBR and A N I M ATED A RT
SDSA pathways are initiated in the same way by resection of DSB ends and strand invasion into a
homologous sequence. Blue strands indicate the broken chromosome and orange strands indicate Double-strand break repair
the homologous chromosome. (b) In the DSBR pathway, the double Holliday junction structure is either
resolved in a crossover or noncrossover manner (small arrows indicate potential sites of crossover).
The black boxes highlight the chromosomal differences that result from the two mechanisms. (c) In
the SDSA pathway, strand displacement followed by annealing leads exclusively to noncrossover
homologous recombination repair.
SUMMARY
The genome sequence in individual cells of an organism is echanisms is largely driven by unique activities of the
m
subject to change during the organism’s lifetime because of enzymes involved. A few types of DNA damage can be
errors in DNA replication as well as damage to DNA caused directly repaired by s ingle enzymes, but most types of DNA
by intrinsic (i.e., cellular) factors and extrinsic (i.e., envi- damage require multiple enzymes and DNA-binding factors
ronmental) factors. Mutations that result from DNA repli- for repair. In general, DNA repair is initiated by proteins that
cation errors and DNA damage can have a wide v ariety of detect DNA damage. In some cases, detection occurs when
phenotypic effects, depending on the type of mutation and the DNA replication or transcription machinery stalls at the
the location of the mutation in the genome. As an exam- site of DNA damage. Following detection, damaged nucle-
ple, DNA replication errors that cause trinucleotide-repeat otides and often surrounding nucleotides are removed by
expansion in a noncoding region of the FMR1 gene lead to helicases and nucleases and the correct sequence is replaced
Fragile X syndrome, which is characterized by intellectual by DNA polymerases, which fills in the gap using the other
disability, while a trinucleotide-repeat expansion in a coding strand as a template, and DNA ligase, which seals the final
region of the Huntingtin gene leads to Huntington’s disease, nick in the phosphodiester backbone. The use of a template
which is characterized by neurodegeneration. In the case of for repair ensures that the repair is error-free. As a case in
point mutations that change a single base pair in a gene, out- point, double-strand break repair by homologous recombi-
comes can range from no effect on gene expression or func- nation, which uses a sister chromatid or the other chromo-
tion of the encoded protein to complete loss of expression some copy as a template, is error-free, but nonhomologous
or protein function. Accordingly, point mutations can cause end joining, which does not use a template, is error-prone.
beneficial, detrimental, or neutral changes in phenotype. For The importance of DNA repair is demonstrated by diseases
example, a point mutation that changes a single amino acid such as xeroderma pigmentosum, Cockayne syndrome, tri-
in the Ras protein causes cancer (Figure 15-3). chothiodystrophy, and hereditary nonpolyposis colorectal
Both bacterial and eukaryotic cells reduce the potential cancer that result from loss of DNA repair factors. Much
mutagenic effects of DNA damage by detecting and repair- remains to be learned by geneticists about how cellular and
ing the damage. DNA repair mechanisms act on specific environmental factors produce DNA damage, how DNA
types of damage and are functional during specific phases damage is repaired to prevent mutations, and how muta-
of the cell cycle (Figure 15-13). The specificity of repair tions lead to disease.
KEY TERMS
abasic site (p. 514) global genomic nucleotide excision oxidative damage (p. 511)
alkylation (p. 511) repair (GG-NER) (p. 517) point mutation (p. 503)
Ames test (p. 514) Holliday junction (HJ) (p. 522) pyrimidine dimer (p. 513)
apurinic (AP) site (p. 514) homologous recombination (HR) recombination (p. 502)
apyrimidinic (AP) site (p. 514) (p. 521) replication slippage (p. 507)
auxotroph (p. 514) indel mutation (p. 503) somatic mutation (p. 502)
base analog (p. 511) induced mutation (p. 506) SOS response (p. 521)
base excision repair (BER) (p. 516) intercalating agent (p. 511) spontaneous mutation (p. 506)
base deletion (p. 503) ionization (p. 507) synonymous (silent) mutation
base insertion (p. 503) ionizing radiation (IR) (p. 514) (p. 503)
base substitution (p. 503) mismatch repair (MMR) (p. 519) synthesis-dependent strand annealing
cancer (p. 505) missense (nonsynonymous) mutation (SDSA) (p. 522)
conservative mutation (p. 503) (p. 503) tautomerization (p. 507)
deamination (p. 509) mutagen (p. 506) transcription-coupled nucleotide
depurination (p. 509) mutation (p. 502) excision repair (TC-NER) (p. 517)
DNA damage (p. 502) nonconservative mutation (p. 503) transition (p. 503)
double-strand break (DSB) (p. 521) nonhomologous end joining (NHEJ) translesion (TLS) polymerase
double-strand break repair (DSBR) (p. 521) (p. 520)
(p. 522) nonsense mutation (p. 503) translesion synthesis (TLS) (p. 520)
frameshift mutation (p. 504) nucleotide excision repair (NER) transversion (p. 503)
germ-line mutation (p. 502) (p. 517) trinucleotide-repeat disease (p. 507)
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the www
WORKING WITH THE FIGURES 16. In Figure 15-16, why might the repair of pyrimidine
( The first 20 questions require inspection of text figures.) dimers be faster and more efficient on the coding
strand of transcriptionally active genes than on other
1. In Figure 15-1, describe a mutation that is neutral. parts of the genome?
2. In Figure 15-2, using an example, explain why it is 17. In Figure 15-17, draw the steps of mismatch repair
essential to know the reading frame to determine with a hemimethylated GATC sequence located 5′ of
how a mutation in an open reading frame affects the the base mismatch. What type of exonuclease, 5′-to-3′
encoded protein. or 3′-to-5′ is required in this case?
3. In Figure 15-3, are there any other single base changes 18. In Figure 15-18, why does repair by translesion synthe-
that would convert a GGC glycine codon into a valine sis not take place in non-dividing cells such as neurons?
codon? Consult Figure 9-8.
19. In Figure 15-19, draw three different DNA end struc-
4. In Figure 15-4, what are the two main types of molec- tures that might be formed by a double-strand break
ular interactions that are affected by point mutations and would be substrates for repair by nonhomologous
in noncoding regions? end joining. Based on these structures, explain why the
5. In Figure 15-5, if the mutation occurred spontaneously proteins involved in nonhomologous end joining are
in the original cell, what percentage of the cells in the described as being distinctive in their versatility.
fifth generation would contain the mutation? 20. In Figure 15-20, how are the noncrossover products of
6. In Figure 15-6, when a G shifts to its rare enol form, DSBR and SDSA different?
what nucleotide can the DNA polymerase add rather
than a C? Also, following additional rounds of DNA BASIC PROBLEMS
replication, wild-type cells have a G ⋅ C base pair; what
21. What is the difference between a DNA lesion and a
base pair will mutant cells have?
DNA mutation?
7. In Figure 15-7, write a DNA template sequence that
22. Consider the following wild-type and mutant
might lead to a two-base insertion due to slippage dur-
sequences:
ing DNA replication.
Wild-type . . . CTTGCAAGCGAATC . . .
8. In Figure 15-8, how might DNA methylation inhibit
transcription of FMR1 (see Section 12.3)? Mutant . . . CTTGCTAGCGAATC . . .
9. In Figure 15-9, oxidation of guanine to 8-oxoguanine The substitution shown seems to have created a stop
leads to a G ⋅ C → T ⋅ A transversion after DNA repli- codon. What additional information do you need to be
cation. Write out the DNA replication steps that lead confident that it has done so?
to the base transversion. 23. Can a missense mutation of proline to histidine be
10. In Figure 15-10, the reaction of EMS with guanine made with a G ⋅ C → A ⋅ T transition-causing mutagen?
generates O-6-ethylguanine, which leads to a G ⋅ C → What about a proline-to-serine missense mutation?
A ⋅ T transition. Write out the DNA replication steps Refer to the genetic code in Figure 9-8.
that lead to the base transition. 24. By base-pair substitution, what are the synonymous
11. In Figure 15-11, draw a 6-4 photoproduct between and nonsynonymous changes that can occur starting
thymine and cytosine. with the codon CGG? Refer to Figure 9-8.
12. In Figure 15-12, what types of DNA damage are likely 25. A mutational lesion results in a sequence containing a
to have led to mutations in spontaneous his+ revertant mismatched base pair:
colonies? 5′ AGCT G CCTT 3′
13. In Figure 15-13, which of the types of DNA damage 3′ TCG ATG GAA 5′
are caused by DNA replication errors? codon
14. In Figure 15-14, do cells need to undergo DNA repli- If mismatch repair occurs in either direction, which
cation for these repair mechanisms to work? Explain amino acids could be found at this site? Refer to
your answer. Figure 9-8.
15. In Figure 15-15, draw the chemical reaction between 26. Why does acridine orange commonly produce null
nucleotides that is catalyzed by DNA ligase. alleles?
27. Defend the statement “Cancer is a genetic disease.” 40. Which of the following statements best describe the
28. Where within a gene might trinucleotide repeat expan- mismatch repair pathway?
sion occur, and how might expansion at those sites a. It is part of the 3′-to-5′ proofreading function of
lead to disease? DNA polymerases.
29. Do you think that mutations in the base excision b. It acts after DNA replication by recognizing mis-
repair factor DNA polymerase β increase the risk of matched base pairs.
cancer? Why or why not? c. It is activated by stalled replication forks.
30. In mismatch repair, only mismatches in the newly syn- d. It is coupled to transcription.
thesized strand are corrected. How are bacteria and
eukaryotes able to detect the newly synthesized strand? CHALLENGING PROBLEMS
31. Why are many chemicals that test positive by the Ames
41. Hydroxylamine (HA) only causes G ⋅ C → A ⋅ T transi-
test also classified as carcinogens?
tions in DNA. Will HA produce nonsense mutations?
32. Differentiate between the elements of the following Will HA revert nonsense mutations?
pairs:
42. You are using methylnitronitrosoguanidine (MNNG)
a. Transitions and transversions to “revert” mutant nic-2 (nicotinamide-requiring)
b. Synonymous and neutral mutations alleles in Neurospora. You treat cells, plate them
on a medium without nicotinamide, and look for
c. Missense and nonsense mutations
prototrophic colonies (i.e., colonies that grow on
d. Frameshift and nonsense mutations minimal media). You obtain the following results for
33. Describe two spontaneous lesions that can lead to two mutant alleles. Explain these results at the molec-
mutations. ular level, and indicate how you would test your
hypotheses. www
34. What are translesion polymerases? How do they differ
from the replicative polymerases? How do their special a. With nic-2www
allele 1, you obtain no prototrophs at
features facilitate their role in DNA repair? all.
35. A certain compound that is an analog of the base cyto- b. With nic-2 allele 2, you obtain three prototrophic
sine can become incorporated into DNA. It normally colonies A, B, and C, and you cross each separately
hydrogen bonds just as cytosine does, but it quite often with a wild-type strain. From the cross prototroph
isomerizes to a form that hydrogen bonds as thymine A × wild type, you obtain 100 progeny, all of which
does. Do you expect this compound to be mutagenic, are prototrophic. From the cross prototroph B ×
and, if so, what types of changes might it induce in wild type, you obtain 100 progeny, of which 78
DNA? are prototrophic and 22 are nicotinamide requir-
ing. From the cross prototroph C × wild type, you
36. In cells that have stopped dividing, what types of DNA
obtain 1000 progeny, of which 996 are prototro-
repair systems are possible?
phic and 4 are nicotinamide requiring.
37. Two pathways, homologous recombination (HR) and
nonhomologous end joining (NHEJ), can repair dou- GENETICS AND SOCIETY
ble-strand breaks in DNA. If HR is an error-free path-
Despite considerable data showing that sunscreen provides
way whereas NHEJ is not always error free, why is
protection from DNA damage following exposure to UV
NHEJ used most of the time in eukaryotes?
irradiation, many individuals are unwilling to protect their
38. Which repair pathways detect DNA damage during skin against the sun. Based on what you now know about
transcription? the causes and consequences of DNA damage and the mech-
39. Which of the following is not possible? anisms of DNA repair, what arguments would you make in
support of public health programs that provide guidance
a. A nonsynonymous mutation in an intron
for sun protection, particularly in younger individuals?
b. A nonsynonymous mutation in an exon
c. An indel mutation in an intron
d. An indel mutation in an exon
527
I
n the 1940s, one of the most remarkable discoveries in the Barbara McClintock and colleagues
history of genetics was made. Namely, there are genetic loci at Cornell University
that can move from one location in the genome to another.
This special class of loci became known as transposable
elements, or transposons for short. A particular transposon
might be on one chromosome in an individual but appear
on a different chromosome in its offspring. In the 1940s, the
idea that there are mobile genetic loci was heretical, and so
it was viewed skeptically and never broadly accepted until
demonstrated at the level of DNA decades later.
Perhaps more remarkable than the discovery of trans-
posons themselves was the woman who discovered them—
Barbara McClintock. McClintock was recognized by her
colleagues as an exceptional, indeed profoundly gifted, sci-
entist from the time she was a graduate student at Cornell
University. She earned the distinct honor of being elected
to the U.S. National Academy of Sciences at the relatively
youthful age of 41. The same year she was elected the first
woman president of the Genetics Society of America. Among
the numerous honors she won in science, she was awarded
FIGURE 16-1 Rollins A. Emerson laboratory members at Cornell
the Nobel Prize in Physiology or Medicine in 1983, the first University, 1929. Standing from left to right: Charles Burnham, Marcus
woman to win the Nobel Prize unshared. In 2005, the United Rhoades, Rollins Emerson, and Barbara McClintock. Kneeling is
States Postal Service issued a stamp in her honor. George Beadle. Both McClintock and Beadle were awarded Nobel
McClintock began her career in maize genetics in the lab- Prizes. [Department of Plant Breeding, Cornell University.]
oratory of Rollins Emerson at Cornell University in the 1920s
(Figure 16-1). Remarkably, at that time, George Beadle, who addition to her Nobel Prize-winning work on transposons,
also won a Nobel Prize for genetics research (the one-gene, she made important contributions to the understanding
one-enzyme hypothesis; Chapters 6 and 9) was another mem- of cytological crossing-over and genetic recombination
ber of the Emerson lab group. McClintock and Beadle are the (Chapter 4) and telomeres (Chapter 2 and 7). She defined
only two contemporary graduate students in a single genet- the morphology of maize chromosomes and published the
ics lab who both went on to win separate Nobel Prizes. The first genetic map for maize. She discovered a cycle in some
Emerson group also included Marcus Rhoades and Charles dividing cells by which a broken chromosome is repaired
Burnham, two leading maize geneticists of this era. Never- and then breaks again with each successive cell division,
theless, even in this august company, McClintock stood out. generating a series of large-scale somatic mutations. This
George Beadle recognized that McClintock’s skills in cytoge- cycle remains important in cancer research today.
netics exceeded his own. The story goes that one day, Beadle In this chapter, you will learn about transposable ele-
prepared some chromosome spreads on microscope slides ments, segments of DNA that can move from one location
that he knew would take him some time to understand. On in the genome to another. We will begin by going through a
leaving the lab that day, he instructed the others, “Don’t show few of McClintock’s experiments and the logic she used to
these slides to Barbara,” as he knew she would solve the infer that some loci are mobile. We will see that transposons
puzzle before he could. are found in virtually all organisms across the tree of life and
McClintock had an exceptional talent for discerning that they can comprise a substantial portion of an organism’s
underlying genetic mechanisms by combining cytological genome. There are two classes of transposons with differ-
observations, the progeny ratios from genetic crosses, and ent biological properties, and the host species regulates their
her knowledge of maize. A self-described introvert with “a movement. Finally, transposons are important sources of
capacity to be alone,” she had a deeply creative mind that new mutations and can be used as tools for reverse genetics
saw well beyond the limits of knowledge of her time. In (Chapter 14) and producing transgenic organisms.
528
16.1 DISCOVERY OF or ripened fruit that has an embryo and a nutritive tissue
surrounding the embryo called the endosperm. Each kernel
TRANSPOSABLE develops after a pollen grain lands on the single silk (style)
ELEMENTS IN MAIZE attached to the immature kernel, germinates, and then the
pollen tube grows through the silk into the immature ker-
LO 16.1 Design and interpret experiments related to the nel (ovary) to complete fertilization. Each kernel contains a
excision, insertion, recombination, repression, and single offspring (embryo) of the parent plant on which the
transcription of transposable elements.
ear is borne.
Like most other flowering plants, the maize life cycle
To understand McClintock’s experiments, let’s begin with includes a process called double fertilization. Each pollen
a little biology of maize. Hundreds of kernels are borne grain possesses two haploid sperm cells—one sperm cell
on each maize ear (Figure 16-2a). Each kernel is a mature combines with the haploid egg cell in the ovary to form
(b)
Pachytene stage of meiosis
Knob
Colored
Shrunken
Knob
Bronze
Waxy
the diploid embryo, and a second sperm cell combines with C allele to produce the blue pigment (Figure 16-3a). Con-
a diploid cell (the central cell) in the ovary to form a trip- sistent with this interpretation, this special stock of maize
loid (3n) cell that divides mitotically to form the triploid showed a high frequency of breakage of chromosome 9
endosperm that surrounds and nourishes the embryo (see when examined cytologically.
Figure 16-2a). The two sperm cells are genetically identi- McClintock made crosses between this special stock
cal, as are the sets of chromosomes in the egg cell and dip- as the male parent and a stock carrying additional marker
loid central cell. The endosperm makes up the bulk of the loci on chromosome 9—the recessive waxy, the recessive
starchy tissue that comprises the maize kernel. Importantly, bronze, and recessive shrunken—as well as C for colored
since the endosperm carries sets of chromosomes from both (blue) kernels (Figure 16-3b). McClintock expected the ker-
the pollen (male) and ear (female) parents, it expresses the nels to lack blue pigment; have normal, not waxy starch;
genotype of the next generation. and be plump rather than shrunken since the dominant
A strength of maize for McClintock’s work is that it alleles from the special stock should obscure the reces-
has relatively large chromosomes that vary in size and bear sive alleles from the female parent. However, she received
distinctive features (knobs) that allowed her to identify another surprise: on some kernels, there were multiple
individual chromosomes on her microscope slides (Figure bronze-colored sectors that had waxy starch and shrunken
16-2b). The knobs are dense segments of heterochromatin rather than plump tissue. In these sectors, all the recessive
that produce bulges at specific and diagnostic locations on alleles were uncovered. This observation could be explained
the different chromosomes. Maize has a haploid number if the chromosome was always breaking in the same loca-
of 10 chromosomes that are numbered 1 to 10 in order of tion somewhere between Wx and the centromere, simulta-
largest to smallest. Each chromosome has a cytologically neously uncovering all four recessive alleles—wx, bz, sh,
defined short (S) and long (L) arm. Finally, McClintock and C. She called this locus for chromosome breaking Ds
took advantage of several linked genes on the short arm (for Dissociation).
of chromosome 9 that are all expressed in the developing McClintock confirmed the presence and location of Ds
kernel and have visible mutant phenotypes in the kernel. on chromosome 9 through her cytological observations.
Waxy (Wx) has a recessive mutant allele (wx) that converts When she examined the meiotic prophase chromosomes
normal starch to waxy starch; Bronze (Bz) has a recessive of the special stock, she observed recurrent chromosome
mutant (bz) that converts a blue pigment to a bronze color; breaks at the same location on the short arm of chromo-
Shrunken (Sh) has a recessive mutant (sh) that gives shriv- some 9, using a chromosome knob as a cytological marker
eled rather than plump kernels. The Colorless (C) locus has for chromosome 9. She could detect Ds both genetically
three alleles: C conditions a blue pigment in kernels; c does by the sectored kernels and cytologically by chromosome
not make blue pigment; and CI is a dominant inhibitor breaks.
allele that represses pigment production; it is dominant
to C. The order of dominance is C I > C > c. KEY CONCEPT Kernel sectors uncovering the recessive
alleles at multiple linked loci on maize chromosome 9 indicated
that the chromosome broke (dissociated) repeatedly in the
McClintock’s experiments: the Ds element same position during kernel development. This inference was
confirmed by cytological observation of recurrent breaks on
In the 1930s, McClintock was working with X-ray-induced
chromosome 9 of the special maize stock being used.
mutants of maize that cause frequent chromosomal breaks.
At this time, she encountered a plant whose progeny exhib-
ited a wide variety of cytological abnormalities on chromo- Another observation McClintock made was that, in
some 9 including chromosome losses, translocations, and some backcross families with the special stock, about
inversions that she could observe cytologically. Among the 50 percent of the kernels were colorless with no sectors,
descendants of this special plant, there was one plant car- and about 50 percent had blue sectors on a colorless back-
rying CI/CI that was expected to produce all heterozygous ground. The 1:1 ratio of these types of kernels suggested
CI/C colorless kernels (lack blue pigment) when used as the another Mendelian factor was required for Ds to break
pollen parent with a C/C ear parent. However, when she the chromosome. If the pollen parent was heterozygous for
made this cross, she received a surprise: some of the kernels this additional factor, then half of the kernels on the ear
had multiple blue sectors (Figure 16-3a). would inherit it, and it would “activate” the breakage of
McClintock suspected that the CI allele was being lost chromosome 9 at the Ds locus, uncovering the C allele and
in the sectors with blue cells during kernel development so giving a blue sector. The other half of the kernels would
that it would no longer inhibit the C allele and its ability not inherit this factor, so Ds would not be activated, the
to make the blue pigment. An obvious way for this loss to chromosome would not break, and CI would be present in
happen would be by breakage of chromosome 9, which she all cells, inhibiting the formation of blue-colored sectors.
had been studying. If chromosome 9 broke between the C McClintock called this second factor Ac (for Activator).
locus and the centromere on the chromosome carrying C I, She knew Ac was unlinked from Ds the same way you
then the CI allele would be on an acentric fragment and learned how to determine that two loci are unlinked in
be lost during mitosis in the descendant cells, allowing the Chapter 4.
Genotypes Phenotypes
(b) Sh
CI Ds
Wx
♂ With Ac, Ds breaks the
chromosome causing
C sh bz wx Ac
bronze, waxy, and shrunken
♀♀ sectors (waxy and shrunken
traits are not visible)
(c) Jump
C I Ds Sh Bz Wx Ds
♂ With Ac, Ds excises and
C sh bz wx Ac jumps to a new location
♀♀
CI
Sh Bz Wx
♂ With Ac, Ds breaks the
C sh bz wx Ac chromosome causing blue
♀♀ sectors with normal starch
FIGURE 16-3 Chromosome 9 of maize breaks at the Ds locus, where produced. (b) Like panel (a) except recessive alleles at shrunken, bronze,
the Ds transposable element has inserted. (a) Chromosome composition and waxy are incorporated on the chromosomes. When C, sh, bz, and
of the endosperm (3n) of a kernel with the Ds near the centromere wx are all uncovered by a chromosome break, bronze sectors with waxy
on the copy of chromosome 9 contributed by the pollen parent. The and shrunken tissue are produced. (c) Like panel (b) except Ds has
endosperm is heterozygous C/C/CI, with two copies of C from the moved to a new location between C and Sh (top). When Ds breaks the
female parent. When Activator (Ac) is not present (top), the Ds element chromosome between Sh and C, only C is uncovered, so the sectors
is stable. When Ac is present on a separate chromosome (bottom), the are blue but have normal, not waxy starch, and they are plump, not
Ds element breaks the chromosome, producing an acentric fragment shriveled (bottom).
with the CI allele and sectors of cells lacking CI in which blue pigment is
Having hypothesized another factor, Ac, McClintock in the position of the breaks along the chromosome. She
wanted to determine where it was located in the genome, could observe this change genetically, because when the
that is, she wanted to map it. She did this by crossing plants chromosome broke, only CI was on the acentric fragment
of stocks with Ac and Ds to other stocks with marker and lost. The dominant Wx, Sh, and Bz alleles were on the
genes on different chromosomes. If Ac was near one of the centromere side of the breakpoint. Thus, the sectors in the
marker genes on another chromosome, she would observe kernels with the broken chromosome were blue since CI
linkage of Ac and the marker locus. When doing these was lost, but they were plump, not shrunken; had normal,
mapping experiments, McClintock received still another not waxy starch; and were blue, not bronze color because
surprise: Ac mapped to different chromosomes in different the dominant Wx, Sh, and Bz alleles were all retained. She
crosses. These observations raised the question: Is Ac able concluded that Ds had moved from a location between Wx
to move around the genome? and the centromere to a location between Sh and C.
Griffiths12e_16_04
McClintock found another piece of the puzzle when Here is one last piece of evidence that convinced
she discovered another derivative of her special stock in McClintock that loci can move around the genome.
which Ds seemed to have moved. With this derivative, McClintock made the cross shown in Figure 16-4. For this
chromosome 9 no longer broke between Wx and the cen- cross, when Ac is not present, we expect the chromosome
tromere; rather, it broke between Sh and C (Figure 16-3c). to remain intact and the kernels to be entirely blue. When
She could observe this change cytologically by the change Ac is present, we expect the chromosome to break at Ds
♀ ♂
c C Ds Ac
c
× C Ds
C Ds
♂ 50% of kernels
Without Ac, Ds is stable and
c the C allele gives blue kernels
♀♀
C
Ac
50% of kernels
♂ With Ac, Ds breaks the
c chromosome causing off white
♀♀ sectors by uncovering c
Ds
Ds jumps into C
creating the c-m1 allele
C Ac
♂ Rare kernel, Ds jumped into
C creating the new c-m1 allele;
c
Ds excision gives blue sectors
♀♀
FIGURE 16-4 A female parent homozygous for the recessive c allele and breaks the chromosome in the sectors, so that the C allele is lost on
is pollinated with a male parent that is homozygous for the dominant the acentric fragment and only the recessive c allele that does not make
C allele that makes blue pigment and for Ds, but that is heterozygous blue pigment is present in the sectors. In one rare kernel, Ds jumps into
for Ac on another chromosome. Fifty percent of the offspring kernels the C locus, creating a new unstable mutant allele (c-m1), giving a
are solid blue because Ac is absent, so chromosome 9 remains intact c/c-m1 heterozygote with a colorless kernel. When Ds jumps out of
and all cells in the kernel are heterozygous—C/c. Fifty percent of the c-m1 in the presence of Ac, blue sectors are formed as c-m1 reverts
offspring kernels are blue with colorless sectors because Ac is present to C.
and the dominant C allele for blue color to be lost, so Second, the transposase that Ac encodes not only can
one should see colorless sectors on an otherwise blue ker- activate the transposition of Ds, but it can activate its own
nel. The result should be a 1 : 1 ratio of blue and sectored transposition. Ac can move from one location to another
kernels. just like Ds. As noted previously, McClintock discovered
When McClintock made the cross in Figure 16-4, she that Ac mapped to different chromosomes in different
did see a 1 : 1 ratio of blue to sectored kernels. However, stocks of maize, suggesting it could move. As further proof
among about 4000 kernels examined, she observed one ker- that Ac transposes, McClintock found an additional unsta-
nel that had the reversed type of sectoring—blue sectors on ble allele of the C gene into which Ac was inserted. This
a colorless background. Subsequent experiments revealed allele is called c-m(Ac). Since Ac activates the movement
that in this kernel, Ds had moved right into the C locus, of transposons, the Ac in c-m(Ac) could activate its own
creating a new recessive allele called c-m1 by McClintock. transposition out of c-m(Ac) to restore a functional allele
The presence of Ds in the C locus disrupted the ability of of C that conferred blue sectors in the kernels. Because
the c-m1 allele to make the blue pigment. However, in the Ac can move on its own, it is said to be an autonomous
presence of Ac, Ds jumps out of c-m1 to restore a functional t ransposable element.
dominant C allele in sectors of the kernel, giving blue spots Third, Ds elements are not all the same but come in
on a white background. The “m” in c-m1 stands for mutable, different forms with different properties. The Ds element
and it is a mutable or unstable allele. The leaving Ds ele- that McClintock first discovered near Wx had a structure
ment is said to excise from the chromosome or transpose. that causes it to break the chromosome when it excises
This was further evidence that Ds was a mobile locus. such that the two ends of the broken chromosome are not
joined back together by the transposase. Other Ds elements
KEY CONCEPT Ds is a genetic element that can jump have a different structure such that when they excise, the
around the genome or break chromosomes in the presence of transposase joins the broken ends of the chromosome back
Ac. Ac also has the ability to jump from one genomic location together. This type of Ds is found in the c-m1 allele. When
to another itself, and Ac provides functionality that enables Ds the Ds in c-m1 excises, the two ends of the chromosome are
to move.
joined together, restoring a functional dominant C allele.
Transposable element in Drosophila nonsense or missense nucleotide substitutions and the rever-
tants were being restored to the original nucleotide, then a
chemical mutagen that increased the single nucleotide muta-
tion rate should increase the rate of reversion. It did not, and
so something else must have been happening.
The scientists took advantage of the ability of λ phage
to pick up the gal operon from E. coli, inserting it into the
phage DNA and thereby into the resultant phage parti-
cles. Such λ phage with gal could be made from wild-type
(λdgal + ) or mutant (λdgal −) E. coli. The “d” is for “defec-
tive” because when gal inserts into the phage, a part of
the phage’s own chromosome is deleted. When a mixture
of λdgal + and λdgal − phage was subjected to density cen-
trifugation (see Chapter 7), the two types of phage parti-
cles could be separated because they had different buoyant
densities (Figure 16-6a). The λdgal − had a higher density
or molecular weight (more DNA) than the λdgal + . Why
should the mutant have more DNA?
The next experiment took advantage of the ability to
hybridize the DNA from λdgal + and λdgal − to one another.
The DNAs from these two phages are mixed, then dena-
tured, and then allowed to anneal to one another. When the
FIGURE 16-5 Excision of the mariner transposable element from the annealed heteroduplexes (double-stranded DNA molecules
white gene causes mosaicism in the eye of a Drosophila. In the red composed of one λdgal + and one λdgal − strand) are exam-
sectors, mariner has been excised from the white gene, restoring a
ined under an electron microscope, one observes a loop
functional allele and wild-type red eye color. [Courtesy of Emilie Robillard,
CNRS.] of single-stranded DNA where the λdgal + has no comple-
ment to the λdgal − strand (Figure 16-6b). The size of the
loop was estimated to be about 1000 bp. When this exper-
16.2 TRANSPOSABLE ELEMENTS iment was done with gal mutants that are caused by point
mutations, there was no loop in the heteroduplex. This
IN BACTERIA experiment showed that the gal− mutants contained extra,
inserted DNA relative to wild type.
LO 16.2 Determine the class of a transposon from its DNA Finally, other experiments showed the extra DNA (the
structure and the proteins it encodes and predict
its behavior based on its class. insertion) in the gal− strain hybridizes to different regions of
the E. coli genome in different E. coli strains, that is, it was
The molecular nature of transposable elements was first moving around like Ac and Ds. The small-phage DNAs could
understood in a bacterium, E. coli. The discovery of trans- be readily sequenced and the sequences of the λdgal + and
posons in E. coli parallels McClintock’s work in several λdgal − compared. This work provided the first DNA sequence
ways—there were unstable mutants that could revert to wild of a transposable element. The sequences revealed that the
type, and a “locus” could appear at several locations around inserted element contains a single open reading frame that
the genome. However, the array of molecular tools that could encodes a transposase flanked by short inverted repeat (IR)
be applied in bacteria, along with the small genome size of sequences of about 20 bp in length (Figure 16-7a). The first
bacteria, enabled resolving these elements to the DNA level. such insertion element isolated from the gal operon was called
IS1 for insertion-sequence (IS) element 1. Subsequently, differ-
ent E. coli strains were found to harbor other similar elements
Evidence for transposable elements named IS2, IS3, etc. The genome of the standard wild-type
in bacteria E. coli is rich in IS elements: it contains eight copies of IS1, five
The story begins with the isolation of some new mutants in copies of IS2, and copies of other less well-studied IS types.
the E. coli gal operon—a cluster of three genes that encode To summarize, some mutants of the gal operon of
the enzymes required for use of galactose as an energy source E. coli contain transposable elements:
(see Chapter 11). These mutants cannot grow on a medium
• The mutants arose spontaneously, and they can revert
in which galactose is the energy source. Like the c-m1 allele
spontaneously to wild type.
of McClintock, these gal mutants could revert to wild type
spontaneously. The ability to revert suggested that they were • The mutants contained an extra segment of DNA
not simple deletions of the operon. Furthermore, adding inserted into the gal operon.
a chemical mutagen to the media did not increase the fre- • The inserted DNA could be found in different locations
quency of reversion. If the mutants were single–base-pair in the E. coli genome; that is, it moves.
IR Transposase gene IR
Density
Centrifuge Transposon Tn3
gal− virus
IR Transposase Resolvase ampR IR
Ampicillin
resistance
Viral DNA Insertion element
FIGURE 16-7 (a) Structure of IS1 including the transposase gene and
terminal inverted repeats (IRs). (b) Tn10, an example of a composite
transposon. The IS elements are inserted in opposite orientation
and form IRs. Each IS element carries a transposase, but only one
gal+ virus gal− virus is usually functional. (c) Tn3, an example of a simple transposon that
encodes its own transposase. The resolvase is a protein that promotes
(b) recombination (see Figure 16-10).
10
IS
Conservative
... A Tn10 B ... ... C D ...
illustrates, one copy of Tn3 is produced from an initial sin- sequence and then joins the free ends together, forming a
gle copy, yielding two copies of Tn3 altogether. Figure 16-10 fused circle called a cointegrate. The transposable element is
shows the details of the intermediates in the transposition duplicated in the fusion event. The cointegrate then resolves
of Tn3 from one plasmid (the donor) to another plasmid by a recombination-like event catalyzed by the resolvase
(the target). During transposition, the donor and recipient (see Figure 16-7) that turns a cointegrate into two smaller
plasmids are temporarily fused together to form a double circles, leaving one copy of the transposable element in
plasmid. The formation of this intermediate is catalyzed by each plasmid. Because a new copy of the element is created,
Tn3-encoded transposase, which makes single-strand cuts this mechanism is called “copy and paste.”
at the two ends of Tn3 and staggered cuts at the target
Conservative transposition Some transposons, such as
Tn10, excise from the chromosome and integrate into the
Replicative transposition of Tn3 target DNA. In these cases, the DNA of the element is not
replicated, and the element is lost from the site of the origi-
nal chromosome (see Figure 16-9). This mechanism is called
Donor
plasmid
“cut and paste.” Like replicative transposition, this reaction
Tn3 is initiated by the element-encoded transposase, which cuts
at the ends of the transposon. However, in contrast with rep-
+ Target sequence licative transposition, the transposase cuts the element out
of the donor site by making a double-stranded cut. (These
double-stranded breaks create the opportunity for a chromo-
Target some break, as seen by McClintock, if the cell fails to ligate
plasmid
the ends at the donor site back together.) It then makes a cut
at a target site and inserts the element into the target site. We
Transposase Formation of cointegrate
will revisit this mechanism in greater detail in a discussion of
the transposition of eukaryotic transposable elements.
An inserted element is flanked by a short repeat transposons. This element occurred in about 35 copies dis-
tributed on the 16 yeast chromosomes. These 5.6-kb ele-
ments had long terminal repeat (LTR) sequences of about
AGG T A AGG T AG
350 bp in length (Figure 16-12a). The geneticists also found
TCCAT TCCATC that there were many more LTRs than the 70 that formed
part of the 35 full-length elements, suggesting that some
Transposase cuts LTRs existed on their own within the genome. One tech-
target-site DNA.
nology used to make these discoveries was standard South-
ern blotting (Chapter 10). Using a cloned DNA probe for
AGG TAAGGTAG Ty, one can visualize about 35 bands on the Southern blot.
TCCATTCC ATC Later, it was learned that Ty elements are flanked by a 5-bp
direct repeat, much like bacterial transposons. Finally, com-
Transposable parison of parent and derivative yeast strains revealed that
element inserts.
new, “transposed” copies of Ty appeared in the genome in
the derivatives. All the evidence suggested that Ty was a
AGG TAAGGTAG transposon.
TCCATTCC ATC Subsequently, geneticists isolated two unstable yeast
mutations that would revert to wild type in the HIS4
Host repairs gene, which catalyzes histidine biosynthesis. The unstable
gaps. mutants were more than 1000 times as likely to revert to
wild type as the other HIS4 mutants. Symbolically, we say
AGGTAAGG TAAGGTAG
that these unstable mutants reverted from His− to His+.
T C C AT T C C ATTCCATC Like the E. coli gal− mutants, these yeast mutants were
found to harbor a large DNA insertion in the HIS4 gene.
Five-base-pair direct
However, for this work published in 1980, the insertion
repeat flanks element. was observed by simply determining its DNA sequence
using Sanger sequencing (Chapter 10) and not by buoyant
FIGURE 16-11 A short sequence of DNA is duplicated at the
density centrifugation as used with the E. coli IS elements.
transposon insertion site. The recipient DNA is cleaved at staggered sites
(a 5-bp staggered cut is shown), leading to the production of two copies The sequences of the insertions in HIS4 showed that
of the five–base-pair sequence flanking the inserted element. they were homologous to the Ty elements identified a decade
earlier. Moreover, they resembled a well-characterized class
of animal viruses called retroviruses ( Figure 16-12b ). A pol encodes the all-important reverse transcriptase, and
retrovirus is a single-stranded RNA virus that employs a env encodes a protein that is embedded in the viral mem-
double-stranded DNA intermediate for replication. The brane. This protein is necessary for the virus to leave the cell
RNA is copied into DNA by the enzyme reverse tran- to infect other cells. Interestingly, Ty elements have genes
scriptase . The double-stranded DNA is integrated into related to gag and pol but not env. These features led to the
host chromosomes, from which it is transcribed to pro- hypothesis that, like retroviruses, Ty elements are transcribed
duce the RNA viral genome and proteins that form new into RNA transcripts that are copied into double-stranded
viral particles. When integrated into host chromosomes as DNA by the reverse transcriptase. However, unlike retrovi-
double-stranded DNA, the double-stranded DNA copy of ruses, Ty elements cannot leave the cell because they do not
the retroviral genome is called a provirus. The life cycle of encode env. Instead, the double-stranded DNA copies are
a typical retrovirus is shown in Figure 16-13. Some retrovi- inserted back into the genome of the same cell. These steps
ruses, such as mouse mammary tumor virus (MMTV) and are diagrammed in Figure 16-14.
Rous sarcoma virus (RSV), are responsible for the induc- In 1985, scientists showed that, like retroviruses, Ty
tion of cancerous tumors. For MMTV, this happens when elements do in fact transpose through an RNA intermedi-
it inserts randomly into the genome next to a gene whose ate. Figure 16-15 diagrams their experimental design. They
altered expression leads to cancer. began by altering a yeast Ty element, cloned on a plasmid.
Figure 16-12 shows the similarity in structure and gene First, near one end of an element, they inserted a promoter
content of a retrovirus and the Ty element isolated from that can be activated by the addition of galactose to the
the HIS4 mutants. Both are flanked by LTR sequences that medium. This enabled them to control expression of the
are several hundred base pairs long. Retroviruses encode at element by adding galactose. Second, they introduced an
least three proteins that take part in viral replication: the intron from another yeast gene into the coding region of
products of the gag, pol, and env genes. The gag-encoded the Ty transposon as a reporter. If Ty was transcribed and
protein has a role in the maturation of the RNA genome, the mRNA processed, then the intron would be spliced out.
DNA
RNA
Double-stranded viral
DNA is integrated into
the DNA of the host
Proviral DNA chromosome. DNA
A retrotransposon is transposed through insertion of an element of the copia family into the white
an RNA intermediate locus. The insertion of LTR-retrotransposons into plant
genes (including maize) also has been shown to contribute
5-bp direct repeat of target DNA to spontaneous mutations in this kingdom.
TAATC TAATC Before we leave retrotransposons, one question needs
LTR gag pol LTR
ATTAG ATTAG to be answered. Recall that the first LTR-retrotransposon
Another was discovered in an unstable His− strain of yeast that
Transcription chromosome
reverted frequently to His+ . However, we have just seen
that LTR-retrotransposons, unlike most DNA transpos-
5' 3'
able elements, do not excise when they transpose. What,
then, is responsible for this allele’s ∼1000-fold increase in
Nu Ty1 copy
Cy
cle inserts reversion frequency when compared to other His− alleles?
us
top The answer is shown in Figure 16-16, which shows that
las
m
the Ty element in the His−allele is located in the promoter
region of the His gene, where it prevents gene transcrip-
tion. In contrast, the revertants contain a single copy of
Translation the LTR, called a solo LTR . This much smaller insertion
does not interfere with the transcription of the His gene.
Reverse
transcription
Reverse
transcriptase Demonstration of transposition through
an RNA intermediate
Plasmid with Ty
FIGURE 16-14 An RNA transcript from the retrotransposon undergoes
reverse transcription into DNA, by a reverse transcriptase encoded by the
retrotransposon. The DNA copy is inserted at a new location in the genome.
AN I M AT E D A RT The Ty1 mechanism of Coding region
retrotransposition LTR LTR
Explanation for His revertants 2.9 kb in length. This size difference is due to the pres-
ence of many defective P elements from which parts of
No
the middle of the element—encoding the transposase
transcription
gene—have been deleted, rendering them nonautono-
His –
LTR gag pol LTR His
mous elements.
The discovery of P elements traces back to the 1970s
Recombination and Margaret Kidwell, a professor at Brown University
who was studying natural (wild) populations of Dro-
Transcription sophila melanogaster and crossing wild and laboratory
Solo
His + strains. When she crossed a wild-strain female with a
LTR His lab-strain male, the F1’s were normal. However, when she
+ crossed a lab-strain female with a wild male, she noticed
FIGURE 16-16 His revertants contain a solo LTR that results from
recombination between the identical DNA sequences in the two LTRs of the the F1’s were sterile. Indeed, the F1’s showed a range of
LTR-retrotransposon in the His promoter. surprising phenotypes in the germ-line cells, including a
high mutation rate and a high frequency of chromosomal
aberration and nondisjunction (Figure 16-18). The defects in
The solo LTR is the product of recombination between the these F1 hybrid progeny made them dysgenic, that is, incapa-
identical LTRs, which results in the deletion of the rest of ble of breeding; hence, this phenomenon was called hybrid
the element (see Chapters 4 and 15 for more on recombina- dysgenesis. It occurs only when females from lab strains are
tion). Solo LTRs are a very common feature in the genomes mated with males derived from natural populations.
of virtually all eukaryotes, indicating the importance of this One observation about the dysgenic flies was that a
process. The sequenced yeast genome contains more than large percentage of the induced mutations were unstable;
fivefold as many solo LTRs as complete Ty elements. that is, they revert to wild type at a high frequency. The
unstable Drosophila mutants had similarities to the unsta-
KEY CONCEPT Retrotransposons, also known as class 1 ble maize mutants characterized by McClintock. Investi-
transposable elements, encode a reverse transcriptase that gators hypothesized that these unstable mutations were
produces a double-stranded DNA copy (from an RNA inter- caused by the insertion of transposable elements into spe-
mediate) that is capable of integrating at a new position in the
cific genes, thereby rendering them inactive. Reversion
genome.
would occur when the transposons excised from the genes.
These observations suggested that hybrid dysgenesis was
Class 2: DNA transposons somehow linked to transposable elements.
The role of transposable elements in hybrid dysgenesis was
Like IS elements, some eukaryotic mobile elements use a “cut
confirmed when an active family of elements was discovered
and paste” mechanism and physically move to a new position
in and molecularly isolated from the dysgenic flies. They called
in the genome after they excise. Elements that transpose in
these P elements. Interestingly, P elements are found in wild fly
this manner are called class 2 elements, or DNA transposons.
populations, but they are absent from laboratory strains. Thus,
The first transposable elements discovered by McClintock in
wild strains have become known as P strains and are said to
maize are now known to be DNA transposons. However, the
have P cytotype (cell type). The lab strains are called M strains
first DNA transposons to be molecularly characterized were
and said to have M cytotype. M stands for “maternal” and
the P elements in Drosophila.
P for “paternal,” symbolizing that dysgenesis occurs when the
lab (M) strain is maternal and the wild (P) strain is paternal.
P elements Of all the transposable elements in Dro-
sophila, the most intriguing and useful to geneticists KEY CONCEPT P elements are DNA transposons found in
are the P elements. The full-size P element resembles the wild strains of D. melanogaster. They were identified to cause
simple transposons of bacteria in that its ends are short hybrid dysgenesis when wild males were mated to lab-strain
(31-bp) inverted repeats and it encodes a single protein— females, but not when lab-strain males were mated to wild
the transposase—that is responsible for its mobilization females.
(Figure 16-17). P elements vary in size, ranging from 0.5 to
Why are P elements found only in wild strains? The answer
P-element structure to this question requires a bit of history. Drosophila melano-
P element
gaster is native to the Old World, and it was brought to the
Americas only in the post-Columbian era. Another species,
Transposase gene
Drosophila willistoni, is native to the Americas, and critically
it contains P elements. In the early 1900s, D. melanogaster
FIGURE 16-17 DNA sequence analysis of the 2.9-kb autonomous was brought out of the wild and into the laboratory for use
P element reveals a gene that encodes transposase. A perfect 31-bp in genetic research by T. H. Morgan. At that time, it was
inverted repeat resides at each of the element’s termini. expected that P elements did not exist in wild populations of
× ×
♀ ♂ ♂ ♀
♀ ♂ ♂ ♀
D. melanogaster at all. As the lab strains were maintained in single protein, a transposase. The nonautonomous Ds ele-
captivity over the years and shared among researchers, they ment does not encode transposase and thus cannot trans-
remained genetically isolated from their wild relatives. Mean- pose on its own. When Ac is in the genome, the transposase
while in nature, D. melanogaster populations were evolving it encodes can bind to the ends of Ac or Ds elements and
separately from their captive cousins. The hypothesis is that promote their transposition (Figure 16-19).
the P element was transferred horizontally from D. willistoni Ac and Ds are members of a single transposon family, and
into wild D. melanogaster, and then it spread rapidly through- there are other families of transposable elements in maize. Each
out the global population. The transfer may have been accom- family contains autonomous elements encoding a transposase
plished by a virus that infected D. willistoni, picked up the that can mobilize elements in the same family but cannot mobi-
P element, and then infected D. melanogaster. lize elements in other families because the transposase can bind
When the P element was first introduced into wild only to the specific inverted repeat DNA sequence at the end of
D. melanogaster, it likely caused hybrid dysgenesis. How- the elements of its family members. Although some organisms
ever, over time, natural populations evolved a mechanism to such as yeast have no DNA transposons, elements structurally
repress P transposition in the germ line. The mechanism to similar to the P or Ac elements have been isolated from many
repress P elements never evolved in the laboratory strains. plant and animal species.
Later in this chapter, we will review the mechanism for
P element repression and why hybrid dysgenesis appears KEY CONCEPT DNA transposons structurally resemble DNA
only when an M female (no P elements) is mated with a bacterial IS elements and are found in many eukaryotes. DNA
P male (P elements), but not in the reciprocal cross. transposons encode a transposase that cuts the transposon
from the chromosome and catalyzes its reinsertion at other
chromosomal locations.
Maize transposable elements revisited Although the
causative agent responsible for unstable mutants was first
shown genetically to be transposable elements in maize, it
Utility of DNA transposons as tools for
was almost 50 years before the maize Ac and Ds elements
were isolated and shown to be related to DNA transposons genetic research
in bacteria and in other eukaryotes. Like the P element of Quite apart from their interest as a genetic phenome-
Drosophila, Ac has terminal inverted repeats and encodes a non, DNA transposons have become important tools for
Activator transposase catalyzes excision of typical maize inbred lines. The active autonomous Mu
and integration element, called MuDr, is only in a few special maize stocks.
By crossing a stock with the active MuDr to one that lacks
Ac MuDr, the nonautonomous elements are activated in the F1.
Activation can be confirmed using an allele of Bronze locus
Transposase
(bz-mum) that has a Mu insertion (Figure 16-20a). When Mu
Transposase binds
the ends of Ac
and Ds elements.
Ds Ac
The maize Mutator transposon
can be used to knock out genes
(a)
Cleavage
Integration into
Target new target site
DNA Mu is off (stable) Mu is on (mutagenic)
(b)
Ac or Ds at Segment of genomic DNA
new location Mu
Ac or Ds gene 1 gene 2
excises from Bronze, there are small blue sectors on a bronze NextGen sequencing (Chapter 10) is applied to pools of the
background in the kernels. Once activated, the nonautono- PCR products that are tagged with “DNA barcodes” so that
mous Mu elements cause new mutations in many genes in the individual sequence reads can be traced back to specific
the F1’s. When the F1’s are selfed, 3/4 of the progeny will have plants. The DNA sequences are collected into a database that
MuDr and sectored kernels, but ¼ will lack MuDr and have can be queried by BLAST (Chapter 14) with the sequence of
stable bronze kernels. Because MuDr was segregated away a known gene from the maize genome sequence. The barcode
in the unsectored bronze kernels, the new insertion alleles is used to trace the sequence back to a specific parent plant
that they carry will be stable mutant alleles. If one creates a for which selfed seed has been saved and which will segregate
large enough population, then potentially every gene in the for the Mu insertion allele in the gene of interest.
genome will have one or more Mu insertions.
The next step is to find a Mu insertion allele in a gene Using P elements to insert genes Geneticists have also
of interest. This is done by creating a database of the DNA shown that P-element DNA can be an effective vehicle for trans-
sequences flanking the Mu insertions. DNAs from the plants ferring donor genes into the germ line of a recipient fly; that is,
with new stable Mu insertion alleles are used as substrates to make a transgenic fly. They devised the following experimen-
for PCR (Figure 16-20b). The DNA is sheared and a short tal procedure (Figure 16-21). Suppose the goal is to transfer the
adapter sequence is ligated onto the ends of the DNA frag- wild-type allele of rosy (ry+ ), which confers a characteristic eye
ments. Then, PCR is performed with a primer that matches color, into the fly genome. The recipient genotype is homozy-
the Mu terminal repeat and one for the adapter sequence. gous for the rosy (ry− ) mutation. From this strain, embryos are
Nuclei
Anterior Posterior
Eventual location
Drosophila embryo of germ cells
ry –/ry –
Grow to adult
rosy + eyes
Chromosome ry + gene
Some
ry +/ry –
progeny
collected at the completion of about nine nuclear divisions. At and these repetitive sequences can make up a large fraction
this stage, the embryo is one multinucleate cell, and the nuclei (up to 90 percent) of eukaryotic genomes.
destined to form the germ cells are clustered at one end. Two Thanks to the many projects to sequence the com-
types of DNA are injected into embryos of this type. The first plete genomes of a wide variety of organisms (including
is a bacterial plasmid carrying a defective P element into which Drosophila, humans, the mouse, Arabidopsis, maize, and
the ry+ gene has been inserted. The defective P element does not rice), we now know that there are many classes of repet-
encode transposase but still has the ends that bind transposase itive sequences in the genomes of higher organisms and
and allow transposition. This defective element is not able to that some of these repeat elements are similar to the DNA
transpose, and so, a second (helper) plasmid encoding trans- transposons and retrotransposons discussed in this chapter.
posase but without the terminal repeats (so it cannot transpose) Most remarkably, these repetitive sequences make up most
also is injected. Flies developing from these embryos are pheno- of the DNA in the genomes of most multicellular eukary-
typically still rosy mutants, but their offspring include a large otes. The portion of the genomes that are composed of
proportion of ry+ flies. Follow-up experiments show that the repetitive sequences varies widely among species.
ry+ gene within the deleted P element was inserted into one of We now see that variation in genome size results from
several distinct chromosome locations. These new ry+ genes are variation in the numbers of repetitive sequences and not the
stable and inherited in a Mendelian fashion. numbers of genes. Barley and rice both have about 40,000
genes. The 10-fold larger genome of barley is due to a much
KEY CONCEPT DNA transposons have been used by sci- greater amount of repetitive DNA, most of which is either
entists in two important ways: (1) to make new mutant alleles transposable elements or decayed transposable elements.
with transposon insertions; and (2) to serve as vectors that can
Rather than correlating with the numbers and sizes of gene
introduce foreign genes into a chromosome.
in a genome, genome size frequently correlates with the
amount of DNA in the genome that is derived from trans-
posable elements. Organisms with big genomes have lots
16.4 THE DYNAMIC GENOME: of sequences that resemble transposable elements, whereas
MORE TRANSPOSABLE organisms with small genomes have many fewer. Two
examples, one from the human genome and the other from
ELEMENTS THAN EVER a comparison of the plant genomes, illustrate this point.
IMAGINED
KEY CONCEPT Genes make up only a small proportion of
LO 16.5 Predict the short-term and evolutionary fate of the genomes of multicellular organisms. Genome size usually
transposable elements in a species. correlates with the amount of transposable-element sequences
and not gene number.
Once geneticists had discovered transposable elements and
determined the DNA sequences and molecular structures,
Transposable elements in the human
new sets of questions could be addressed. How prevalent
are transposable elements in genomes? Are there other
genome
families of transposable elements in the genome that have Almost half of the human genome is derived from trans-
remained unknown because they had not caused a muta- posable elements. The vast majority of these transposable
tion that could be studied in the laboratory? Do all organ- elements are two types of retrotransposons called long
isms have transposable elements in their genomes? In this interspersed elements, or LINEs, and short i nterspersed
section, we will review how these questions were addressed elements, or SINEs ( Figure 16-22 ). LINEs move like a
and the answers that were found. r etrotransposon with the help of an element-e ncoded
reverse transcriptase but lack some structural features of
retrovirus-like elements, including LTRs (see Figure 16-12d).
Large genomes are largely SINEs can be best described as nonautonomous LINEs
transposable elements because they have the structural features of LINEs but do
Long before the advent of DNA-sequencing projects, scien- not encode their own reverse transcriptase. Presumably,
tists using a variety of molecular techniques discovered that they are mobilized by reverse transcriptase enzymes that are
genome size varied dramatically among eukaryotic species encoded by LINEs residing in the genome.
and did not correlate with an organism’s biological complex- The most abundant SINE in humans is called Alu, so named
ity. For example, the genomes of salamanders are 20 times as because it contains a target site for the Alu restriction enzyme.
large as the human genome, whereas the genome of barley is The human genome contains more than 1 million whole and
more than 10 times as large as the genome of rice. In addi- partial Alu sequences, scattered between genes and within
tion, the genomes of all eukaryotes contain repetitive DNA introns. These Alu sequences make up more than 10 percent of
elements. Specifically, there are some short DNA sequences the human genome. The full Alu sequence is about 300 nucleo
(hundreds to a few thousand bp in length) that are repeated tides long and bears a remarkable resemblance to 7SL RNA,
thousands, even hundreds of thousands, of times in a genome, an RNA that is part of a complex by which newly synthesized
Fraction
Element Transposition Structure Length Copy number
of genome
ORF1 ORF2 (pol )
LINEs Autonomous AAA 1– 5 kb 20,000 – 40,000 21%
transposase
DNA Autonomous 2 – 3 kb
transposons 300,000 3%
Nonautonomous 80 – 3000 bp
FIGURE 16-22 Several general classes of transposable elements are found in the human genome.
[Data from Nature 409, 880 (15 February 2001), “Initial Sequencing and Analysis of the Human Genome,” The
International Human Genome Sequencing Consortium.]
polypeptides are targeted to the endoplasmic reticulum (see plants survive and thrive with so much mobile DNA in their
Figure 9-25). Presumably, the Alu sequences originated as genomes? Several factors come into play. As we will see in
reverse transcripts of these RNA molecules. Section 16.5, organisms suppress transposon activity so that
There is about 20 times as much DNA in the human the elements are inactive most of the time and do not cause new
genome derived from transposable elements as there is mutations. When transposons do move, they may insert into
DNA encoding all human proteins. Figure 16-23 illustrates exons, introns, or noncoding regions of the genome between
how diverse types of transposons are distributed within genes. If a transposon inserts into an exon, it will disrupt the
and between genes in the human genome. Alu elements are coding sequence, which is apt to destroy protein function. Such
often found within introns. a deleterious insertion will be removed from the population by
If such a large fraction of the human and other eukary- Darwinian selection (Chapters 19 and 20). If they insert into an
otic genomes is composed of intact or decayed transposable intron, the mRNA produced by the gene will not include any
elements, an obvious question arises: how do animals and sequences from transposable elements because they will have
Gene
Gene
Exon LINE
FIGURE 16-23 Distribution of transposable elements in a typical segment of the human genome.
Transposons are found in introns and intergenic regions, but not in exons of functional alleles.
been spliced out of the pre-mRNA with the surrounding intron. Plants: LTR-retrotransposons thrive in
So, many of the transposons that survive in genomes are safely large genomes
hidden away in introns. Similarly, if they insert in the noncod-
ing regions between genes, they may not affect gene function In plants, differences in the genome sizes of different species
unless they insert into a regulatory element. Thus, transposons have been shown to correlate primarily with the number of
accumulate over evolutionary time in parts of the genome one class of elements, the LTR-retrotransposons. Plants share
where they do not affect gene function. a common biology and homologous organs, including roots,
When transposons do insert into critical regions of stems, leaves, and flowers, and as such, their genomes are sim-
the genome, they can cause severe mutations that remain ilar with respect to gene content. Despite these similarities, the
in the population at least transiently. Three separate inser- genome size of plants varies widely, from about 125 Mbp for
tions of LINEs have disrupted the coagulation factor VIII Arabidopsis thaliana (mustard weed) to 5100 Mbp for barley
gene, causing hemophilia A. At least 11 Alu insertions into (Figure 16-24). For plants like Arabidopsis with small genomes,
human genes have been shown to cause several diseases, LTR-retrotransposons comprise less than 10 percent of the
including hemophilia B (in the coagulation factor IX gene), genome, but for plants like barley with large genomes, this class
neurofibromatosis (in the NF1 gene), and breast cancer (in of transposon comprises about 80 percent of the genome. Most
the BRCA2 gene). of the expansion in plant genome size is due to growth in the
numbers of LTR-retrotransposons.
KEY CONCEPT Transposable elements compose the largest
fraction of the human genome, with LINEs and SINEs being
the most abundant. The vast majority of transposable Safe havens
elements can no longer move or increase their copy number.
The abundance of transposable elements in the genomes of
A few elements remain active, and their movement into genes
multicellular organisms led some investigators to postulate that
can cause disease.
successful transposable elements (those that attain very high
100
90 LTR-retrotransposons
5000
80
Transposon percentage of genome
70
4000
50 3000
40
2000
30
20
1000
10
0 0
is ry m fa e t e a le to o n ze y
s r
di
u lfa
l
Ri
c ille ra
p
na
n
hu
m
pp a at
be
a ai ar
le
op be M rg A ot m M
bi
d
ra
w
ypo A G Ba So
P To So
y B
Ar
a St ch
Bra
FIGURE 16-24 Flowering plants first evolved about 200 million years ago. Since that time,
transposable elements have accumulated to different levels in each species. Maize and barley have
genomes with large numbers of LTR-retrotransposons. Arabidopsis has a small genome with fewer
LTR-retrotransposons. The inset shows how LTR-retrotransposons preferentially insert into other
LTR-retrotransposons (safe havens) rather than critical genes.
copy numbers) have evolved mechanisms to prevent harm to How are transposable elements able to spread to new
their hosts by not inserting into host genes. Instead, success- sites in genomes with few safe havens? Investigators have
ful transposable elements insert into so-called safe havens in identified hundreds of Ty elements in the sequenced yeast
the genome. For the grasses, a safe haven for new insertions genome and have determined that they are not randomly
appears to be into other retrotransposons that are located in distributed. Instead, each family of Ty elements inserts
the regions between genes (see Figure 16-24, insert). Another into a specific genomic region. For example, the Ty3 fam-
safe haven is the heterochromatin of centromeres, where there ily inserts almost exclusively near but not in tRNA genes,
are very few genes but lots of repetitive DNA (see Chapter 12 at sites where they do not interfere with the production
for more on heterochromatin). Many classes of transposable of tRNAs and, presumably, do not harm their hosts. This
elements in both plant and animal species tend to insert into the region-specific integration is made possible by a mecha-
centromeric heterochromatin. nism that evolved in Ty elements: the proteins necessary
In contrast to the genomes of multicellular eukary- for integration interact with specific yeast proteins bound
otes, the genome of unicellular yeast is very compact, with to genomic DNA. Ty3 proteins, for example, recognize and
closely spaced genes and very few introns. Because almost bind to subunits of the RNA polymerase complex that have
70 percent of its genome consists of exons, there is a high assembled at tRNA promoters (Figure 16-25a).
probability that new insertions of transposable elements The ability of some transposons to insert preferentially
will disrupt a coding sequence. Yet, as we have seen earlier into certain sequences or genomic regions is called targeting.
in this chapter, the yeast genome supports a collection of A remarkable example of targeting is illustrated by the R1 and
LTR-retrotransposons called Ty elements. R2 elements of arthropods, including Drosophila. R1 and R2
TFIIIB
Integrase TFIIIC
tRNA gene
Ty3 DNA
rRNA gene R1 R1 R1
2 kb
R2 RT
1 kb
R1 RT
FIGURE 16-25 Some transposable elements are targeted to specific safe havens. (a) The yeast
Ty3 retrotransposon inserts near transfer RNA genes. (b) The Drosophila R1 and R2 non-LTR-
retrotransposons (LINEs) insert into the genes encoding ribosomal RNA that are found in long tandem
arrays on the chromosome. Only the reverse transcriptase (RT) genes of R1 and R2 are noted.
are LINEs (see Figure 16-22) that insert only into the genes to excise from the unc-22/Tc1 allele in the germ line and
that produce ribosomal RNA. In arthropods, several hundred revert the twitching phenotype to wild type (unc-22). To
rRNA genes are organized in tandem arrays (Figure 16-25b). this end, they exposed the mutant unc-22/Tc1 strain to a
With so many genes encoding the same product, the host tol- chemical (called a mutagen; see Chapter 16) that greatly
erates insertion into a subset. However, too many insertions of increased the frequency of mutation and examined their
R1 and R2 have been shown to decrease insect viability, pre- progeny under a microscope, searching for rare worms that
sumably by interfering with ribosome assembly. no longer twitched.
This and subsequent genetic screens identified over 25
KEY CONCEPT A successful transposable element increases C. elegans genes that, when mutated, allowed the host to
copy number without harming its host. One way in which an excise Tc1 in the germ line. Significantly, many of the prod-
element safely increases copy number is to target new inser- ucts of these genes are integral components of the RNAi
tions into safe havens, regions of the genome where there are
silencing pathway, including proteins found in Dicer and
few genes.
RISC (see Chapters 8 and 12). Recall from Chapter 8 that
Dicer binds to long dsRNAs and cleaves them into small
dsRNA fragments. These fragments are then unwound so
16.5 REGULATION OF that one strand, the siRNA, can target RISC to chop up
TRANSPOSABLE ELEMENT complementary mRNAs (see Figure 8-29).
MOVEMENT BY THE HOST Beginning with this elegant genetic screen, many years
of experimentation have led to the following model for
LO 16.5 Predict the short-term and evolutionary fate of the repression of transposable elements in the germ line of
transposable elements in a species. C. elegans. With 32 Tc1 elements scattered throughout the
C. elegans genome, a few elements near genes are tran-
The repression of transposable elements was first investigated scribed along with the nearby gene (Figure 16-27). Because
in the late 1990s, using the model organism Caenorhab- the ends of Tc1 are 54-bp terminal inverted repeats, the Tc1
ditis elegans (a nematode; see the Model Organism box in RNA spontaneously forms dsRNA. Like all dsRNAs pro-
Chapter 13). This story starts with the observation of a strik- duced in most eukaryotes, this RNA is recognized by Dicer
ing difference between the mobility of a transposable ele-
ment called Tc1 in two different cell types of this model
organism. Tc1 is a DNA transposon that, like the Ac ele- A mutant screen leads to genes required to
ment of maize, can lead to an unstable mutant phenotype repress transposition
when it excises from a gene with a visible phenotype.
unc-22 Tc1 unc-22
There are 32 Tc1 elements in the sequenced genome of
the common laboratory strain of C. elegans. Significantly, unc-22 / Tc1
allele (in all cells)
Tc1 transposes in somatic but not in germ-line cells. That
observation suggested that transposition is repressed in
the germ line by the host. Evidently, germ-line repression
results from the silencing of the transposase genes of all 32
Tc1 copies in germ-line cells.
1 Mutagenize strain.
2 Isolate and grow progeny.
RNAi silencing of transposable elements
3 Observe under microscope.
Researchers set out to identify C. elegans genes responsi- Tc1
ble for silencing the transposase gene. They began with a Most progeny
C. elegans strain that had Tc1 inserted in the unc-22 gene have the
twitching
(designated unc-22/Tc1; Figure 16-26). Whereas wild-type Twitching phenotype
C. elegans glides smoothly on the surface of the agar in worms with the
a petri dish (as illustrated by horizontal arrows in Fig- unc-22 / Tc1
allele.
ure 16-26), worms with the mutant unc-22/Tc1 gene
have a twitching movement (as illustrated by vertical In rare gliders,
arrows in Figure 16-26) that can be easily observed with Smooth gliding Tc1 has been
a microscope. Because Tc1 cannot normally transpose in worm excised from
unc-22.
the germ line, it remains inserted in the unc-22 gene and
continues to disrupt its function. Thus, the strain with
the mutant unc-22/Tc1 gene should express a twitch- FIGURE 16-26 Experimental design used to identify genes required to
repress transposition. Investigators look for mutants that have regained
ing phenotype from generation to generation. However, normal movement because mutations in these individuals would have
researchers reasoned that mutations that inactivated disabled the repression mechanism that prevents the transposition of the Tc1
C. elegans genes required for repression would allow Tc1 element from the unc-22 gene.
Dicer
All transposase
mRNA targeted
for degradation
RISC siRNA
RISC
and ultimately siRNA is produced, which directs RISC to all transposable-element (TE) family members. Two other
chop up complementary Tc1 transcripts. Because all Tc1 types of genome radar (also called genome surveillance) have
RNA is efficiently chopped up in the germ line, the ele- been described that utilize different classes of small noncod-
ment-encoded transposase gene is silenced. Without trans- ing RNAs to target “invasive” nucleic acids including trans-
posase, the element cannot excise. It has been hypothesized posons and viruses (see Chapter 6). These mechanisms are
that Tc1 can transpose in somatic cells because RNAi is not presented here because they illustrate how different solutions
as efficient and some transposase can be produced. evolve to solve similar biological problems.
Over the past decade, numerous laboratories work-
ing with both plants and animals have discovered that
piRNAs in animals In the germ lines of animal species
mutations that disrupt the RNAi pathway often lead to
including Drosophila, active transposons are repressed
the activation of transposable elements in their respective
through the action of piRNAs (short for Piwi-interacting
genomes. Because of the abundance of transposable ele-
RNAs). Animal genomes contain several long (often >100 kb)
ments in eukaryotic genomes, it has been suggested that
loci called pi-clusters that serve as traps that ensnare active
the RNAi pathway evolved to maintain genome stability
transposons as they insert randomly around the genome.
by repressing the movement of transposable elements.
A pi-cluster can contain remnants of several different trans-
KEY CONCEPT Eukaryotes use RNAi to repress the expres-
poson families that represent a historical record of prior
sion of active transposable elements in their genomes. In this insertions of active transposons into it. Long RNAs are tran-
way, a single element that inserts near a gene can be tran- scribed from the pi-clusters and then processed into short sin-
scribed to produce dsRNA that will trigger the silencing of all gle-stranded piRNAs of 23–30 nt in length. piRNAs form an
copies of the element in the genome. RNA-protein complex with the protein Piwi-Argonaute. The
piRNA-Piwi complex has two effects. First, the piRNA-Piwi
complex guides the degradation of mRNAs complementary to
Genome surveillance the piRNA, that is, mRNAs from the transposons that com-
The RNAi silencing pathway is akin to radar in that enables pose the pi-cluster (Figure 16-28). By degrading the mRNAs
the host to detect new insertions of transposons into the complementary to transposases, any transposons represented
genome if they generate antisense RNA. The host then in a pi-cluster are silenced. Second, the piRNA-Piwi complex
responds by producing siRNAs that target the transposase directs the placement of histone marks on the pi-cluster chro-
mRNA, silencing the gene and preventing the movement of matin to promote the transcription of the long RNAs from it.
Transcription Transcription
Inactive element
not transcribed
Processing Processing
piRNA mRNA
piwi-
Argonaute Translation
Transposition of
Genome surveillance “yellow” elements in
genome
This latter step enables the production of the piRNAs needed deposited in cells of the primordial germ line. Here, these
for the first step. piRNAs direct histone marks to be placed on the pi-clusters,
allowing them to be transcribed to make piRNAs for the
piRNA-Piwi complex. Essentially, the maternal piRNAs from
piRNAs and hybrid dysgenesis in Drosophila Now,
the P female jump-start the system to repress P elements and
let’s return to Drosophila P elements. As discussed earlier,
allow healthy P(female) × M (male)F1’s to develop.
P elements entered the genome of wild D. melanogaster a
In the reciprocal cross, an M female has no P ele-
little over 100 years ago. At first, they were probably active
ments in her pi-clusters, and so no maternal piRNAs com-
and jumping, causing many mutations. As the elements
plementary to P elements are loaded into her embryos
inserted randomly around the genome, eventually one or
(Figure 16-29, left). Sperm from a P stock do not deposit
more landed in a pi-cluster, after which P elements could
piRNAs into the embryo. The result is that the F 1 has
be repressed in wild Drosophila by the piRNA-Piwi com-
Introduction to Genetic Analysis, 11e P elements from the male parent but no maternally loaded
Figure 15.27 #1531 plex. Since the laboratory strains of D. melanogaster lacked
piRNAs to activate the piRNA-Piwi complex to repress
06/30/14 P elements, so too did their pi-clusters. Thus, there would
07/23/14
their movement. As development proceeds, the P elements
be no means for the piRNA-Piwi complex to repress P ele-
08/04/14 are mobilized, cause genomic havoc in the germ line, and
ment activity in lab stocks.
08/25/14 the M (female) × P(male)F1’s are dysgenic.
08/26/14 Next, let’s consider why hybrid dysgenesis does not occur
Dragonfly Media Group when a P cytotype female is mated with an M cytotype male
but does occur in the reciprocal mating. The P elements of KEY CONCEPT Like siRNAs, piRNAs in animals interact with
Drosophila are repressed in the germ line by piRNA-Piwi protein complexes and guide them to degrade complementary
sequences in transposons. These small noncoding RNAs
early in embryonic development (Figure 16-29, right). Repres-
have their origin in long RNAs transcribed from pi-clusters that
sion is initially activated by maternally produced piRNAs
capture fragments of invasive DNA.
from the P female that are loaded into the embryo and
♀ ♂ piRNA ♂ ♀ piRNA
Fertilized eggs
and early embryos
♀ ♂ ♂ ♀
SUMMARY
Transposable elements were discovered in maize by Barbara elements (DNA transposons). The P element of Drosophila
McClintock as an astounding feature of the genome, was the first eukaryotic class 2 DNA transposon to be isolated
stretches of DNA (loci) that could move from one location molecularly. It was isolated from unstable mutations that were
to another. McClintock determined that some transposons, induced in crosses between P and M cytotype strains. When P
like Ac, are autonomous and able to move on their own; elements are mobilized in P × M crosses, the F1’s exhibit hybrid
but others, like Ds, are nonautonomous and require the dysgenesis. P elements have been developed into vectors for
presence of an autonomous element for mobilization. Trans- the introduction of foreign DNA into Drosophila germ cells.
posons that insert into genes can produce unstable mutant Ac, Ds, and P are examples of DNA transposons, so
alleles that revert to wild type when the element excises. named because the transposition intermediate is the DNA
Bacterial insertion-sequence elements were the first element itself. Autonomous elements such as Ac encode a
transposable elements isolated molecularly. There are many transposase that binds to the ends of autonomous and non-
different types of IS elements in E. coli strains, and they are autonomous elements and catalyzes excision of the element
usually present in at least several copies. Composite trans- from the donor site and reinsertion into a new target site
posons contain IS elements flanking one or more genes, such elsewhere in the genome.
as genes conferring resistance to antibiotics. Transposons Retrotransposons were first molecularly isolated from
with resistance genes can insert into plasmids and can then yeast mutants, and their resemblance to retroviruses was
be transferred by conjugation to nonresistant bacteria. immediately apparent. Retrotransposons are class 1 ele-
There are two major groups of transposable elements in ments, as are all transposable elements that use RNA as
eukaryotes: class 1 elements (retrotransposons) and class 2 their transposition intermediate.
DNA sequencing of whole genomes, including the rare because of two factors. First, most of the transposable
human genome, has led to the remarkable finding that elements in eukaryotic genomes cannot move because inac-
almost half of the human genome is derived from transpos- tivating mutations prevent the production of normal trans-
able elements. For some plants like maize, transposable ele- posase and reverse transcriptase. Second, expression of the
ments compose fully 80 percent of the DNA in the genome. vast majority of the remaining elements is silenced by the
Despite having so many transposable elements, eukaryotic RNAi and/or the piRNA pathways.
genomes are extremely stable, as transposition is relatively
KEY TERMS
Activator (Ac) element (p. 530) insertion-sequence (IS) element retrovirus (p. 539)
Alu (p. 545) (p. 534) reverse transcriptase (p. 539)
autonomous transposable element inverted repeat (IR) sequence (p. 534) R plasmid (p. 535)
(p. 533) long interspersed element (LINE) safe haven (p. 548)
class 1 element (retrotransposon) (p. 545) short interspersed element (SINE)
(p. 540) long terminal repeat (LTR) (p. 538) (p. 545)
class 2 element (DNA transposon) LTR-retrotransposon (p. 540) simple transposon (p. 535)
(p. 541) M cytotype (p. 541) solo LTR (p. 540)
cointegrate (p. 537) nonautonomous transposable element targeting (p. 548)
composite transposon (p. 535) (p. 533) target-site duplication (p. 537)
conservative transposition (p. 536) P cytotype (p. 541) transposable element (p. 528)
copia-like element (p. 540) P element (p. 541) transposase (p. 533)
Dissociation (Ds) element (p. 530) pi-cluster (p. 550) transpose (p. 533)
DNA transposon (p. 541) piRNAs (p. 550) transposition (p. 536)
excise (p. 533) provirus (p. 539) transposon (p. 528)
genome surveillance (p. 550) replicative transposition (p. 536) Ty element (p. 538)
hybrid dysgenesis (p. 541) retrotransposon (p. 540) unstable allele (p. 533)
PROBLEMS
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the www
WORKING WITH THE FIGURES are caused by a transposon in the chalcone synthase
(The first 8 questions require inspection of text figures.) gene, how would this be manifest by the results of the
PCR with the three leaf DNA samples? The PCR with
1. For Figure 16-3b , if the Ds element were located
leaf DNA indicates there is a transposon in the chal-
between C and Sh, what phenotype would you expect
cone synthase gene. How could you investigate this
in the sectors?
hypothesis further using isolated DNA from the corol-
2. In Figure 16-5 , the fly has a mariner transposon las of the sectored flowers?
inserted in the white locus.
11. Propose a model for the generation of a multiple-
a. The wild-type allele is w+ . Make up an allele desig- drug-resistant plasmid.
nation for the mariner insertion allele.
12. Propose an experiment to prove that the transposition
b. Referring back to what you learned in Chapter 2, of the Ty element in yeast takes place through an RNA
what is the genotype for the white locus in the red intermediate.
and white sectors if the fly shown is male and if it
13. Explain how the properties of P elements in Drosoph-
is female?
ila make gene-transfer experiments possible in this
3. For Figure 16-8, draw out a series of steps that could organism.
explain the origin of this large plasmid containing
14. Although DNA transposons are abundant in the
many transposable elements.
genomes of multicellular eukaryotes, class 1 elements
4. Draw a figure for the mode of transposition not shown usually make up the largest fraction of very large
in Figure 16-9, retrotransposition. genomes. Given what you know about class 1 and
5. In Figure 16-11, show where the transposase would class 2 elements, what is it about their distinct mech-
have to cut to generate a 6-bp target-site duplication. anisms of transposition that would account for this
Also show the location of the cut to generate a 4-bp consistent difference in abundance?
target-site duplication. 15. As you saw in Figure 16-23, the genes of multicellular
6. If the transposable element in Figure 16-15 were a eukaryotes often contain many transposable elements.
DNA transposon that had an intron in its transposase Why do most of these elements not affect the expres-
gene, would the intron be removed during transposi- sion of genes?
tion? Justify your answer. 16. What are safe havens? Are there any places in the
7. For Figure 16-23, draw the pre-mRNA that is tran- much more compact bacterial genomes that might be
scribed from one of the genes shown and then draw its a safe haven for insertion elements?
mRNA. 17. Nobel Prizes are usually awarded many years after the
8. For Figure 16-29, how would you modify this figure actual discovery. For example, James Watson, Francis
for a P cytotype × P cytotype cross and an M cytotype × Crick, and Maurice Wilkens were awarded the Nobel
M cytotype cross? Prize in Medicine or Physiology in 1962, almost a
decade after their discovery of the double-helical
structure of DNA. However, Barbara McClintock was
BASIC PROBLEMS awarded the Nobel Prize in 1983, almost four decades
9. At a garden shop, you notice that there are three variet- after her discovery of transposable elements in maize.
ies of petunia for sale: one with blue pigmented corollas, Why do you think it took this long for the significance
one with white corollas, and one with small patches (sec- of her discovery to be recognized in this manner?
tors) of blue on a white background. What hypotheses 18. Why can’t retrotransposons move from one cell to
would you propose to explain these three phenotypes? another like retroviruses?
Can you design any experiments to test your hypotheses? a. Because they do not encode the Env protein.
10. Working with the petunias from Problem 9, you iso- b. Because they are nonautonomous elements.
late DNA from leaves of each and design PCR prim-
ers to amplify the chalcone synthase gene, an enzyme c. Because they require reverse transcriptase.
involved in pigment biosynthesis. If sectored flowers d. Both a and b are true.
19. Which of the following is true of reverse transcriptase? 24. In Drosophila, there is a singed allele (sncm) with some
a. It is required for the movement of DNA trans- unusual characteristics. Females homozygous for this
posons. X-linked allele have singed bristles, but they have
numerous patches of sn+ (wild-type) bristles on their
b. It catalyzes the synthesis of DNA from RNA. heads, thoraxes, and abdomens. When these flies are
c. It is required for the transposition of retrotrans- mated with sn− males, some females give only singed
posons. progeny, but others give both singed and wild-type
d. Both b and c are correct. progeny in variable proportions. Explain these results.
20. What is the major reason why the barley genome is 25. Consider two maize plants:
much larger than the rice genome? a. Genotype C /c m ; Ac /Ac+ ,where c m is an unstable
a. Barley has more genes than rice. allele caused by a Ds insertion
b. Rice has more genes than maize. b. Genotype C/c m , where c m is an unstable allele
caused by Ac insertion
c. Barley has more DNA transposons than rice.
What kernel phenotypes would be produced and in
d. Barley has more retrotransposons than rice. what proportions when (1) each plant is crossed with a
21. If 80% or even 90% of an organism’s genome is com- stable recessive loss of function mutant c/c and (2) the
posed of largely decayed retrotransposons and other plant in part a is crossed with the plant in part b? Assume
classes of transposons, what cost, if any, would you that Ac and c are unlinked, that the chromosome-
expect this to exert on the organism? Why do you sup- breakage frequency is negligible, and that mutant c/c
pose the organism does not simply delete all of this is Ac+.
“junk” DNA? 26. The yeast His− mutant with the Ty retroelement inser-
tion is unstable and undergoes a high rate of reversion
CHALLENGING PROBLEMS to wild-type His+ . The high rate of reversion is notable
because retroelements have no means for excision. The
22. The insertion of transposable elements into genes can
mechanism of reversion is recombination between the
alter the normal pattern of expression. In the following
direct repeats to leave only a solo LTR, while the rest
situations, describe the possible consequences on gene
of the element is deleted. Diagram the recombination
expression.
event that would give this outcome.
a. A LINE inserts into an enhancer of a human gene.
27. DNA transposons use a “cut and paste” mechanism by
b. A transposable element contains a binding site for which they are excised from one genomic location and
a transcriptional repressor and inserts adjacent to a pasted into another. This is a conservative or nonrepli-
promoter. cative process. Yet, DNA elements, like the Drosophila
c. An Alu element inserts into the 3′ splice (AG) site P element, can enter a genome as a single copy and
of an intron in a human gene. rise to have 30 or more copies over time. Describe sev-
eral mechanisms by which the copy number of such
d. A Ds element that was inserted into the exon of
elements increases over time in a species.
a maize gene excises imperfectly and leaves three
base pairs behind in the exon. 28. The evidence indicates that Drosophila P elements first
entered the D. melanogaster genome in the Americas in
e. Another excision by that same Ds element leaves
the early 1900s and then spread worldwide through-
two base pairs behind in the exon.
out all D. melanogaster populations by the 1980s
f. A Ds element that was inserted into the middle of such that today there are no longer wild flies that lack
an intron excises imperfectly and leaves five base P elements. How can a fundamentally parasitic DNA
pairs behind in the intron. element like P spread so rapidly and pervasively when
23. Before the integration of a transposon, its transposase its effect is to cause deleterious mutations by inserting
makes a staggered cut in the host target DNA. If the into genes?
staggered cut is at the sites of the arrows below, draw 29. You are leading the bioinformatics team analyzing
what the sequence of the host DNA will be after the a newly discovered species of nematode related to
transposon has been inserted. Represent the trans- C. elegans . You have its complete genome DNA
poson as a rectangle. sequence and a database of all the RNAs expressed in
↓ this species. How would you proceed to (1) identify all
AATTTGGCCTAGTACTAATTGGTTGG class 1 and class 2 transposons, (2) distinguish between
TTAAACCGGATCATGATTAACCAACC autonomous and nonautonomous elements, and
↑ (3) determine if any of the family of elements is active?
30. The yeast genome has class 1 elements (Ty1, Ty2, and recessive allele (su) that conditions kernels that accu-
so forth) but no class 2 elements. What is a possible mulate sugar but not starch (su is the sweet corn allele).
reason why DNA elements have not been successful in We have two parent stocks—one stock is homozygous
the yeast genome? c-m1/c-m1 on chromosome 9 and homozygous Su/Su
31. In addition to Tc1, the C. elegans genome contains for starchy kernels on chromosome 4; the other stock
other families of DNA transposons such as Tc2, Tc3, is homozygous C/C and su/su, and homozygous for an
Tc4, and Tc5. Like Tc1, their transposition is repressed Ac element at an unknown location. Remember, c-m1
in the germ line but not in somatic cells. Predict the is an unstable allele containing a Ds element. You cross
behavior of these elements in the mutant strains where these two stocks and self-pollinate their F1.
Tc 1 is no longer repressed due to mutations in the a. What are the phenotypes of the two parent stocks?
RNAi pathway. Justify your answer. b. Make a table showing the genotypes and pheno-
32. Based on the mechanism of gene silencing, what fea- types and their proportions among the F2’s, given
tures of transposable elements does the RNAi pathway that the Ac is so tightly linked to su that there are
exploit to ensure that the host’s own genes are not also no crossovers between them.
silenced? c. Make the same table, given that the Ac is on a
33. What are the similarities and differences between ret- chromosome other than chromosomes 4 and 9.
roviruses and retrotransposons? It has been hypothe- 38. You are using nitrosoguanidine to “revert” mutant
sized that retroviruses evolved from retrotransposons. nic-2 (nicotinamide-requiring) alleles in Neurospora.
Do you agree with this model? Justify your answer. You treat cells, plate them on a medium without nic-
34. You have isolated a transposable element from otinamide, and look for prototrophic colonies. You
the human genome and have determined its DNA obtain the following results from two mutant alleles.
sequence. How would you use this sequence to deter- a. With nic-2 allele 1, you obtain no prototrophs at all.
mine the copy number of the element in the human
genome based on analytical methods you learned in b. With nic-2 allele 2, you obtain three prototrophic
Chapters 10 and 14? colonies, A, B, and C, and you cross each sepa-
rately with a wild-type strain of Neurospora. From
35. Devise a genetic screen to identify the molecular com- the cross prototroph A × wild type, you obtain
ponents of the mechanism that represses P element 100 progeny, all of which are prototrophic. From
transposition in Drosophila . Describe the genetic the cross prototroph B × wild type, you obtain
stocks you would use, how you would treat them, 100 progeny, of which 78 are prototrophic and 22
any crosses you would make, and the results that you are nicotinamide requiring. From the cross pro-
expect to obtain. totroph C × wild type, you obtain 1000 progeny,
36. When a typical maize solo Ds element excises, the of which 996 are prototrophic and 4 are nicotin-
two ends of the chromosome at the excision site are amide requiring. Explain these results at the molec-
normally ligated back together. Thus, when Ds is in a ular level. www
gene, the gene can revert to wild type when Ds excises. www
The first Ds element that McClintock discovered was GENETICS AND SOCIETY
unusual in that it induced a high frequency of chromo- Researchers in France have developed a technique to tran-
some breakage because the broken ends of the chro- siently mobilize transposable elements in crop plants to
mosome were not ligated back together. Suppose two rapidly generate large numbers of new mutations with the
Ds elements are close neighbors on a chromosome. expectation that some of these mutants will improve the
Diagram a mechanism by which the transposase could crop. Plant breeders would screen the mutants and identify
cause an excision event involving these neighbor ele- those that are beneficial. In the past, plant breeders have
ments that promotes the failure of the two ends of the used X-ray mutagenesis to create new mutations in crop
chromosome to be ligated back together. plants. The technique does not involve “genetic engineer-
37. McClintock mapped Ac and found that it mapped to ing,” the insertion of foreign genes into a species. How
different locations in different crosses. Let’s consider might consumers or governmental agencies respond to this
a cross that would allow us to map Ac relative to new technology? How would you guide their decision-
the Sugary (Su) locus on chromosome 4 which has a making process based on your knowledge of transposons?
Large-Scale Chromosomal
Changes
557
A
young couple is planning to have children. The hus- significant? What tests might be necessary to investigate this
band knows that his grandmother had a child with situation? The analysis of such questions is the topic of this
Down syndrome by a second marriage. Down syn- chapter.
drome is a set of physical and mental disorders caused by We have seen throughout the book that gene mutations
the presence of an extra chromosome 21 (Figure 17-1). No are an important source of change in the genomic sequence.
records of the birth, which occurred early in the twentieth However, the genome can also be remodeled on a larger
century, are available, but the couple knows of no other scale by alterations to chromosome structure or by changes
cases of Down syndrome in their families. in the number of copies of chromosomes in a cell. These
The couple has heard that Down syndrome results from large-scale variations are termed chromosome mutations
a rare chance mistake in egg production and therefore decide to distinguish them from gene mutations. Broadly speak-
that they stand only a low chance of having such a child. ing, gene mutations are defined as changes that take place
They decide to have children. Their first child is unaffected, within a gene, whereas chromosome mutations are changes
but the next conception aborts spontaneously (a miscar- in a chromosome region encompassing multiple genes.
riage), and their second child is born with Down syndrome. Gene mutations are never detectable microscopically; a
Was their having a Down syndrome child a coincidence, or chromosome bearing a gene mutation looks the same under
did a connection between the genetic makeup of the child’s the microscope as one carrying the wild-type allele. In con-
father and that of his grandmother lead to their both hav- trast, many chromosome mutations can be detected by
ing Down syndrome children? Was the spontaneous abortion microscopy, by genetic or molecular analysis, or by a com-
bination of these techniques. Chromosome mutations have
been best characterized in eukaryotes, and all the examples
Child with Down syndrome in this chapter are from that group.
Chromosome mutations are important to scientists and
clinicians for several reasons. First, they can be sources of
insight into how genes act in concert on a genomic scale.
Second, they reveal several important features of meiosis
and chromosome architecture. Third, they constitute useful
tools for experimental genomic manipulation. Fourth, they
are sources of insight into evolutionary processes. Fifth,
chromosomal mutations are regularly found in humans,
and some of these mutations cause genetic disease.
Many chromosome mutations cause abnormalities in
cell and organismal function. Most of these abnormalities
stem from changes in gene number or gene position. In
some cases, a chromosome mutation results from chromo-
some breakage. If the break occurs within a gene, the result
is functional disruption of that gene.
For our purposes, we will divide chromosome muta-
tions into two groups: changes in chromosome number
and changes in chromosome structure. These two groups
represent two fundamentally different kinds of events.
Changes in chromosome number are not associated with
structural alterations of any of the DNA molecules of the
cell. Rather, it is the number of these DNA molecules that
is changed, and this change in number is the basis of their
genetic effects. Changes in chromosome structure, on the
other hand, result in novel sequence arrangements within
one or more DNA double helices. These two types of chro-
mosome mutations are illustrated in Figure 17-2, which is a
summary of the topics of this chapter. We begin by explor-
FIGURE 17-1 Down syndrome results from having an extra copy of ing the nature and consequences of changes in chromosome
chromosome 21. [Terry Harris/Shutterstock.] number.
558
Relocation of Loss of
Deletion
genetic material genetic material
Translocation
Missing chromosome(s)
From another
chromosome Wild-type sequence
Extra chromosome(s)
Inversion
Gain of
Duplication genetic material
FIGURE 17-2 The illustration is divided into three colored regions to depict the main types of
chromosome mutations that can occur: the loss, gain, or relocation of entire chromosomes or chromosome
segments. The wild-type chromosome is shown in the center.
17.1 CHANGES IN CHROMOSOME are individual organisms that have more than two chro-
mosome sets. They can be represented by 3n (triploid), 4n
NUMBER (tetraploid), 5n (pentaploid), 6n (hexaploid), and so forth.
The number of chromosome sets is called the ploidy or
LO 17.1 Distinguish among the different types of ploidy level. The number of chromosomes in a set is called
polyploidy and predict their effects on meiosis and
subsequent progeny.
the haploid chromosome number. An individual member
of a normally diploid species that has only one chromo-
LO 17.2 Predict the outcome of first and second division
meiotic nondisjunction. some set (n) is called a monoploid to distinguish it from an
individual member of a normally haploid species (also n).
In genetics as a whole, few topics impinge on human affairs Examples of these conditions are shown in Table 17-1.
quite so directly as that of changes in the number of chromo-
somes present in our cells. Foremost is the fact that a group TABLE 17-1 Chromosome Constitutions in a
of common genetic disorders results from the presence of Normally Diploid Organism with Three
an abnormal number of chromosomes. Although this group Chromosomes (Identified as A, B, and C)
in the Basic Set*
of disorders is small, these disorders are the leading genetic
cause of miscarriage, birth defects, and developmental dis- Number of
Name Designation Constitution chromosomes
abilities in humans. Also of relevance to humans is the role
of chromosome mutations in plant breeding: plant breeders Normal Euploid
have routinely manipulated chromosome number to improve Diploid 2n AA BB CC 6
commercially important agricultural crops. Aberrant Euploids
Changes in chromosome number are of two basic types: Monoploid n ABC 3
changes in whole chromosome sets, resulting in a condition Triploid 3n AAA BBB CCC 9
called aberrant euploidy, and changes in parts of chromo-
Tetraploid 4n AAAA BBBB 12
some sets, resulting in a condition called aneuploidy. CCCC
Aneuploids
Aberrant euploidy
Monosomic 2n − 1 A BB CC 5
Organisms with multiples of the basic chromosome set
AA B CC 5
(genome) are referred to as euploid. You learned in earlier
AA BB C 5
chapters that familiar eukaryotes such as plants, animals,
and fungi carry in their cells either one chromosome set Trisomic 2n + 1 AAA BB CC 7
(haploidy) or two chromosome sets (diploidy). In these spe- AA BBB CC 7
cies, both the haploid and the diploid states are cases of AA BB CCC 7
normal euploidy. Organisms that have more or fewer than *In the case shown, the number of chromosomes in the basic set (the
the normal number of sets are aberrant euploids. Polyploids haploid chromosome number) is three.
Monoploids Male bees, wasps, and ants are monoploid. Higher ploidy produces larger size
In the normal life cycles of these insects, males develop by
parthenogenesis (the development of a specialized type of
unfertilized egg into an embryo without the need for fertil-
ization). In most other species, however, monoploid zygotes
fail to develop. The reason is that virtually all members
of a diploid species carry a number of deleterious reces-
sive mutations, together called a genetic load. The delete-
rious recessive alleles are masked by wild-type alleles in
the diploid condition, but the effects of these alleles can be
observed in a monoploid derived from a diploid. Mono-
ploids that do develop to advanced stages are abnormal. If
they survive to adulthood, their germ cells cannot proceed
through meiosis normally because the chromosomes have
no pairing partners. Thus, monoploids are characteristi-
cally sterile. (Male bees, wasps, and ants bypass meiosis; in
these groups, gametes are produced by mitosis.)
Pairing possibilities
or
Aneuploid gametes do not generally give rise to viable fibers in cells undergoing division. The resulting polyploid
offspring. In plants, aneuploid pollen grains are generally tissue (such as a polyploid branch of a plant) can be
inviable and hence unable to fertilize the female gamete. In detected by examining stained chromosomes from the tis-
any organism, zygotes that might arise from the fusion of a sue under a microscope. Such a branch can be removed and
haploid and an aneuploid gamete will themselves be aneu- used as a cutting to generate a polyploid plant or allowed to
ploid, and typically these zygotes also are inviable. We will produce flowers, which, when selfed, would produce poly-
examine the underlying reason for the inviability of aneu- ploid offspring. A commonly used antitubulin agent is col-
ploids when we consider gene balance later in the chapter. chicine, an alkaloid extracted from the autumn crocus. In
colchicine-treated cells, the S phase of the cell cycle takes
KEY CONCEPT Polyploids with odd numbers of chromo- place, but chromosome segregation or cell division does
some sets, such as triploids, are sterile or highly infertile not. As the treated cell enters telophase, a nuclear mem-
because their gametes and offspring are aneuploid. brane forms around the entire doubled set of chromosomes.
Thus, treating diploid (2n) cells with colchicine for one cell
Autotetraploids arise by the doubling of a 2n comple- cycle leads to tetraploids (4n) with exactly four copies of
ment to 4n. This doubling can occur spontaneously, but it each type of chromosome (Figure 17-5). Note that all alleles
can also be induced artificially by applying chemical agents in the genotype are doubled. Therefore, if a diploid cell of
that disrupt microtubule polymerization. As stated in genotype A/a ; B/b is doubled, the resulting autotetraploid
Chapter 2, chromosome segregation is powered by spindle will be of genotype A/A/a/a ; B/B/b/b. Treatment for an
fibers, which are polymers of the protein tubulin. Hence, additional cell cycle produces octoploids (8n), and so forth.
disruption of microtubule polymerization blocks chro- This method works in both plant and animal cells, but, gen-
mosome segregation. The chemical treatment is normally erally, plants seem to be much more tolerant of polyploidy.
applied to somatic tissue during the formation of spindle Many natural autotetraploid plants, such as potatoes,
A B b a A B b a
With
colc
hicin One tetraploid cell
e
4n = 8
A B b a
Genotype = A/A/a/a; B/B/b/b
FIGURE 17-5 Colchicine may be applied to generate a tetraploid sister chromatids after the centromere has split. A single cell is created
from a diploid. Colchicine added to mitotic cells during metaphase and that contains pairs of identical chromosomes that are homozygous at all
anaphase disrupts spindle-fiber formation, preventing the migration of loci.
Tetraploid watermelon parts are larger each set pair and segregate. There are several possibilities, as
than diploid ones shown in Figure 17-7. If the chromosomes pair as bivalents
or quadrivalents, the chromosomes segregate normally, pro-
ducing diploid gametes. The fusion of gametes at fertilization
regenerates the tetraploid state. If trivalents form, segregation
leads to nonfunctional aneuploid gametes and, hence, sterility.
Pairing
possibilities
FIGURE 17-7 There are three different pairing possibilities at meiosis I in tetraploids. The four homo- A N I M ATED A RT
logous chromosomes may pair as two bivalents or as a quadrivalent, and each can yield functional
gametes. A third possibility, a trivalent plus a univalent, yields nonfunctional gametes. Segregation outcomes of autotetra-
ploid meiosis
chromosome, and functional gametes of the type n1 + n2 are Origin of the amphidiploid Raphanobrassica
produced. These gametes fuse to give 2n1 + 2n2 allopolyploid
n=9
progeny, which also are fertile. This kind of allopolyploid
is sometimes called an amphidiploid, or doubled diploid Gametes
(Figure 17-8). Treating a sterile hybrid with colchicine greatly
n=9
s
increases the chances that the chromosome sets will double.
g ou
lin ne
Amphidiploids are now synthesized routinely in this manner.
ub ta
do pon
When Karpechenko’s allopolyploid was crossed with Raphanus
S
either parental species—the cabbage or the radish—sterile 2n = 18
offspring resulted. The offspring of the cross with cabbage
Parents ×
were 2n1 + n2 , constituted from an n1 + n2 gamete from
Sterile F1 hybrid
the allopolyploid and an n1 gamete from the cabbage. The Brassica n+n=9+9 Raphanobrassica
n2 chromosomes had no pairing partners; hence, a normal 2n = 18 2n = 18
meiosis could not take place, and the offspring were sterile. Fertile amphidiploid
2n + 2n = 18 + 18
Thus, Karpechenko had effectively created a new species, 4n = 36
with no possibility of gene exchange with either cabbage or
radish. He called his new plant Raphanobrassica.
In nature, allopolyploidy seems to have been a major
force in the evolution of new plant species. One convinc- FIGURE 17-8 In the progeny of a cross of cabbage (Brassica) and
ing example is shown by the genus Brassica, as illustrated radish (Raphanus), the fertile amphidiploid arose from spontaneous
in Figure 17-9. Here, three different parent species have doubling in the 2n = 18 sterile hybrid. The sketches depict the seedpods
of each parental species and the hybrids.
B. oleracea,
2n = 18
Cabbage
Cauliflower
Broccoli
Kale
Kohlrabi
Brussels sprouts
n=9 n=9
B. carinata, B. napus,
2n = 34 2n = 38
Abyssinian mustard Rutabaga
Oil rape
n=8 n = 10
FIGURE 17-9 Allopolyploidy is important in the production of new species. In the example shown, three diploid species of
Brassica (light green boxes) were crossed in different combinations to produce their allopolyploids (tan boxes). Some of the
agricultural derivatives of some of the species are shown within the boxes.
hybridized in all possible pair combinations to form new An aneuploid is an individual organism whose chromo-
amphidiploid species. Natural polyploidy was once viewed some number differs from the wild type by part of a chro-
as a somewhat rare occurrence, but recent work has shown mosome set. Generally, the aneuploid chromosome set
that it is a recurrent event in many plant species. The use differs from the wild type by only one chromosome or by
of DNA markers has made it possible to show that poly- a small number of chromosomes. An aneuploid can have a
ploids in any population or area that appear to be the same chromosome number either greater or smaller than that of
are the result of many independent past fusions between the wild type. Aneuploid nomenclature (see Table 17-1) is
genetically distinct individuals of the same two parental based on the number of copies of the specific chromosome
species. An estimated 50 percent of all angiosperm plants in the aneuploid state. For autosomes in diploid organisms,
are polyploids, resulting from auto- or allopolyploidy. As the aneuploid 2n + 1 is trisomic, 2n − 1 is monosomic, and
a result of multiple polyploidizations, the amount of allelic 2n − 2 (the “−2” represents the loss of both homologs of a
variation within a polyploid species is much higher than chromosome) is nullisomic. In haploids, n + 1 is disomic.
formerly thought, perhaps contributing to its potential for Special notation is used to describe sex-chromosome aneu-
adaptation. ploids because it must deal with the two different chro-
Allopolyploidy has also been important in the produc- mosomes. The notation merely lists the copies of each sex
tion of modern crop plants. In addition to the examples of chromosome, such as XXY, XYY, XXX, or XO (the “O”
Brassica species, New World cotton is a natural allopoly- stands for absence of a chromosome and is included to
ploid that arose spontaneously, as is wheat. show that the single X symbol is not a typographical error).
Allopolyploid plant cells can also be produced artifi-
cially by fusing diploid cells from different species. First, Nondisjunction The cause of most aneuploidy is
the walls of two diploid cells are removed by treatment ondisjunction in the course of meiosis or mitosis. Dis-
n
with an enzyme, and the membranes of the two cells fuse junction is another word for the normal segregation of
and become one. The nuclei often fuse, too, resulting in the homologous chromosomes or chromatids to opposite poles
polyploid. If the cell is nurtured with the appropriate hor- at meiotic or mitotic divisions. Nondisjunction is a failure
mones and nutrients, it divides to become a small allopoly- of this process, in which two chromosomes or chromatids
ploid plantlet, which can then be transferred to soil. incorrectly go to one pole and none to the other.
Mitotic nondisjunction can occur as cells divide during
KEY CONCEPT Allopolyploid plants can arise in nature or development. Sections of the body will be aneuploid
be synthesized by crossing related species and doubling the (aneuploid sectors) as a result. Meiotic nondisjunction is
chromosomes of the hybrid or by fusing diploid cells. Allopoly- more commonly encountered. In this case, the products
ploidy has played an important role in the formation of many
of meiosis are aneuploid, leading to descendants in which
natural and agricultural plant species.
the entire organism is aneuploid. In meiotic nondisjunc-
tion, the chromosomes may fail to disjoin at either the first
or the second meiotic division ( Figure 17-10). Either way,
Polyploid animals As noted earlier, polyploidy is more
n −1 and n + 1 gametes are produced. If an n −1 gamete is
common in plants than in animals, but there are cases of
fertilized by an n gamete, a monosomic (2n − 1) zygote is
naturally occurring polyploid animals. Polyploid species
produced. The fusion of an n + 1 and an n gamete yields a
of flatworms, leeches, and brine shrimps reproduce by par-
trisomic 2n + 1.
thenogenesis. Triploid and tetraploid Drosophila have been
synthesized experimentally. Naturally occurring polyploid
amphibians and reptiles are surprisingly common. They have KEY CONCEPT Aneuploid organisms result mainly from
nondisjunction in a parental meiosis.
several modes of reproduction: polyploid species of frogs and
toads participate in sexual reproduction, whereas polyploid
Nondisjunction occurs spontaneously. Like most
salamanders and lizards are parthenogenetic. The Salmoni-
gene mutations, it is an example of a chance failure of
dae (the family of fishes that includes salmon and trout) pro-
a basic cellular process. The precise molecular processes
vide a familiar example of the numerous animal species that
that fail are not known, but in experimental systems, the
appear to have originated through ancestral polyploidy.
frequency of nondisjunction can be increased by interfer-
The sterility of triploids has been commercially exploited
ence with microtubule polymerization, thereby inhibit-
in animals as well as in plants. Triploid oysters have been
ing normal chromosome movement. Disjunction appears
developed because they have a commercial advantage over
to be more likely to go awry in meiosis I. This failure
their diploid relatives. The diploids go through a spawning
is not surprising, because normal anaphase I disjunction
season, when they are unpalatable, but the sterile triploids
requires that the homologous chromatids of the tetrad
do not spawn and are palatable year-round.
remain paired during prophase I and metaphase I, and
it requires crossovers. In contrast, proper disjunction at
Aneuploidy anaphase II or at mitosis requires that the centromere
Aneuploidy is the second major category of chromosomal split properly but does not require chromosome pairing
aberrations in which the chromosome number is abnormal. or crossing over.
A N I M ATED A RT
n+1
Nondisjunctional meiosis I
n–1 A N I M ATED A RT
Nondisjunctional meiosis II
n–1
First Nondisjunction
division at second
division n+1
n–1
Crossovers are a necessary component of the normal Monosomics (2n − 1) Monosomics are missing one copy
disjunction process. Somehow the formation of a chi- of a chromosome. In most diploid organisms, the absence of
asma helps to hold a bivalent together and ensures that one chromosome copy from a pair is deleterious. In humans,
the two dyads will go to opposite poles. In most organ- monosomics for any of the autosomes die in utero. Many
isms, the amount of crossing over is sufficient to ensure X-chromosome monosomics also die in utero, but some
that all bivalents will have at least one chiasma per meio- are viable. A human chromosome complement of 44 auto-
sis. In Drosophila, many of the nondisjunctional chromo- somes plus a single X produces a condition known as Turner
somes seen in disomic (n + 1) gametes are nonrecombinant, syndrome, represented as XO. Affected persons have a char-
showing that they arise from meioses in which there is no acteristic phenotype: they are sterile females, short in stature,
crossing over on that chromosome. Similar observations and often have a web of skin extending between the neck and
have been made in human trisomies. In addition, in several shoulders (Figure 17-11). Although their intelligence is near
different experimental organisms, mutations that interfere normal, some of their specific cognitive functions are defec-
with recombination have the effect of massively increasing tive. About 1 in 5000 female births show Turner syndrome.
the frequency of meiosis I nondisjunction. All these obser- Geneticists have used viable plant monosomics to map
vations provide evidence for the role of crossing over in newly discovered recessive mutant alleles to a specific chro-
maintaining chromosome pairing; in the absence of these mosome. For example, one can make a set of monosomic
associations, chromosomes are vulnerable to anaphase I lines, each known to lack a different chromosome. Homo-
nondisjunction. zygotes for the new mutant allele are crossed with each
monosomic line, and the progeny of each cross are inspected
for the recessive phenotype. The appearance of the recessive
KEY CONCEPT Crossovers are needed to keep bivalents phenotype identifies the chromosome that has one copy miss-
paired until anaphase I. If crossing over fails for some reason,
ing as the one on which the gene is normally located. The test
first-division nondisjunction occurs.
works because half the gametes of a fertile 2n − 1 monosomic
Short stature
Characteristic
facial features
Low hairline
Fold of skin
Constriction
Shield-shaped of aorta
thorax Poor breast
Widely spaced development
nipples
Elbow
Shortened deformity
metacarpal IV
Small Rudimentary
fingernails ovaries
Gonadal streak
(underdeveloped
gonadal
structures)
Brown spots (nevi)
No menstruation
FIGURE 17-11 Turner syndrome results from the presence of a single X chromosome (XO), which
can be seen in the image of chromosomes from an individual with Turner syndrome. [Zuzanae/
Shutterstock.com.]
will be n −1, and, when an n −1 gamete is fertilized by a As an illustration, let’s assume that a gene A with a mutant
gamete bearing a new mutation on the homologous chromo- allele a is on chromosome 2. Crosses of a/a and monosomics
some, the mutant allele will be the only allele of that gene for chromosome 1 and chromosome 2 are predicted to produce
present and hence its phenotype can be observed. different results (chromosome is abbreviated chr):
a A
chr 1/chr 1 chr 2/chr 2 × chr 1/O chr 2/chr 2
a A
chr 2 is mutant chr 1 is monosomic
A A
chr 1/chr 1 chr 2/chr 2 chr 1/O chr 2/chr 2
a a
100% wild-type (A/a) progeny
a A
chr 1/chr 1 chr 2/chr 2 × chr 1/chr 1 chr 2/O
a
chr 2 is mutant chr 2 is monosomic
a A
chr 1/chr 1 chr 2/O chr 1/chr 1 chr 2/chr 2
a
50% mutant (a/O) progeny 50% wild-type (A/a) progeny
Osteoporosis
Female-type
pubic hair Small testes
pattern
FIGURE 17-13 Klinefelter syndrome results from the presence of two X chromosomes and a
Y chromosome, which can be seen in the image of chromosomes from an individual with Klinefelter
syndrome. [Zuzanae/Shutterstock.com.]
Y does not pair and is not transmitted to gametes. Therefore, Of human trisomies, the most familiar type is Down
the gametes contain either X or Y, never YY or XY. Triplo- syndrome (Figure 17-14), discussed briefly at the beginning
X trisomics (XXX) are phenotypically normal and fertile of the chapter. The frequency of Down syndrome is about
females. Meiosis shows pairing of only two X chromosomes; 0.15 percent of all live births. Most affected persons have
the third does not pair. Hence, eggs bear only one X and, like an extra copy of chromosome 21 caused by nondisjunc-
that of XYY males, the condition is not passed on to progeny. tion of chromosome 21 in a parent who is chromosomally
normal. In this sporadic type of Down syndrome, there is common phenomenon in many animals. In female humans,
no family history of aneuploidy. Some rarer types of Down all oocytes are arrested at diplotene before birth. Meiosis
syndrome arise from translocations (a type of chromosomal resumes at each menstrual period, which means that the
rearrangement discussed later in the chapter); in these chromosomes in the bivalent must remain properly asso-
cases, as we will see, Down syndrome recurs in the pedigree ciated for as long as five or more decades. If we speculate
because the translocation may be transmitted from parent that these associations have an increasing probability of
to child. breaking down by accident as time passes, we can envision
The combined phenotypes that make up Down syn- a mechanism contributing to increased maternal nondis-
drome include mental retardation (with an IQ in the 20 to junction with age. Consistent with this speculation, most
50 range); a broad, flat face; eyes with an epicanthic fold nondisjunction related to the effect of maternal age is due
(a skin fold of the upper eyelid that covers the inner corner to nondisjunction at anaphase I, not anaphase II. However,
of the eye); short stature; short hands with a crease across recent research suggests that additional factors contribute
the middle; and a large, wrinkled tongue. Females may be to the prevalence of aneuploidy in older mothers, and iden-
fertile and may produce normal or trisomic progeny, but tifying those factors as well as preventative interventions
males are sterile with very few exceptions. Due to improve- are active areas of research.
ments in the care of individuals with Down syndrome, the The only other human autosomal trisomics to survive to
average life expectancy is now 60 years. birth are those with trisomy 13 (Patau syndrome) and tri-
The incidence of Down syndrome is related to mater- somy 18 (Edwards syndrome). Both have severe physical and
nal age: older mothers run a greatly elevated risk of hav- mental abnormalities. The phenotypic syndrome of trisomy
ing a child with Down syndrome (Figure 17-15). For this 13 includes a harelip, a small and malformed head, “rocker-
reason, fetal chromosome analysis (by cell-free fetal DNA bottom” feet, and a mean life expectancy of 130 days.
testing, amniocentesis, or chorionic villus sampling) is That of trisomy 18 includes “faunlike” ears, a small jaw,
now recommended for older expectant mothers. Although a narrow pelvis, and rocker-bottom feet; almost all babies
paternal age has not been linked to Down syndrome, a with trisomy 18 die within the first few weeks after birth.
paternal-age effect on point mutations also has been found All other trisomics die in utero.
(see Chapter 1).
Even though the maternal-age effect has been known
for many years, the causes are still not known. Nonethe- The concept of gene balance
less, there are some interesting biological correlations. In considering aberrant euploidy, we noted that an increase
With age, possibly the chromosome bivalent is less likely in the number of full chromosome sets correlates with
to stay together during prophase I of meiosis. Meiotic increased organism size but that the general shape and pro-
arrest of oocytes (female meiocytes) in late prophase I is a portions of the organism remain very much the same. In
contrast, autosomal aneuploidy typically alters the organ-
ism’s shape and proportions in characteristic ways.
The maternal-age effect in Down syndrome Plants tend to be somewhat more tolerant of aneuploidy
than are animals. Studies in jimsonweed (Datura stramonium)
Incidence of Down syndrome per number of births
FIGURE 17-16 Each of the 12 possible trisomics of Datura is show the fruit of a normal Datura (2n) or different trisomic Datura (2n + 1),
disproportionate in a different way. (a) Datura fruit. (b) The drawings each of which has been named. [(a) Konrad Lange/Getty Images.]
are much less abnormal than are monosomics for chro- product is made (Figure 17-17). This relation between the
mosome 4. In humans, no autosomal monosomic survives number of copies of a gene and the amount of the gene’s
to birth, but, as already stated, three types of autosomal product made is called a gene-dosage effect.
trisomics can do so. As is true of aneuploid jimsonweed, We can infer that normal physiology in a cell depends
each of these three trisomics shows unique phenotypic on the proper ratio of gene products in the euploid cell.
syndromes. This ratio is the normal gene balance. If the relative dos-
Why does aneuploidy for each chromosome have its age of certain genes changes—for example, because of the
own characteristic phenotypic effects? Why are mono- removal of one of the two copies of a chromosome (or
somics typically more severely affected than are the corre- even a segment thereof)—physiological imbalances in cel-
sponding trisomics? And why are aneuploids generally so lular pathways can arise. These imbalances are the reason
much more abnormal than polyploids? The answers seem that aneuploids are generally much more abnormal than
certain to be a matter of gene balance. In a euploid, the polyploids.
ratio of genes on any one chromosome to the genes on In some cases, the imbalances of aneuploidy result
other chromosomes is always 1:1, regardless of whether from the effects of a few “major” genes whose dosage has
we are considering a monoploid, diploid, triploid, or tetra- changed, rather than from changes in the dosage of all
ploid. For example, in a tetraploid, for gene A on chromo- the genes on a chromosome. Such genes can be viewed as
some 1 and gene B on chromosome 2, the ratio is 4 A:4 B, haplo-abnormal (resulting in an abnormal phenotype if
or 1:1. In contrast, in an aneuploid, the ratio of genes on present only in one copy) or triplo-abnormal (resulting in
the aneuploid chromosome to genes on the other chromo- an abnormal phenotype if present in three copies). They
somes differs from the wild type up or down by 50 percent: contribute significantly to the aneuploid phenotypic syn-
50 percent for monosomics and 150 percent for trisomics. dromes. For example, the study of persons trisomic for only
For example, in a trisomic for chromosome 2 with gene A part of chromosome 21 has made it possible to localize
on chromosome 1 and gene B on chromosome 2, we find genes contributing to Down syndrome to various regions
that the ratio of the A and B genes is 2 A:3 B. Thus, the of chromosome 21; the results hint that some aspects of
aneuploid genes are out of balance. How does their being the phenotype might be due to triplo-abnormality for a
out of balance help us answer the questions raised at the specific gene in these chromosome regions. However, other
opening of this paragraph? aspects of aneuploid syndromes are likely to result from the
In general, the amount of transcript produced by a gene cumulative effects of aneuploidy for numerous genes whose
is directly proportional to the number of copies of that gene products are all out of balance. This is supported by the fact
in a cell. That is, for a given gene, the rate of transcription is that the only trisomy that survives to adulthood in humans
directly related to the number of DNA templates available. is for chromosome 21, which is the smallest human chro-
Thus, the more copies of the gene, the more transcripts mosome. Trisomies of larger chromosomes with even more
are produced and the more of the corresponding protein genes out of balance do not survive. Thus, the characteristic
Aneuploidy affects the balance of gene dosage in a cell phenotypic effects associated with aneuploidy for each
different chromosome, as seen in jimsonweed or humans,
(a) Diploid, 2n likely result from a combination of the imbalance of genes
Gene A Gene B with very specific phenotypic effects and the total number
Chr 1 Chr 2 of genes that are out of balance.
However, the concept of gene balance does not tell
us why having too few gene products (monosomy) is
much worse for an organism than having too many gene
products (trisomy). In a parallel manner, we can ask why
there are many more haplo-abnormal genes than triplo-
Gene A Gene B abnormal ones. A key to explaining the extreme abnor-
Chr 1 Chr 2
mality of monosomics relative to trisomics is that the phe-
notypes of any deleterious recessive alleles present on a
monosomic autosome will be automatically observed due
to the absence of the wild-type allele.
A:B = 1:1
Sex chromosome gene balance How do we apply the
idea of gene balance to cases of sex-chromosome aneu-
(b) Aneuploid, 2n+1 ploidy? Gene balance holds for sex chromosomes as well,
Gene A Gene B but we also have to take into account the special proper-
Chr 1 Chr 2 ties of the sex chromosomes. In many organisms with XY
sex determination, such as mammals, the Y chromosome
seems to be a degenerate version of the X chromosome in
which there are very few functional genes other than some
concerned with sex determination itself, sperm production,
or both (Figure 17-18). The X chromosome, on the other
Gene A Gene B
Chr 1 Chr 2 hand, contains many genes concerned with basic cellu-
lar processes (“housekeeping genes”) that just happen to
reside on the chromosome that eventually evolved into the
X chromosome. XY sex-determination mechanisms have
likely evolved independently at least a hundred times in dif-
ferent taxonomic groups. For example, there appears to be
Gene B one sex-determination mechanism for all mammals, but it
Chr 2 is completely different from the mechanism governing sex
determination in fruit flies (see Chapter 13).
In a sense, X chromosomes are naturally aneuploid.
In species with an XY sex-determination system, females
have two X chromosomes, whereas males have only one.
A:B = 2:3
Nonetheless, the X chromosome’s housekeeping genes
are expressed to approximately equal extents per cell
(c) Aneuploid, 2n–1 in females and in males. In other words, there is dosage
c ompe nsat ion. How is this compensation accom-
Gene A Gene B
Chr 1 Chr 2 plished? The answer depends very much on the organ-
ism. In fruit flies, the male’s X chromosome appears to be
hyperactivated, allowing it to be transcribed at twice the
rate of either X chromosome in the female. As a result, the out that a few genes scattered throughout an “inactive X”
XY male Drosophila has an X gene dosage equivalent to that are still transcriptionally active. In XXY males, these genes
of an XX female. In mammals, in contrast, the rule is that no are transcribed at twice the level that they are in XY males.
matter how many X chromosomes are present, there is only In XXX females, on the other hand, the few transcribed
one transcriptionally active X chromosome in each somatic genes are active at only 1.5 times the level that they are in
cell. This rule gives the XX female mammal an X gene dos- XX females. This lower level of “functional aneuploidy”
age equivalent to that of an XY male. Dosage compensation in XXX than in XXY, plus the fact that the active X genes
in mammals is achieved by X-chromosome inactivation. A appear to lead to feminization, may explain the feminized
female with two X chromosomes, for example, is a mosaic phenotype of XXY males. The severity of Turner syndrome
of two cell types in which one or the other X chromosome (XO) may be due to the deleterious effects of monosomy
is active. We examined this phenomenon in Chapter 12. and to the lower activity of the transcribed genes of the
Thus, XY and XX individuals produce the same amounts of X (compared with XX) females. As is usually observed
X-chromosome housekeeping-gene products. X-chromosome for aneuploids, monosomy for the X chromosome pro-
inactivation also explains why triplo-X humans are pheno- duces a more abnormal phenotype than does having an
typically normal: only one of the three X chromosomes is extra copy of the same chromosome (triplo-X females or
transcriptionally active in a given cell. Similarly, an XXY XXY males).
male is only moderately affected because only one of his two Gene dosage is also important in the phenotypes of
X chromosomes is active in each cell. polyploids. Human polyploid zygotes do arise through var-
Why are XXY individuals abnormal at all, given that ious kinds of mistakes in cell division. Most die in utero.
triplo-X individuals are phenotypically normal? It turns Occasionally, triploid babies are born, but none survive.
This fact seems to violate the principle that polyploids are 4. Repair systems in the cell correct the double-stranded
more normal than aneuploids. The explanation for this breaks by joining broken ends back together (see
contradiction seems to lie with X-chromosome dosage Chapter 15 for a detailed discussion of DNA repair).
compensation. Part of the rule for gene balance in organ- 5. If the two ends of the same break are rejoined, the
isms that have a single active X seems to be that there original DNA order is restored. If the ends of two
must be one active X for every two copies of the autoso- different breaks are joined, however, one result is some type
mal chromosome complement. Thus, some cells in triploid of chromosomal rearrangement (see Figure 17-19, left side).
mammals are found to have one active X, whereas others,
surprisingly, have two. Neither situation is in balance with Another important cause of rearrangements is crossing
autosomal genes. over between repetitive (duplicated) DNA segments. This type
of unequal crossing over is termed nonallelic homologous
KEY CONCEPT Aneuploidy is nearly always deleterious recombination (NAHR). In organisms with repeated DNA
because of gene imbalance: the ratio of gene products is dif- sequences within one chromosome or on different chromo-
ferent from that in euploids, and this difference interferes with somes, there is ambiguity about which of the repeats will pair
the normal function of the genome. with each other at meiosis. If sequences pair up that are not
in the same relative positions on the homologs, crossing over
can produce aberrant chromosomes. Deletions, duplications,
inversions, and translocations can all be produced by such
crossing over (see Figure 17-19, right side).
17.2 CHANGES IN CHROMOSOME Regardless of the mechanism of formation, the only
STRUCTURE chromosomal rearrangements that survive meiosis are those
that produce DNA molecules that have one centromere and
LO 17.3 Distinguish among the major types of chromosome two telomeres. If a rearrangement produces a chromosome
rearrangements (translocations, inversions, that lacks a centromere, such an acentric chromosome
deletions, duplications) and diagnose their will not be dragged to either pole at anaphase of mitosis
presence in progeny analysis.
or meiosis and will not be incorporated into either progeny
LO 17.4 In a cross involving a known chromosome
nucleus. Therefore, acentric chromosomes are not inherited.
rearrangement, predict the inheritance of genes
linked and unlinked to the rearrangement. If a rearrangement produces a chromosome with two cen-
tromeres (a dicentric chromosome), it will often be pulled
Changes in chromosome structure, called rearrangements, simultaneously to opposite poles at anaphase, forming an
encompass several major classes of events (see Figure 17-2). anaphase bridge. Anaphase-bridge chromosomes typically
A chromosome segment can be lost, resulting in a deletion, will not be incorporated into either progeny cell. If a chro-
or doubled, to form a duplication. The orientation of a mosome break produces a chromosome lacking a telomere,
segment within the chromosome can be reversed, called an DNA will be progressively lost from the end of the chromo-
inversion, or a segment can be moved to a different chro- some with every round of replication. Recall from Chapter 7
mosome, constituting a translocation. These rearrange- that this is because telomeres are needed to prime proper
ments can be caused by either DNA breakage or crossing DNA replication at the ends (see Figure 7-24).
over between repetitive DNA (Figure 17-19). There are two general types of rearrangements: unbal-
DNA breakage is a major cause of each of these chro- anced and balanced. Unbalanced rearrangements change
mosomal rearrangements. Both DNA strands must break the gene dosage of a chromosome segment. As with aneu-
at two different locations, followed by a rejoining of the ploidy for whole chromosomes, the loss of one copy of a
broken ends to produce a new chromosomal arrangement segment or the addition of an extra copy can disrupt nor-
(Figure 17-19, left side). Chromosomal rearrangements mal gene balance. The larger the segment that is lost or
by breakage can be induced artificially by using ionizing duplicated, the more likely it is that gene imbalance will
radiation. This kind of radiation, particularly X rays and cause phenotypic abnormalities. The two simple classes of
gamma rays, is highly energetic and causes numerous unbalanced rearrangements are deletions and duplications.
double-stranded breaks in DNA. To understand how chro- A deletion is the loss of a segment within one chromosome
mosomal rearrangements are produced by breakage, sev- arm and the juxtaposition of the two segments on either
eral points should be kept in mind: side of the deleted segment, as in this example, which shows
loss of segment C–D:
1. Each chromosome is a single double-stranded DNA
molecule. A B C D E
2. The first event in the production of a chromosomal
rearrangement is the generation of two or more dou-
ble-stranded breaks in the chromosomes of a cell (see A B E
Figure 17-19, top row at left).
3. Double-stranded breaks are potentially lethal, unless A duplication is the repetition of a segment of a chro-
they are repaired. mosome arm. In the simplest type of duplication, the two
1 2 3 4
1 2 3 4 1 4
Deletion 3 2 Loss
3 2 4 1 4
Loss 1
1 2 3 4 1 2 4 1 2 3 4 1 4
Deletion
and
duplication 1 2 3 4 1 2 3 3 4 1 2 3 4 1 2 3 2 3 4
1 2 3 4
1 2 3 4 1 3 2 4
Inversion 2 3
1 3 2 4
1 4
1 2 3 4 1 2 8 9 10 1 2 3 4 1 2 8 9 10
Translocation
5 6 7 8 9 10 5 6 7 3 4 5 6 7 8 9 10 5 6 7 3 4
FIGURE 17-19 Each of the four types of chromosomal rearrangements can be produced by either of
two basic mechanisms: chromosome breakage and rejoining or crossing over between repetitive DNA.
Chromosome regions are numbered 1 through 10. Homologous chromosomes are the same color.
A B C D E E F G
A B C C D E A F G
Deletions a b c d e f g
A deletion is simply the loss of a part of one chromosome arm.
The process of deletion requires two chromosome breaks to
cut out the intervening segment. The deleted fragment has no + + + + + + +
centromere; consequently, it cannot be pulled to a spindle pole Phenotype
in cell division and is lost. The effects of deletions depend on +++++++
Deletion
their size. A small deletion within a gene, called an intragenic
deletion, inactivates the gene and has the same effect as that
of other null mutations of that gene. If the homozygous null
a b c d e f g
phenotype is viable (as, for example, in human albinism), the
homozygous deletion also will be viable.
+ + + + +
For most of this section, we will be dealing with
Phenotype
multigenic deletions, in which several to many genes are
+ b c ++++
missing. The consequences of these deletions are more severe
than those of intragenic deletions. If such a deletion is made
homozygous by inbreeding (that is, if both homologs have
If there is no dele-
the same deletion), the combination is nearly always lethal. Deletion loops
tion, none of the pheno-
This fact suggests that all regions of the chromosomes are
types associated with the
essential for normal viability and that complete elimination (a) Meiotic chromosomes
seven recessive alleles are
of any segment from the genome is deleterious. Even an indi-
expected to be observed; C D
vidual organism heterozygous for a multigenic d eletion—
however, if the pheno- E F
that is, having one normal homolog and one that carries the
types associated with B
deletion—may not survive. Principally, this lethal outcome is A
b and c are observed, F
due to disruption of normal gene balance. Alternatively, the E
then a deletion span- B
deletion may “uncover” deleterious recessive alleles, allowing
ning the b+ and c+ genes A
the effects of the mutant allele to be observed.
has probably occurred
KEY CONCEPT The lethality of large heterozygous deletions on the other homo- (b) Polytene chromosomes
can be explained by gene imbalance and the expression of log. Because recessive
Loop
deleterious recessive alleles. alleles seem to be show-
ing dominance in such
Small deletions are sometimes viable in combination cases, the effect is called
with a normal homolog. Such deletions may be identified pseudodominance.
by examining meiotic chromosomes under the microscope. In the reverse case—
The failure of the corresponding segment on the normal if we already know the
homolog to pair creates a visible deletion loop in meiotic location of the deletion—
chromosomes (Figure 17-20a). In Drosophila, deletion loops we can apply the pseudo
FIGURE 17-20 (a) Schematic
are also visible in the polytene chromosomes. These chro- dominance effect in the of meiotic pairing in a deletion
mosomes are found in the cells of salivary glands and other opposite direction to map heterozygote. The normal homolog
specific tissues of certain insects. In these cells, the homo- the positions of mutant forms a loop because the genes in
logs pair and replicate many times without separating, and alleles. This procedure, this loop have no alleles with which to
so each homologous chromosome pair is represented by called deletion mapping, synapse. (b) A deletion loop is visible
in Drosophila polytene chromosomes
a thick bundle of replicates. These polytene chromosomes pairs mutations against where the normal chromosome
provide a rare opportunity to visualize interphase chromo- a set of defined overlap- homolog is unable to align with the
somes under light microscopy. Each chromosome has a set ping deletions. An exam- deletion chromosome homolog.
of dark-staining bands of fixed position and number. These ple from Drosophila is [William M. Gelbart, Harvard
bands act as useful chromosomal landmarks. An example shown in Figure 17-21. In University.]
of a polytene chromosome in which one original homolog this diagram, the recom-
carried a deletion is shown in Figure 17-20b. A deletion can bination map is shown at the top, marked with distances in
be assigned to a specific chromosome location by examin- map units from the left end. The location of a specific dele-
ing polytene chromosomes microscopically and determin- tion is identified by the presence of deletion loops in polytene
ing the position of the deletion loop. chromosomes, as described earlier. The horizontal bars below
Another clue to the presence of a deletion is that the loss the chromosome show the extent of the deletions listed at
of a segment on one homolog sometimes unmasks recessive the left. Each deletion is paired with each mutation under
alleles present on the other homolog, leading to the appear- test, and the phenotype is examined to see if the mutation
ance of phenotypes associated with those mutations. Consider, is pseudodominant. The mutation pn (prune), for example,
for example, the deletion shown in the following diagram: shows pseudodominance only with deletion 264-38, and
Mapping mutant alleles by pseudodominance are microencephaly (abnormally small head) and a moon-
like face. Like syndromes caused by other deletions, cri
Linkage map (m.u.) 0.8 1.5 1.7 3.0 du chat syndrome includes mental retardation. Fatality
pn w rst fa rates are low, and many persons with this deletion reach
adulthood.
Small segment
Another instructive example is Williams syndrome.
of banded polytene This syndrome is autosomal dominant and is character-
chromosome ized by unusual development of the nervous system and
certain external features. Williams syndrome is found at a
frequency of about 1 in 10,000 people. Patients often have
1 2 3 1 2 3 4 5 6 1 2 3 4 5 6 7 8 910 1 2 3 4 1 2 3 4 5 6 7 8 9 101112
pronounced musical or singing ability, as well as hyper-
2E 2F 3A 3B 3C sociality. The syndrome is almost always caused by a 1.5-Mb
258-11 deletion on one homolog of chromosome 7, specifically at
258-14 band 7q11.23. Sequence analysis has shown that this dele-
N-8Mohr
tion encompasses between 26 and 28 genes of known and
264-38
unknown function. It is likely that haploinsufficiency of dif-
264-36
Deletions
264-30
ferent genes within the deletion contribute to different phe-
264-31 notypes associated with the syndrome. Sequence analysis
264-32 also reveals the origin of this deletion because the normal
264-33 sequence is bounded by repeated copies of a gene called
264-37 PMS, which happens to encode a DNA-repair protein. As
we have seen, repeated sequences can act as substrates for
pn fa
13 13
KEY CONCEPT Deletions can be recognized by deletion 1 1
loops and pseudodominance.
14 14
15 15
Deletions in human chromosomes are not uncommon. 21 21
Although these deletions are usually small, they still may
encompass many genes. In this case, a deletion will likely 22 22
have adverse effects, even when heterozygous, due to the 2
23
2
23
disruption of normal gene balance for many genes. Dele-
tions of specific human chromosome regions cause unique
syndromes of phenotypic abnormalities. One example is 31 31
cri du chat syndrome, caused by a heterozygous deletion
32 32
of the tip of the short arm of chromosome 5 (Figure 17-22).
The specific bands deleted in cri du chat syndrome are 3 33 3 33
5p15.2 and 5p15.3, the two most distal bands identifiable 34 34
on 5p. (The short and long arms of human chromosomes 35 35
are traditionally called p and q, respectively.) The most
characteristic phenotype in the syndrome is the one that Normal 5 Deleted 5
gives it its name, the distinctive catlike mewing cries made FIGURE 17-22 Cri du chat syndrome is caused by the loss of the tip
by affected infants. Other manifestations of the syndrome of the short arm of one of the homologs of chromosome 5.
unequal crossing over. A crossover between flanking cop- some extent regardless of their genetic content. In diploid
ies of PMS on opposite ends of the chromosomal segment plants, on the other hand, the pollen produced by a deletion
leads to a Williams syndrome deletion and a duplication, heterozygote is of two types: functional pollen carrying
as shown in Figure 17-23. Although less is currently known the normal chromosome and nonfunctional (aborted) pol-
about the traits associated with having a duplication of the len carrying the deficient homolog. Thus, pollen cells seem
7q11.23 region, affected individuals seem to have traits to be sensitive to changes in the amount of chromosomal
that are the opposite of those individuals with a deletion material, and this sensitivity might act to weed out dele-
of this region of chromosome 7. This finding is consistent tions. This effect is analogous to the sensitivity of pollen
with the concept of gene balance that we considered in to whole-chromosome aneuploidy, described earlier in this
the previous section when comparing the effects of whole- chapter. Unlike animal sperm cells, whose metabolic activ-
chromosome monosomies and trisomies. ity relies on enzymes that have already been deposited in
Most human deletions, such as those that we have just them during their formation, pollen cells must germinate
considered, arise spontaneously in the gonads of a normal and then produce a long pollen tube that grows to fertilize
parent of an affected person; thus, no signs of the deletions the ovule. This growth requires that the pollen cell manu-
are usually found in the chromosomes of the parents. Less facture large amounts of protein, thus making it sensitive
commonly, deletion-bearing individuals appear among to genetic abnormalities in its own nucleus. Plant ovules,
the offspring of an individual having an undetected bal- in contrast, are quite tolerant of deletions, presumably
anced rearrangement of chromosomes. For example, cri du because they receive their nourishment from the surround-
chat syndrome can result from a parent heterozygous for ing maternal tissues.
a reciprocal translocation, because (as we will see) segre-
gation produces deletions. Deletions may also result from
recombination within a heterozygote having a pericentric Duplications
inversion (an inversion spanning the centromere) on one The processes that cause chromosomal mutations some-
chromosome. Both mechanisms will be detailed later in the times produce an extra copy of a chromosome region.
chapter. The duplicate regions can be located adjacent to each
Animals and plants show differences in the survival of other, called a tandem duplication; or the extra copy can
gametes or offspring that bear deletions. A male animal be located elsewhere in the genome, called an insertional
with a deletion in one chromosome produces sperm carry- duplication. A diploid cell containing a duplication will
ing one or the other of the two chromosomes in approx- have three copies of the chromosome region in question:
imately equal numbers. These sperm seem to function to two in one chromosome set and one in the other. This is
known as a duplication heterozygote. In meiotic prophase,
tandem-duplication heterozygotes are seen as a
loop consisting of the unpaired extra region.
Probable origin of the Williams syndrome deletion Analyses of genome DNA sequences have
Normal sequence PMS PMS revealed a high level of duplications in humans
and in most model organisms. Simple sequence
repeats, which are extensive throughout the
26–28 genes genome and useful as molecular markers in
mapping, were discussed in earlier chapters.
However, another class of duplications is based
Unequal crossover PMS PMS on duplicated units that are much bigger than
simple sequence repeats. Duplications in this
class are termed segmental duplications. The
duplicated units in segmental duplications gen-
PMS PMS erally range from 10 to 50 kilobases in length
and encompass whole genes and the regions
in between. An example of segmental dupli-
Williams syndrome
deletion cations on human chromosome 7 is shown in
PMS
Figure 17-24. Most of the duplications are dis-
plus
7q11.23 persed within chromosome 7, but there are
duplication some tandem duplications, and even some
syndrome
duplicated segments from chromosome 7 that
Duplication are found on other chromosomes. One of the
FIGURE 17-23 A crossover between left and right repetitive flanking genes results in
tandem segmental duplications overlaps with
two reciprocal rearrangements, one corresponding to the Williams syndrome deletion, the previously discussed chromosomal rear-
and the other to the 7q11.23 duplication syndrome. rangement associated with Williams syndrome.
Chr 7
Williams syndrome
FIGURE 17-24 A map of human chromosome 7 shows the positions where recombination between duplications has given rise to the Williams
of duplications greater than 10 kilobases in size. Blue connecting syndrome deletion is indicated. [Data from J. A. Bailey et al., “Recent
lines show intrachromosomal duplications (the great majority). Segmental Duplications in the Human Genome,” Science 297, 2002,
Interchromosomal duplications are shown with red bars. The location 1003–1007.]
Duplicated units dispersed within the same chromosome Because inversions are balanced rearrangements, they do
are called intrachromosomal, while duplicated segments not change the overall amount of genetic material, and so
found on other chromosomes are known as interchro- they do not result in gene imbalance. Individuals with inver-
mosomal. Every chromosome in the human genome con- sions are generally normal, if there are no breaks within
tains numerous intra- and interchromosomal segmental genes. A break that disrupts a gene produces a mutation
duplications. that may be detectable as an abnormal phenotype. If the
Segmental duplications are thought to have an impor- gene has an essential function, then the break point acts as
tant role as substrates for nonallelic homologous recombi- a lethal mutation linked to the inversion. In such a case,
nation (NAHR). As shown in Figure 17-19, crossing over the inversion cannot be bred to homozygosity. However,
between segmental duplications can lead to various chromo- many inversions can be made homozygous, and further-
somal rearrangements. Such rearrangements seem to have more, inversions can be detected in haploid organisms. In
been important in human evolution. For example, major these cases, the break points of the inversion are clearly not
inversions that are key differences between human and ape in essential regions. Some of the possible consequences of
sequences have almost certainly come from NAHR mediated inversion at the DNA level are shown in Figure 17-25.
by segmental duplications. It also seems likely that NAHR Most analyses of inversions are carried out on diploid
is responsible for rearrangements that cause some human cells that contain one normal chromosome set plus one set
diseases. In addition to the previously discussed association carrying the inversion. This type of cell is called an inver-
between a segmental duplication on chromosome 7 with sion heterozygote, but note that this designation does not
Williams syndrome (Figure 17-23), segmental duplications imply that any gene locus is heterozygous; rather, it means
are found more frequently than expected in regions of the that for the chromosome pair with the inversion, there is
genome that have been associated with dozens of other one normal chromosome and one inverted chromosome
human genetic disorders, including neurofibromatosis, hemo- present in the cell. The location of the inverted segment can
philia A, and red-green color blindness. often be detected microscopically. In meiosis, one chromo-
some twists once at the ends of the inversion to pair with its
KEY CONCEPT Crossing over between segmental duplica- untwisted homolog; in this way, the paired homologs form
tions can lead to other chromosomal rearrangements. a visible inversion loop (Figure 17-26).
In a paracentric inversion, crossing over within the inver-
sion loop at meiosis connects homologous centromeres
Inversions in a dicentric bridge while also producing an acentric
We have seen that, to create an inversion, a segment of a fragment (Figure 17-27). Then, as the chromosomes separate
chromosome is cut out, flipped, and reinserted. Inversions are in anaphase I, the centromeres remain linked by the bridge.
of two basic types. If the centromere is outside the inversion, The acentric fragment cannot align itself or move; conse-
the inversion is said to be a paracentric inversion. Inversions quently, it is lost. Tension eventually breaks the dicentric
spanning the centromere are pericentric inversions. bridge, forming two chromosomes with terminal deletions.
Either the gametes containing such chromosomes or the
A B C D E F
Normal sequence zygotes that they eventually form will probably be invia-
A B C E D F ble. Hence, a crossover event, which normally generates the
Paracentric recombinant class of meiotic products, is instead lethal to
A D C B E F those products. The overall result is a drastically lower fre-
Pericentric
quency of viable recombinants. In fact, for genes within the
ANIMAT E D ART inversion, the recombinant frequency is close to zero. (It is
not exactly zero because rare double crossovers between only
Formation of paracentric inversions
two chromatids are viable.) For genes flanking the inversion,
Inversion
One breakpoint between genes and one within gene C: A –“C” – B – “C” – D
A “C ” B “C ” D
Inversion
A D C B A D
Inversion
the recombination frequency (RF) is reduced in proportion to Hence, the RF value of genes within a pericentric inversion
the size of the inversion because, for a longer inversion, there also is zero.
is a greater probability of a crossover occurring within it and Inversions affect recombination in another way, too.
producing an inviable meiotic product. Inversion heterozygotes often have mechanical pairing
In a heterozygous pericentric inversion, the net genetic problems in the region of the inversion. The inversion loop
effect is the same as that of a paracentric inversion— causes a large distortion that can extend beyond the loop
crossover products are not recovered—but the reasons are itself. This distortion reduces the opportunity for crossing
different. In a pericentric inversion, the centromeres are over in the neighboring regions.
contained within the inverted region. Consequently, the Let us consider an example of the effects of an inversion
chromosomes that have engaged in crossing over separate on recombinant frequency. A wild-type Drosophila speci-
in the normal fashion, without the creation of a bridge men from a natural population is crossed with a homozy-
(Figure 17-28). However, the crossover produces chroma- gous recessive laboratory stock dp cn/dp cn. (The dp allele
tids that contain a duplication and a deletion for different encodes dumpy wings and cn encodes cinnabar eyes. The
parts of the chromosome. In this case, if a gamete carry- two genes are known to be 45 map units apart on chromo-
ing a crossover chromosome is fertilized, the zygote dies some 2.) The F1 generation is wild type. When an F1 female
because of gene imbalance. Again, the result is that only is crossed with the recessive parent, the progeny are
noncrossover chromatids are present in viable progeny.
250 wild type + +/dp cn
Inversion loops at meiosis 246 dumpy cinnabar dp cn/dp cn
Inversion loop
5 Dumpy dp +/dp cn
b c c
a 7 Cinnabar + cn /dp cn
d e b d
Normal
a
c In this cross, which is effectively a dihybrid testcross,
a b d e
d
Paracentric inversions can lead to deletion products Pericentric inversions can lead to duplication-and-
deletion products
A B C D E
Paracentric- A B C D
inversion heterozygote Pericentric-
A D C B E inversion
A C B D heterozygote
Pairing Pairing
A E A D
Segregation
Segregation
FIGURE 17-27 A crossover AN IM ATED Note that the length ratio of the long arm to the short
in the loop of a paracentric- A RT arm has been changed from about 4:1 to about 1:1 by
inversion heterozygote gives rise to
chromosomes containing deletions.
Meiotic behavior of the inversion. Paracentric inversions do not alter the arm
paracentric inversions ratio, but they may be detected microscopically by observ-
ing changes in banding or other chromosomal landmarks,
Because the expected RF was based on measurements made if available. The ability to sequence whole genomes (see
on laboratory strains, the wild-type fly from nature was the Chapter 14) has also enabled the discovery of inversions
most likely source of the inverted chromosome. Hence, chro- that are not cytologically visible and revealed that many
mosome 2 in the F1 can be represented as follows: inversions exist both within and between species. For
example, a comparison of the chimpanzee and human
dp cn genomes revealed that there are over 1500 inversions
between these two species, including 33 that encompass
( ) more than 100 kb (see Figure 17-29). Previously, only nine
+ Inversion + pericentric inversions between these two species had been
identified using cytogenetic methods. The phenotypic con-
KEY CONCEPT The main diagnostic features of heterozy- sequences of these inversions are not known, but such
gous inversions are inversion loops, reduced recombinant
inversions have been proposed to play key roles in forming
frequency, and reduced fertility because of unbalanced or
and maintaining reproductive barriers between species (see
deleted meiotic products.
Chapter 20).
Reciprocal translocations products is deficient for a different arm of the cross and has
a duplicate of the other. These products are inviable. On
ANIMATED ART
the other hand, the two normal chromosomes may segre-
Formation of reciprocal translocations gate together, as will the reciprocal parts of the translocated
ones, to produce N1 + N 2 and T1 + T2 products. This segre-
There are several types of translocations, but here we gation pattern is called alternate segregation. These prod-
consider only reciprocal translocations, the simplest type. ucts are both balanced and viable.
Recall that, to form a reciprocal translocation, two nonho- Adjacent-1 and alternate segregations are equal in num-
mologous chromosomes trade acentric fragments created ber, and so half the overall population of gametes will be
by two simultaneous chromosome breaks (see the Chapter nonfunctional, a condition known as semisterility or “half-
17 opening photo on page 557). As with other rearrange- sterility.” Semisterility is an important diagnostic tool for
ments, meiosis in heterozygotes having two translocated identifying translocation heterozygotes. However, semisteril-
chromosomes and their normal counterparts produces ity is defined differently for plants and animals. In plants, the
characteristic configurations. Figure 17-30 illustrates mei- 50 percent of meiotic products that are from the adjacent-1
osis in an individual that is heterozygous for a reciprocal segregation generally abort at the gametic stage (Figure 17-31).
translocation. Note the cross-shaped pairing configura- In animals, these products are viable as gametes but lethal to
tion. Because the law of independent assortment is still in the zygotes that they produce on fertilization.
force, there are two common patterns of segregation. Let Remember that heterozygotes for inversions also may
us use N1 and N2 to represent the normal chromosomes show some reduction in fertility but by an amount depen-
and T1 and T2 as the translocated chromosomes. The seg- dent on the size of the affected region. Thus, the precise
regation of each of the structurally normal chromosomes 50 percent reduction in viable gametes or zygotes is
with one of the translocated ones (T1 + N 2 and T2 + N1 ) usually a reliable diagnostic clue for the presence of a
is called adjacent-1 segregation. Each of the two meiotic translocation.
N1 N2
Normal
T1 T2
Translocated
T1 N2
Pairing configuration
N1 T2
T1 N2 T1 N2
N1 T2 N1 T2
T1 N1 N1 T1
N2 T2 N2 T2
FIGURE 17-31 Pollen of a semisterile corn plant. The clear pollen grains
contain chromosomally unbalanced meiotic products of a reciprocal-
translocation heterozygote. The opaque pollen grains, which contain
either the complete translocation genotype or normal chromosomes, are
functional in fertilization and development. [William Sheridan.]
Pseudolinkage of genes in a translocation heterozygote result from a translocation in one of the parents. We have
seen that translocations can produce progeny that have
Translocation heterozygote Normal-sequence tester
extra material from part of the genome, and so a translo-
a b a b
cation concerning chromosome 21 can produce progeny
× with extra material from that chromosome. In Down syn-
drome, the translocation responsible is of a type called a
b a a b Robertsonian translocation, which involves breakage of
two chromosomes at or near their centromeres, and subse-
Viable progeny
a b a b
quent fusions of the long arms of the chromosomes as well
as loss of the short arms of the chromosomes. Note that
in humans, Robertsonian translocations usually involve the
five chromosomes (13, 14, 15, 20, and 21) with almost no
b a a b
unique genes on their short arms. Thus, loss of these short
Phenotype ab Phenotype a b arms can be tolerated. In the case of a Robertsonian trans-
location involving chromosome 21, the progeny therefore
FIGURE 17-32 When a AN IM AT E D
translocated fragment carries a A RT
will carry an almost complete extra copy of chromosome
marker gene, this marker can 21. The translocation and its segregation are illustrated in
Pseudolinkage of genes by
show linkage to genes on the other Figure 17-33. Note that, in addition to complements caus-
chromosome. All viable progeny that reciprocal translocations
ing Down syndrome, other aberrant chromosome comple-
inherit the translocation chromosome
ments are produced, most of which abort. In our example,
will show the wild-type phenotypes associated with genes a and b,
while all viable progeny that do not inherit the translocation will the man may have this translocation, which he may have
show the mutant phenotypes associated with genes a and b. inherited from his grandmother. To confirm this possibility,
his chromosomes would be checked. His unaffected child
might have normal chromosomes or might have inherited
Genetically, genes on translocated chromosomes act as
his translocation.
though they are linked if their loci are close to the translo-
cation break point. Figure 17-32 shows a translocation het- Applications of inversions
erozygote that has been established by crossing an a/a ; b/b
individual with a translocation homozygote bearing the
and translocations
wild-type alleles. When the heterozygote is testcrossed, Inversions and translocations have proven to be useful
recombinants are created but do not survive because they genetic tools; some examples of their uses follow.
carry unbalanced genomes (duplication-and-deletions).
The only viable progeny are those bearing the parental Balancer chromosomes In some model experimental
genotypes; so linkage is seen between loci that were orig- systems, notably Drosophila and the nematode Caenorhab-
inally on different chromosomes. The apparent linkage of ditis elegans, inversions have a practical use as balancers.
genes normally known to be on separate nonhomologous A balancer chromosome contains multiple inversions; so,
chromosomes—sometimes called pseudolinkage —is a when it is combined with the corresponding wild-type chro-
genetic diagnostic clue to the presence of a translocation. mosome, there can be no viable crossover products. In some
analyses, it is important to keep all the alleles on one chro-
KEY CONCEPT Heterozygous reciprocal translocations mosome together with no recombination between them. The
are diagnosed genetically by semisterility and by the apparent
geneticist creates individuals having genotypes that combine
linkage of genes whose normal loci are on separate
such a chromosome of interest with a balancer. This combi-
chromosomes.
nation eliminates progeny with crossovers, so only parental
combinations appear in the progeny. For convenience, bal-
ancer chromosomes are marked with a dominant morpho-
Robertsonian translocations logical mutation. The marker allows the geneticist to track
Let’s return to the family with the Down syndrome child, the segregation of the entire balancer or its normal homolog
introduced at the beginning of the chapter. The birth of two by noting the presence or absence of the marker.
children with Down syndrome in the family can indeed be
a coincidence—after all, coincidences do happen. However, Gene mapping Inversions and translocations are useful
the miscarriage gives a clue that something else might be for mapping and subsequent isolation of specific genes.
going on. A large proportion of spontaneous abortions The gene for human neurofibromatosis was isolated in
carry chromosomal abnormalities, so perhaps that is the this way. The critical information came from people who
case in this example. If so, the couple may have had two not only had the disease, but also carried chromosomal
conceptions with chromosome mutations, which would be translocations. All the translocations had one break point
very unlikely unless there was a common cause. However, in common, in a band close to the centromere of chromo-
a small proportion of Down syndrome cases are known to some 17. Hence, this band appeared to be the locus of the
Gametes from normal parent Gametes from translocation carrier Gametes from translocation carrier
Down syndrome
+
Translocation carrier
Normal
+
Lethal
+
Lethal
Lethal trisomy 14
FIGURE 17-33 In a small minority of cases, the origin of Down syndrome is a parent heterozygous
for a Robertsonian translocation concerning chromosome 21. Meiotic segregation results in some
gametes carrying a chromosome with a large additional segment of chromosome 21. In combination
with a normal chromosome 21 provided by the gamete from the opposite sex, the symptoms of Down
syndrome are produced even though there is not full trisomy 21.
neurofibromatosis gene, which had been disrupted by the is perhaps surprising that there is great variation in both
translocation break point. Subsequent analysis showed the number and structure of chromosomes among species.
that the chromosome 17 break points were not at identi- For example, the human and chimpanzee genomes differ
cal positions; however, because they must have been within not only by over 1500 inversions, but also by chromosome
the gene, the range of their positions revealed the segment number. Human chromosome 2 is the result of a fusion
of the chromosome that constituted the neurofibromato- between two chromosomes, represented by chromosomes
sis gene. The isolation of DNA fragments from this region 12 and 13 in the chimpanzee genome. Across mammalian
eventually led to the recovery of the gene itself. species, the diploid chromosome number varies greatly,
ranging from 6 to 102. The mammalian species with the
Synthesizing specific duplications or deletions lowest chromosome number is the Indian muntjac: females
Translocations and inversions are routinely used to delete have only 6 chromosomes (3 pairs) and males have 7
or duplicate specific chromosome segments. Recall, for (because the Y chromosome is unpaired). Remarkably,
example, that pericentric inversions as well as transloca- the closely related Chinese muntjac has 46 chromosomes
tions generate products of meiosis that contain a duplica- (23 pairs), suggesting there have been many chromosomal
tion and a deletion (see Figures 17-28 and 17-30). If the fusion events in the Indian muntjac ( Figure 17-34). These
duplicated or the deleted segment is very small, then the two species have been observed to mate and produce viable
duplication-and-deletion meiotic products are often viable. offspring in captivity. However, these offspring are sterile,
Duplications and deletions are useful for a variety of experi- consistent with problems in meiotic pairing of these chro-
mental applications, including the mapping of genes and the mosomes. Dramatic differences in chromosome number
varying of gene dosage for the study of regulation, as seen and structure can also occur within species. For example,
in preceding sections. some populations of house mice have a reduced number of
chromosomes due to the presence of multiple Robertsonian
17.3 PHENOTYPIC CONSEQUENCES translocations.
These mammalian examples are just a few to high-
OF CHROMOSOMAL CHANGES light the remarkable diversity of chromosome number
and structure found in nature. Currently, almost nothing
LO 17.5 Distinguish among the main human syndromes is known about the mechanisms that evolved to overcome
resulting from chromosomal changes. the deleterious effects of these chromosome rearrange-
ments. However, many scientists have hypothesized that
Chromosome rearrangements and evolution chromosome rearrangements, such as inversions, might
As we have seen in this chapter, chromosomal changes actually facilitate evolutionary change. Recent research
within species can have many detrimental effects. Thus, it in a number of different systems, including plants, ants,
1 2 3 4 5
1 2 3+X
6 7 8 9 10 11 12
13 14 15 16 17 18
19 20 21 22 1 2 3 3+X Y
X Y
butterflies, and birds, supports this hypothesis. A classic a cancer. A localized knot of proliferated cells is called a
example is Müllerian mimicry in butterflies, in which two tumor, whereas cancers of mobile cells such as blood cells
distasteful species mimic each other’s wing patterns. In this disperse throughout the body. Cancer is most often caused
co-mimicry, predators more easily learn to avoid a particu- by mutations in the coding or regulatory sequence of genes
lar wing pattern, providing a benefit for both species. It has whose normal function is to regulate cell division. Such
been observed that the genes controlling different aspects genes are called proto-oncogenes before a cancer-causing
of the wing pattern are tightly linked to each other, cre- mutation occurs, and oncogenes after a cancer-causing
ating a so-called supergene. Any recombination between mutation occurs. Chromosomal rearrangements, especially
these genes would create hybrid wing patterns that might translocations, can interfere with the normal function of
no longer be recognized as distasteful by predators, and such proto-oncogenes.
thus increase the risk that the butterfly would be eaten. As There are two basic ways in which translocations can
we have seen, recombination between genes is rare within alter the function of proto-oncogenes. In the first mecha-
inversions, and it was therefore predicted that these wing nism, the translocation relocates a proto-oncogene next
pattern supergenes would be found within inversions. to a new regulatory element. A good example is provided
Indeed, there are now examples of butterfly species in by Burkitt lymphoma. The proto-oncogene in this cancer
which wing pattern supergenes are found in chromosomal encodes the protein MYC, a transcription factor that acti-
inversions (Figure 17-35). Similar evidence is accumulating vates genes required for cell proliferation. Normally, the
in other systems, which emphasizes that chromosome rear- MYC gene is transcribed only when a cell needs to undergo
rangements are an important substrate for evolution. proliferation, but in cancerous cells, the proto-oncogene
MYC is relocated next to the regulatory region of immu-
noglobulin (Ig) genes (Figure 17-36a). These immunoglobu-
Chromosome rearrangements and cancer lin genes are constitutively transcribed; that is, they are on
Cancer is a disease of abnormal cell proliferation. As a all the time. Consequently, the MYC gene is transcribed at
result of some insult inflicted on it, a cell of the body all times, and the cell-proliferation genes are continuously
divides out of control to form a population of cells called activated.
3 species of Melinaea
(unpalatable)
Mimicry
3 mimetic forms of
Heliconius numata
Supergene
(H. numata)
FIGURE 17-35 Heliconius numata is a distasteful butterfly species that is a Müllerian mimic with different
species of the butterfly genus Melinaea. When the two species co-occur, H. numata has a wing pattern
that mimics the local Melinaea species. However, these wing patterns differ between Melinaea species, and
so H. numata wing patterns differ from location to location. The genes controlling these differences in wing
pattern in H. numata are found in a supergene on chromosome 15. Distinct wing patterns are associated with
different chromosomal inversions around the supergene. [Courtesy of Mathieu Joron.]
RegIG MYC
Translocation
Translocation
break point
Chr 22
BCR1
Hybrid oncogene
Translocation
FIGURE 17-36 The two main ways that translocations can cause cancer in a body (somatic)
cell are illustrated by the cancers Burkitt lymphoma (a) and chronic myelogenous leukemia (b).
The genes MYC, BCR1, and ABL are proto-oncogenes.
The other mechanism by which translocations can cellular processes are prone to a high level of error.
cause cancer is the formation of a hybrid gene. An exam- Figure 17-37 shows the estimated distribution of chro-
ple is provided by the disease chronic myelogenous leuke- mosome mutations among human conceptions that
mia (CML), a cancer of white blood cells. This cancer can develop sufficiently to implant in the uterus. Of the
result from the formation of a hybrid gene between the estimated 15 percent of conceptions that abort spon-
two proto-oncogenes BCR1 and ABL (Figure 17-36b). taneously (pregnancies that terminate naturally), fully
The ABL proto-oncogene encodes a protein kinase in a half show chromosomal abnormalities. Some medical
signaling pathway. The protein kinase passes along a sig- geneticists believe that even this high level is an under-
nal initiated by a growth factor that leads to cell pro- estimate because many cases are never detected. Among
liferation. The protein kinase activity of the BCR1-ABL live births, 0.6 percent have chromosomal abnormali-
fusion protein is always on. The fusion protein contin- ties, resulting from both aneuploidy and chromosomal
ually propagates its growth signal onward, regardless of rearrangements.
whether the initiating signal is present.
1,000,000
conceptions
850,000 150,000
live births spontaneous abortions
833,000 17,000
children perinatal deaths
5,165 75,000
chromosome abnormalities chromosome abnormalities
758 balanced
Robertsonian 13,500 XO
translocations 1,183 autosomal
trisomics
42 trisomy 13
100 trisomy 18
1,041 trisomy 21
FIGURE 17-37 The proportion of chromosomal mutations is much higher in spontaneous abortions.
[Data from K. Sankaranarayanan, Mutat. Res. 61, 1979, 249–257.]
SUMMARY
Polyploidy is an abnormal condition in which there is a hence can produce balanced gametes and progeny. Poly-
larger-than-normal number of chromosome sets. Polyploids ploidy can result in an organism of larger dimensions; this
such as triploids (3n) and tetraploids (4n) are common discovery has permitted important advances in horticulture
among plants and are represented even among animals. and in crop breeding.
Organisms with an odd number of chromosome sets are In plants, allopolyploids (polyploids formed by combin-
sterile because not every chromosome has a partner at meio- ing chromosome sets from different species) can be made by
sis. Unpaired chromosomes segregate randomly to the poles crossing two related species and then doubling the progeny
of the cell in meiosis, leading to unbalanced sets of chromo- chromosomes through the use of colchicine. These techniques
somes in the resulting gametes. Such unbalanced gametes do have potential applications in crop breeding because allo-
not yield viable progeny because the normal balance of gene polyploids combine the features of the two parental species.
dosage is disrupted. In polyploids with an even number of When cellular accidents change parts of chromosome
sets, each chromosome has a potential pairing p artner and sets, aneuploids result. Aneuploidy itself usually results in
unbalanced dosage of gene products, which leads to abnor- dosage imbalance or because they uncover recessive deleteri-
mal phenotypes. Examples of aneuploids include monosomics ous alleles, or they may be nonlethal. When a deletion in one
(2n − 1) and trisomics (2n + 1). Down syndrome (trisomy 21), homolog allows the phenotypic expression of recessive alleles
Klinefelter syndrome (XXY), and Turner syndrome (XO) in the other, the unmasking of the recessive alleles is called
are well-documented examples of aneuploid conditions in pseudodominance.
humans. The spontaneous level of aneuploidy in humans is Duplications are generally produced from other
quite high and accounts for a large proportion of genetically rearrangements or by aberrant crossing over. They also
based disorders in human populations. The phenotype of an unbalance the genetic material, producing a deleterious
aneuploid organism depends very much on the particular chro- p henotypic effect or death of the organism. Segmental
mosome affected. In some cases, such as human trisomy 21, duplications are also a substrate for additional chro-
there is a highly characteristic constellation of associated mosomal rearrangements due to nonallelic homologous
phenotypes. recombination. Many human chromosome disorders are
Most instances of aneuploidy result from accidental associated with regions of the genome that harbor segmen-
chromosome missegregation at meiosis (nondisjunction). tal duplications.
The error is spontaneous and can occur in any particu- An inversion is a 180-degree turn of a part of a chromo-
lar meiocyte at the first or second division. In humans, a some. In the homozygous state, inversions may cause little
maternal-age effect is associated with nondisjunction of problem for an organism unless one of the breaks disrupts
chromosome 21, resulting in a higher incidence of Down a gene. On the other hand, inversion heterozygotes show
syndrome in the children of older mothers. inversion loops at meiosis, and crossing over within the
The other general category of chromosome mutations loop results in inviable products. The crossover products
comprises structural rearrangements, which include dele- of pericentric inversions, which span the centromere, dif-
tions, duplications, inversions, and translocations. These fer from those of paracentric inversions, which do not, but
changes result either from breakage and incorrect reunion both show reduced recombinant frequency in the affected
or from crossing over between repetitive elements (nonallelic region and often result in reduced fertility.
homologous recombination). In individuals heterozygous for A translocation moves a chromosome segment to another
a chromosome rearrangement (i.e., with one normal chro- position in the genome. A simple example is a reciprocal
mosome homolog and one rearranged chromosome homo- translocation, in which parts of nonhomologous chromo-
log), there are unusual pairing structures at meiosis resulting somes exchange positions. In the heterozygous state, trans-
from the strong pairing affinity of homologous chromosome locations produce duplication-and-deletion meiotic products,
regions. For example, heterozygous inversions show loops, which can lead to unbalanced zygotes. New gene linkages
and reciprocal translocations show cross-shaped structures. can be produced by translocations. The random segregation
Segregation of these structures results in abnormal meiotic of centromeres in a translocation heterozygote results in 50
products unique to the rearrangement. percent unbalanced meiotic products and, hence, 50 percent
A deletion is the loss of a section of chromosome, either sterility (semisterility).
because of chromosome breaks followed by loss of the Chromosomal rearrangements are an important cause
intervening segment or because of segregation in hetero- of sterility, birth defects, and disease in human populations.
zygous translocations or inversions. If the genes removed However, they are also substrates for evolution and useful
in a deletion are essential to life, a homozygous deletion is in engineering special strains of organisms for experimental
lethal. Heterozygous deletions may be lethal because of gene and applied genetics.
KEY TERMS
acentric chromosome (p. 574) chromosome mutation (p. 558) gene balance (p. 571)
acentric fragment (p. 579) deletion (p. 574) gene-dosage effect (p. 571)
adjacent-1 segregation (p. 582) deletion loop (p. 576) genetic load (p. 560)
allopolyploid (p. 560) deletion mapping (p. 576) haploid (p. 559)
alternate segregation (p. 582) dicentric bridge (p. 579) haploid chromosome number (p. 559)
amphidiploid (p. 563) dicentric chromosome (p. 574) hexaploid (p. 559)
anaphase bridge (p. 574) diploid (p. 559) homeologous chromosomes (p. 560)
aneuploid (p. 560) disomic (p. 564) insertional duplication (p. 578)
autopolyploid (p. 560) dosage compensation (p. 572) intragenic deletion (p. 576)
balanced rearrangement (p. 575) Down syndrome (p. 569) inversion (p. 574)
balancer chromosome (p. 584) duplication (p. 574) inversion heterozygote (p. 579)
bivalent (p. 560) euploid (p. 559) inversion loop (p. 579)
Klinefelter syndrome (p. 567) pentaploid (p. 559) tandem duplication (p. 578)
monoploid (p. 559) pericentric inversion (p. 579) tetraploid (p. 559)
monosomic (p. 564) polyploid (p. 559) translocation (p. 574)
multigenic deletion (p. 576) polytene chromosome (p. 576) triploid (p. 559)
nonallelic homologous recombination pseudodominance (p. 576) trisomic (p. 564)
(NAHR) (p. 574) pseudolinkage (p. 584) trivalent (p. 560)
nondisjunction (p. 564) rearrangement (p. 574) Turner syndrome (p. 565)
nullisomic (p. 564) Robertsonian translocation (p. 584) unbalanced rearrangement (p. 574)
paracentric inversion (p. 579) segmental duplication (p. 578) univalent (p. 560)
parthenogenesis (p. 560) semisterility (p. 582)
SOLVED PROBLEMS
SOLVED PROBLEM 1 All the progeny from this cross will be heterozygous
A corn plant is heterozygous for a reciprocal translocation for the chromosome carrying the brachytic allele,
and is therefore semisterile. This plant is crossed with a chro- but what about the chromosomes taking part in the
mosomally normal strain that is homozygous for the recessive translocation? In this chapter, we have seen that only
allele brachytic (b), located on chromosome 2. A semisterile F1 alternate-segregation products survive and that half
plant is then backcrossed to the homozygous brachytic strain. of these survivors will be chromosomally normal and
The progeny obtained show the following phenotypes: half will carry the two rearranged chromosomes. The
rearranged combination will regenerate a translocation
Nonbrachytic Brachytic
heterozygote when it combines with the chromosom-
Semisterile Fertile Semisterile Fertile ally normal complement from the normal parent. These
334 27 42 279 latter types—the semisterile F1’s—are diagrammed as
part of the backcross to the parental brachytic strain:
a. What ratio would you expect to result if the chromo-
some carrying the brachytic allele does not take part in Semisterile F1
the translocation?
b. Do you think that chromosome 2 takes part in the b
translocation? Explain your answer, showing the con-
formation of the relevant chromosomes of the semister- b
ile F1 and the reason for the specific numbers obtained.
SOLUTION
Tester
a. We should start with the methodical approach and
simply restate the data in the form of a diagram, where
b
normal progeny from this cross, and so the hypothesis • Second intercrosses of normal progeny: V × V /v.
is already overthrown. Linkage could save the hypoth- These intercrosses give 25 percent each of V/V, V/v,
esis temporarily if we assumed that the mutation was V (all normals), and v (waltzers). This also fits.
in the normal parent, giving a gamete V S. Then the This hypothesis is therefore consistent with the data.
F1 waltzer would be V S/v s, and, if linkage were tight
d. Deletion of V in normal parent
enough, few or no V s gametes would be produced, the
type that are necessary to combine with the v s gamete Let’s call the deletion D. The F1 waltzer would be D/v,
from the male to give V s/v s normals. However, if the and the subsequent matings would be
linkage hypothesis were true, the cross with the normal • D/v × v /v, which gives v/v and D/v, which are waltz-
males would be V S /v s × V s /V s, and this would give ers. This fits.
a high percentage of V S/V s progeny, which would be
• D/v × V /V , which gives V/v and D/V, which are nor-
waltzers, none of which were seen.
mal. This fits.
c. Nondisjunction in the normal parent
• First intercrosses of normal progeny: D/V × D/V ,
This explanation would give a nullisomic gamete that which give D/ V and V/V, all normal. This fits.
would combine with v to give the F1 waltzer the hemi- Again, note that the D/D progeny produced by this
zygous genotype v. The subsequent matings would be intercross would likely not survive.
• v × v /v, which gives v/v and v progeny, all waltzers. • Second intercrosses of normal progeny: D/V × V /v,
This fits. which give 25 percent each of V/V, V/v, D/V (all
• v × V /V , which gives V/v and V progeny, all nor- normals), and D/v (waltzers). This also fits.
mals. This also fits. Once again, the hypothesis fits the data provided; so we
• First intercrosses of normal progeny: V × V . These are left with two hypotheses that are compatible with the
intercrosses give V and V/V, which are normal. This results, and further experiments are necessary to distinguish
fits. Note that this intercross would also produce them. One way of doing so would be to examine the chro-
progeny that are homozygous for the loss of the mosomes of the exceptional female under the microscope:
chromosome, but these progeny would not survive. aneuploidy should be easy to distinguish from deletion.
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading. Problems with the
icon have a Problem Solving Video. Problems with the icon have an Unpacking the Problem exercise.
www
www
WORKING WITH THE FIGURES the genome of B. rapa as “C.” Which species shown in
(The first 20 questions require inspection of text figures.) this figure would have an “AC” genome?
1. Based on Table 17-1, how would you categorize the 6. In Figure 17-10, what would be the constitution of
following genomes? (Letters H through J stand for an individual formed from the union of a monosomic
four different chromosomes.) from a first-division nondisjunction in a female and
a disomic from a second-division nondisjunction in a
HH II J KK male, assuming the gametes were functional?
HH II JJ KKK
7. In Figure 17-12, what would be the expected percent-
HHHH IIII JJJJ KKKK age of each type of segregation?
2. Based on Figure 17-4, how many chromatids are in a 8. In Figure 17-19, is there any difference between the
trivalent? inversion products formed from breakage and those
3. Based on Figure 17-5, if colchicine is used on a plant in formed from crossing over?
which 2n = 18, how many chromosomes would be in 9. Referring to Figure 17-19, draw a diagram showing
the abnormal product? the process whereby an inversion formed from cross-
4. Basing your work on Figure 17-8, use colored pens to ing over could generate a normal sequence.
represent the chromosomes of the fertile amphidiploid. 10. In Figure 17-21, would the phenotype associated with
5. In Figure 17-9 , we can designate the genome of the recessive fa allele be visible when paired with dele-
B. oleracea as “A,” the genome of B. nigra as “B,” and tion 264-32? With 258-11?
11. Look at Figure 17-22 and state which bands are miss- 29. A disomic product of meiosis is obtained. What is
ing in the cri du chat deletion. its likely origin? What other genotypes would you
12. Explain why the phenotypes associated with the expect among the products of that meiosis under your
Williams syndrome deletion and the 7q11.23 duplica- hypothesis?
tion shown in Figure 17-23 are not the same. 30. Can a trisomic A/A/a ever produce a gamete of geno-
13. Referring to Figure 17-25, draw the product if breaks type a?
occurred within genes A and B. 31. Which, if any, of the following sex-chromosome aneu-
14. In Figure 17-27, what would be the consequence of a ploids in humans are fertile: XXX, XXY, XYY, XO?
crossover between the centromere and locus A? 32. If you observed a dicentric bridge at meiosis, what
15. In Figure 17-29, which of the three pairs of homolo- rearrangement would you predict had taken place?
gous human and chimpanzee chromosomes shown dif- 33. Why do acentric fragments get lost?
fer in the ratio of their arm length? 34. Diagram a translocation arising from repetitive DNA.
16. Based on Figure 17-30 , are normal gametes ever Repeat for a deletion.
formed from an adjacent-1 segregation? 35. From a large stock of Neurospora rearrangements
17. Referring to Figure 17-32, draw an inviable product available from the fungal genetics stock center, what
from the same meiosis. type would you choose to synthesize a strain that had
18. Based on Figure 17-33 , what fraction of progeny a duplication of the right arm of chromosome 3 and a
would be phenotypically normal in a cross between a deletion for the tip of chromosome 4?
translocation heterozygote and a normal parent? 36. You observe a very large pairing loop at meiosis. Is it
19. Based on Figure 17-36, write a sentence stating how more likely to be from a heterozygous inversion or a
translocation can lead to cancer. Can you think of heterozygous deletion? Explain.
another genetic cause of cancer? 37. A new recessive mutant allele does not show
20. Using Figure 17-37, calculate what percentage of con- pseudodominance with any of the deletions that
ceptions are triploid. The same figure shows XO in the span Drosophila chromosome 2. What might be the
spontaneous-abortion category; however, we know explanation?
that many XO individuals are viable. In which of the 38. Compare and contrast the origins of Turner syndrome,
viable categories would XO be grouped? Williams syndrome, cri du chat syndrome, and Down
syndrome. (Why are they called syndromes?)
BASIC PROBLEMS
39. List the diagnostic features (genetic or cytological) that
21. In keeping with the style of Table 17-1, what would are used to identify these chromosomal alterations:
you call organisms that are MM N OO; MM NN OO; a. Deletions
MMM NN PP? b. Duplications
22. A large plant arose in a natural population. Quali- c. Inversions
tatively, it looked just the same as the others, except
d. Reciprocal translocations
much larger. Is it more likely to be an allopolyploid or
an autopolyploid? How would you test that it was a 40. The normal sequence of nine genes on a certain
polyploid and not just growing in rich soil? Drosophila chromosome is 123 ⋅ 456789, where the
dot represents the centromere. Some fruit flies were
23. Is a trisomic an aneuploid or a polyploid?
found to have aberrant chromosomes with the follow-
24. Seedless watermelons are sterile triploids. However, ing structures:
sometimes viable seeds are found in these watermel-
a. 123 ⋅ 476589 c. 123 ⋅ 46789
ons. What is an explanation for this finding?
b. 1654 ⋅ 32789 d. 123 ⋅ 4566789
25. How many different types of gametes can be produced
by a triploid organism with a haploid number of three Name each type of chromosomal rearrangement, and
(i.e., chromosomes A, B, C)? draw diagrams to show how each would synapse with
the normal chromosome.
26. In a tetraploid B/B/b/b, how many quadrivalent possi-
ble pairings are there? Draw them (see Figure 17-5). 41. The two loci P and Bz are normally 36 m.u. apart on
the same arm of a certain plant chromosome. A para-
27. Someone tells you that cauliflower is an amphidiploid. centric inversion spans about one-fourth of this region
Do you agree? Explain. but does not include either of the loci. What approxi-
28. Why is Raphanobrassica fertile, whereas its progenitor mate recombinant frequency between P and Bz would
was not? you predict in plants that are
a. heterozygous for the paracentric inversion? homozygous for the inversion was impossible even after
b. homozygous for the paracentric inversion? many attempts. What is the most likely explanation for
this inability to produce a homozygous inversion?
42. As stated in Solved Problem 2, recessive mutation
in certain mice called waltzers causes them to exe- 45. Orangutans are now recognized as a group of three
cute bizarre steps. W. H. Gates crossed waltzers with endangered species in their natural environments (one
pure-breeding normal mice and found, among several species on the island of Borneo and two species on
hundred normal progeny, a single waltzing female the island of Sumatra). Before the distinction between
mouse. This mouse was mated with a waltzing male, the three species was clear, a captive-breeding pro-
and her offspring were waltzers. When mated with a gram was established using orangutans held in zoos
homozygous normal male, all her progeny were nor- throughout the world. One component of this program
mal. Some of these normal males and females were is research into orangutan cytogenetics. This research
intercrossed, and, unexpectedly, none of their progeny has shown that all orangutans from Borneo carry one
were waltzers. T. S. Painter examined the chromosomes form of chromosome 2, as shown in the accompanying
of some of Gates’s waltzing mice that showed a breed- diagram, and all orangutans from Sumatra carry the
ing behavior similar to that of the original, unusual other form. Before this cytogenetic difference became
waltzing female. He found that these mice had the nor- known, some matings were carried out between ani-
mal number of 40 chromosomes. In the unusual waltz- mals from different islands, and 14 hybrid progeny are
ers, however, one member of a chromosome pair was now being raised in captivity.
abnormally short. Interpret these observations as com-
cen
q3.1 q2.2 q2.1 q1.2 q1.1 p1.1 p1.2
pletely as possible, both genetically and cytologically.
Borneo
43. A salivary-gland chromosome of Drosophila has six
bands as shown in the following illustration. Below
cen
q3.1 q2.2 q2.1 p1.1 q1.1 q1.2 p1.2
the chromosome are shown the extent of five deletions
Sumatra
(Del 1 to Del 5):
a. What key features of the data are different from the 20. What effect could lead to the absence of half the
expected results? seeds?
b. State a concise hypothesis that explains the results. 21. Did half the seeds die? If so, was the female or the
c. Show genotypes of parents and progeny. male parent the reason for the deaths?
Now try to solve the problem. If you are unable to do so,
d. Draw a diagram showing the arrangement of alleles
try to identify the obstacle and write a sentence or two
on the chromosomes.
describing your difficulty. Then go back to the expansion
e. Explain the origin of the two classes of progeny questions and see if any of them relate to your difficulty. If
having three and five members. this approach does not work, inspect the Learning Objec-
www
tives and Key Concepts of this chapter and ask yourself
UNPACKING PROBLEM 46 which might be relevant to your difficulty.
www
11. What does the approximate equality of the first two 48. In corn, the allele Pr stands for green stems, pr for pur-
progeny classes tell you? ple stems. A corn plant of genotype pr/pr that has stan-
dard chromosomes is crossed with a Pr/Pr plant that
12. What does the approximate equality of the second is homozygous for a reciprocal translocation between
two progeny classes tell you? chromosomes 2 and 5. The F1 is semisterile and pheno-
13. What were the gametes from the unusual plant, and typically Pr. A backcross with the parent with standard
what were their proportions? chromosomes gives 764 semisterile Pr, 145 semisterile
14. Which gametes were in the majority? pr, 186 normal Pr, and 727 normal pr. What is the map
distance between the Pr locus and the translocation
15. Which gametes were in the minority? point? www
Deduce the phenotypic proportions in the progeny of 55. A cross is made in tomatoes between a female plant
the following crosses of trisomics. that is trisomic for chromosome 6 and a normal dip-
a. b+ /b /b × b /b loid male plant that is homozygous for the recessive
allele for potato leaf (p/p). A trisomic F1 plant is back-
b. b+ /b+ /b × b /b crossed to the potato-leaved male.
c. b+ /b+ /b × b+ /b a. What is the ratio of normal-leaved plants to
52. A woman with Turner syndrome is found to be color- potato-leaved plants when you assume that p is
blind (an X-linked recessive phenotype). Both her located on chromosome 6?
mother and her father have normal vision. b. What is the ratio of normal-leaved to potato-leaved
a. Explain the simultaneous origin of Turner syn- plants when you assume that p is not located on
drome and color blindness by the abnormal behav- chromosome 6?
ior of chromosomes at meiosis. 56. A tomato geneticist attempts to assign five recessive
b. Can your explanation distinguish whether the mutations to specific chromosomes by using trisomics.
abnormal chromosome behavior occurred in the She crosses each homozygous mutant (2n) with each
father or in the mother? of three trisomics, in which chromosomes 1, 7, and
c. Can your explanation distinguish whether the 10 take part. From these crosses, the geneticist selects
abnormal chromosome behavior occurred at the trisomic progeny (which are less vigorous) and back-
first or second division of meiosis? crosses them to the appropriate homozygous recessive.
The diploid progeny from these crosses are examined.
d. Now assume that a color-blind Klinefelter man has Her results, in which the ratios are wild type:mutant,
parents with normal vision, and answer parts a, b, are as follows:
and c.
Trisomic Mutation
53. a. How would you synthesize a pentaploid?
chromosome d y c h cot
b. How would you synthesize a triploid of genotype
A/a/a? 1 48:55 72:29 56:50 53:54 32:28
c. You have just obtained a rare recessive mutation a* 7 52:56 52:48 52:51 58:56 81:40
in a diploid plant, which Mendelian analysis tells 10 45:42 36:33 28:32 96:50 20:17
you is A/a*. From this plant, how would you syn-
thesize a tetraploid (4n) of genotype A/A/a*/a*? Which of the mutations can the geneticist assign to
which chromosomes? (Explain your answer fully.)
d. How would you synthesize a tetraploid of genotype
A/a/a/a? 57. A petunia is heterozygous for the following autosomal
homologs:
54. Suppose you have a line of mice that has cytologically
distinct forms of chromosome 4. The tip of the chro- A B C D E F G H I
mosome can have a knob (called 4K) or a satellite (4S)
a b c d h g f e i
or neither (4). Here are sketches of the three types:
58. Two groups of geneticists, in California and in Chile, C gives colored aleurone; c, white aleurone.
begin work to develop a linkage map of the medfly Bz gives green leaves; bz, bronze leaves.
(see Chapter 4). They both independently find that
the loci for body color (B = black, b = gray) and eye Wx gives starchy seeds; wx, waxy seeds.
shape (R = round, r = star) are linked 28 m.u. apart. Sh gives smooth seeds; sh, shrunken seeds.
They send strains to each other and perform crosses; a D gives tall plants; d, dwarf.
summary of all their findings is shown here:
A plant from a standard stock that is homozygous
Progeny of for all five recessive alleles is crossed with a wild-type
Cross F1 F1 × any b r /b r plant from Mexico that is homozygous for all five
B R /B R (Calif.) B R/b r B R/b r 36% dominant alleles. The F1 plants express all the dom-
inant alleles and, when backcrossed to the recessive
× b r /b r (Calif.) b r/b r 36
parent, give the following progeny phenotypes:
B r/b r 14
colored, green, starchy, smooth, tall 360
b R/b r 14
white, bronze, waxy, shrunk, dwarf 355
B R /B R (Chile) B R/b r B R/b r 36
× b r /b r (Chile) colored, bronze, waxy, shrunk, dwarf 40
b r/b r 36
white, green, starchy, smooth, tall 46
B r/b r 14
colored, green, starchy, smooth, dwarf 85
b R/b r 14
white, bronze, waxy, shrunk, tall 84
B R /B R (Calif.) B R/b r B R/b r 48
×b r /b r (Chile) or colored, bronze, waxy, shrunk, tall 8
b r/b r 48
b r /b r (Calif.) white, green, starchy, smooth, dwarf 9
B r/b r 2
× B R /B R (Chile) colored, green, waxy, smooth, tall 7
b R/b r 2
white, bronze, starchy, shrunk, dwarf 6
a. Provide a genetic hypothesis that explains the three
Propose a hypothesis to explain these results. Include
sets of testcross results.
a. a general statement of your hypothesis, with dia-
b. Draw the key chromosomal features of meiosis in
grams if necessary.
the F1 from a cross of the Californian and Chilean
lines. b. why there are 10 classes.
59. An aberrant corn plant gives the following RF values c. an account of the origin of each class, including its
when testcrossed: frequency.
d–f f–b b–x x–y y–p 61. Chromosomally normal corn plants have a p locus on
chromosome 1 and an s locus on chromosome 5.
Control 5 18 23 12 6
P gives dark green leaves; p, pale green leaves.
Aberrant plant 5 2 2 0 6
S gives large ears; s, shrunken ears.
(The locus order is centromere-d–f–b–x–y–p.) The An original plant of genotype P/p ; S/s has the expected
aberrant plant is a healthy plant, but it produces far phenotype (dark green, large ears) but gives unex-
fewer normal ovules and pollen than does the control pected results in crosses as follows:
plant.
•• On selfing, fertility is normal, but the frequency of
a. Propose a hypothesis to account for the abnormal p/p ; s/s types is 1/4 (not 1/16 as expected).
recombination values and the reduced fertility in
•• When crossed with a normal tester of genotype
the aberrant plant.
p/p ; s/s, the F1 progeny are 12 P/p ; S/s and 12 p/p ; s/s ;
b. Use diagrams to explain the origin of the recombi- fertility is normal.
nants according to your hypothesis.
•• When an F1 P/p ; S/s plant is crossed with a normal
60. The following corn loci are on one arm of chromo- p/p ; s/s tester, it proves to be semisterile, but, again,
some 9 in the order indicated (the distances between the progeny are; 12 P/p ; S/s and ; 12 p/p ; s/s.
them are shown in map units):
Explain these results, showing the full genotypes of the
c–bz–wx–sh–d–centromere original plant, the tester, and the F1 plants. How would
12 8 10 20 10 you test your hypothesis?
62. A male rat that is phenotypically normal shows repro- a. What types of linear asci (see Chapter 3) do you
ductive anomalies when compared with normal male predict, and in what frequencies, in a normal cross
rats, as shown in the table above. Propose a genetic of un-3 ad-3 ×wildtype? (Specify genotypes of
explanation of these unusual results, and indicate how spores in the asci.)
your idea could be tested. b. Most of the time such crosses behave predictably;
63. A tomato geneticist working on Fr, a dominant mutant but in one case, a standard un-3 ad-3 strain was
allele that causes rapid fruit ripening, decides to find crossed with a wild type isolated from a field of
out which chromosome contains this gene by using sugarcane in Hawaii. The results follow:
a set of lines of which each is trisomic for one chro-
mosome. To do so, she crosses a homozygous diploid
un-3 ad-3 un-3 ad-3
mutant with each of the wild-type trisomic lines.
un-3 ad-3 un-3 ad-3
a. A trisomic F 1 plant is crossed with a diploid
un-3 ad-3
wild-type plant. What is the ratio of fast- to slow-
ripening plants in the diploid progeny of this un-3 ad-3
abort
second cross if Fr is on the trisomic chromosome? un-3ad-3
Use diagrams to explain. un-3ad-3
b. What is the ratio of fast- to slow-ripening plants in un-3ad-3 un-3ad-3
the diploid progeny of this second cross if Fr is not un-3ad-3 un-3ad-3
located on the trisomic chromosome? Use diagrams
to explain.
(and its upside- (and other
c. Here are the results of the crosses. On which chro- down version) spore pair orders)
mosome is Fr, and why? 80% 20%
Trisomic Fast ripening : slow ripening in Explain these results, and state how you could test
chromosome diploid progeny your idea. (Note: In Neurospora, ascospores with
1 45:47 extra chromosomal material survive and are the nor-
mal black color, whereas ascospores lacking any chro-
2 33:34
mosome region are white and inviable.)
3 55:52
65. Two mutations in Neurospora, ad-3 and pan-2, are
4 26:30 located on chromosomes 1 and 6, respectively. An
5 31:32 unusual ad-3 line arises in the laboratory, giving the
6 37:41 results shown in the table below. Explain all three
results with the aid of clearly labeled diagrams. (Note:
7 44:79 In Neurospora, ascospores with extra chromosomal
8 49:53 material survive and are the normal black color,
9 34:34 whereas ascospores lacking any chromosome region
are white and inviable.)
10 37:39
RF between
(Problem 63 is from Tamara Western.) Ascospore ad-3 and
appearance pan-2
CHALLENGING PROBLEMS 1. Normal ad-3 All black 50%
64. The Neurospora un-3 locus is near the centromere ×normal pan-2
on chromosome 1, and crossovers between un-3 and 2. Abnormal ad-3 About 12 black and 1%
the centromere are very rare. The ad-3 locus is on the ×normal pan-2 1 white (inviable)
2
other side of the centromere of the same chromosome,
and crossovers occur between ad-3 and the centromere 3. Of the black spores from cross 2, about half were
in about 20 percent of meioses (no multiple crossovers completely normal and half repeated the same
occur). behavior as the original abnormal ad-3 strain.
10. Draw first- and second-division meiotic nondisjunc- a. Provide reasonable explanations for these three
tion in the female parent of cross 2, as well as in the exceptional octads.
resulting gametes. b. Diagram the meiosis that gave rise to octad 2.
11. Are any of the gametes from part 10 aneuploid?
71. The life cycle of the haploid fungus Ascobolus is sim-
12. Would you expect aneuploid gametes to give rise to ilar to that of Neurospora. A mutational treatment
viable progeny? Would these progeny be nullisomic, produced two mutant strains, 1 and 2, both of which
monosomic, disomic, or trisomic? when crossed with wild type gave unordered tetrads,
13. What progeny phenotypes would be produced by all of the following type (fawn is a light brown color;
the various gametes considered in parts 9 and 10? normally, crosses produce all black ascospores):
14. Consider the phenotypic ratio in the progeny of spore pair 1 black spore pair 3 fawn
cross 3. Many genetic ratios are based on halves spore pair 2 black spore pair 4 fawn
and quarters, but this ratio is based on thirds and a. What does this result show? Explain.
sixths. To what might this ratio point?
The two mutant strains were crossed. Most of the
15. Could there be any significance to the fact that the
unordered tetrads were of the following type:
crosses concern genes on a very small chromosome?
When is chromosome size relevant in genetics? spore pair 1 fawn spore pair 3 fawn
spore pair 2 fawn spore pair 4 fawn
16. Draw the progeny expected from cross 3 under
the two hypotheses, and give some idea of relative b. What does this result suggest? Explain.
proportions. When large numbers of unordered tetrads were
Now try to solve the problem. If you are unable to do so, screened under the microscope, some rare ones that
try to identify the obstacle and write a sentence or two contained black spores were found. Four cases are
describing your difficulty. Then go back to the expansion shown here:
questions and see if any of them relate to your difficulty. If
Case A Case B Case C Case D
this approach does not work, inspect the Learning Objec-
tives and Key Concepts of this chapter and ask yourself spore pair 1 black black black black
which might be relevant to your difficulty. spore pair 2 black fawn black abort
70. In the fungus Ascobolus (similar to Neurospora), asco- spore pair 3 fawn fawn abort fawn
spores are normally black. The mutation f, producing spore pair 4 fawn fawn abort fawn
fawn-colored ascospores, is in a gene just to the right
of the centromere on chromosome 6, whereas muta- (Note: Ascospores with extra genetic material survive,
tion b, producing beige ascospores, is in a gene just but those with less than a haploid genome abort.)
to the left of the same centromere. In a cross of fawn c. Propose reasonable genetic explanations for each of
and beige parents (+f × b+), most octads showed four these four rare cases.
fawn and four beige ascospores, but three rare excep-
tional octads were found, as shown in the accompa- d. Do you think the mutations in the two original
nying illustration. In the sketch, black is the wild-type mutant strains were in one single gene? Explain.
phenotype, a vertical line is fawn, a horizontal line is
beige, and an empty circle represents an aborted (dead) GENETICS AND SOCIETY
ascospore. Based on the data shown in Figures 17-15 and 17-37, at what
age would you recommend screening of expectant mothers
for fetal chromosome abnormalities? And, for which chro-
mosomal abnormalities? What aspects of an expectant moth-
er’s health history would influence this judgment?
Double mutants (b f )
1 2 3
Population Genetics
603
I
n 2009, Sean Hodgson was released from a British prison offspring and not all offspring survive. When geneticists
after serving 27 years behind bars for the 1979 mur- began developing the principles of population genetics, they
der of Teresa De Simone, a clerk and part-time barmaid. had rather limited tools to actually measure genetic vari-
Hodgson, who suffers from mental illness, initially confessed ation. With the development of DNA-based technologies
to the crime but withdrew his confession during the trial. over the past three decades, geneticists now have the ability
Throughout his years in prison, he maintained his innocence. to observe directly differences between the DNA sequences
More than two decades after the crime, the courts analyzed of individuals throughout their genomes, and they can
DNA of the assailant found at the crime scene and deter- measure these differences in large samples of individuals
mined that it did not come from Mr. Hodgson. Hodgson’s in many species. The result has been a revolution in our
conviction was overturned, and the police subsequently iden- understanding of genetic variation in populations.
tified David Lace as the likely murderer. Strangely, Lace had In this chapter, we will consider the concept of the gene
come forward in 1983 and confessed, but with Hodgson pool and how geneticists estimate allele and genotype fre-
already convicted, the police refused to believe Lace. In 1988, quencies in populations. Next, we will examine the impact
Lace committed suicide long before the DNA evidence impli- that mating systems have on the frequencies of genotypes
cated him. As you will learn in this chapter, the DNA-based in a population. We will also discuss how geneticists mea-
analysis used to exonerate Mr. Hodgson and hundreds of sure variation using DNA-based technologies. We will then
other wrongly convicted prisoners was dependent on popu- discuss the forces that modulate the levels of genetic varia-
lation genetic analysis. tion within populations. Finally, we will look at some case
The principles of population genetics are at the heart studies involving the application of population genetics to
of many questions facing society today. What are the risks questions of interest to society.
that a couple will have a child with a genetic disease? Have
the practices of plant and animal breeding caused a loss of
genetic diversity on the farm, and does this loss of diver-
sity place our food supply at risk? As the human popula- 18.1 DETECTING GENETIC
tion continues to expand and wildlife retreats into smaller VARIATION
and smaller parts of the earth, will wildlife species be able
to avoid inbreeding and survive? The principles of popula- LO 18.1 Describe and analyze data to determine how much
tion genetics are also fundamental to understanding many genetic variation exists within populations.
historical and evolutionary questions. How are human
populations from different regions of the world related The methods of population genetics can be used to analyze
to one another? How has the human genome responded any variable or polymorphic locus in the DNA sequences of
as humans have spread out across the globe and become a population of organisms. Over the past several decades,
adapted to different environments and lifestyles? How do multiple technologies, such as DNA sequencing, DNA
populations and species evolve over time? microarrays, and PCR (see Chapters 10 and 14), have been
A population is a group of individuals of the same spe- developed that allow geneticists to observe differences in
cies. Population genetics analyzes the amount and distribu- the DNA sequences among large samples of individuals.
tion of genetic variation in populations and the forces that In population genetics, a locus is simply a location in
control this variation. It has its roots in the early 1900s, the genome; it can be a single nucleotide site or a stretch of
when geneticists began to study how Mendel’s laws could many nucleotides. The simplest form of variation one might
be extended to understand genetic variation within whole observe among individuals at a locus is a difference in the
populations of organisms. While Mendel’s laws explain nucleotide present at a single nucleotide site, whether ade-
how genes are passed from parent to offspring in known nine, cytosine, guanine, or thymine. These types of variants
pedigrees, these laws are insufficient to understand the are called single nucleotide polymorphisms (SNPs), and
transmission of genes from one generation to the next in they are the most widely studied variants in human popu-
natural populations, in which not all individuals produce lation genetics (Figure 18-1; see also Chapter 4). Population
604
Nucleotide position
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Chromosomes from
seven individuals
1 G G C A T C G C G C C G T T A C G T A G A G A G A G A G G T G A A T C
2 G G C A A C G C G C C G T T A C G T A G A G A G A G A G G T G A A T C
3 G G G A T C G C G C C G T T A C G T A G A G A G A G A G G T G A A T C
4 G C C A T C G C T C C G T T A C T T A G A G A G - - - - G T T A G T C
5 G C C A T C G C T C - - - T A C T T A G A G A G - - - - G T T A G T C
6 G C C A T C G C T C - - - T A C T T A G A G A G - - - - G T T A G T C
7 G C C A T C G C T C C G T T A C T T A G A G A G - - - - C T T A G T C
* * * * * * * *
Indel Microsatellite
FIGURE 18-1 Variation in the aligned DNA sequences of seven chromosomes from different people.
The asterisks show the location of SNPs. The location of an indel (insertion/deletion of a string of
nucleotide pairs) and a microsatellite are also indicated.
genetics also makes extensive use of microsatellite loci (see partial genome sequences of these 48 individuals with one
Chapter 4). These loci have a short sequence motif, 2 to 6 another. This initial effort led to the discovery of more than
base-pairs long, that is repeated multiple times with differ- 1 million SNPs.
ent alleles having different numbers of repeats. For example, Once SNPs have been discovered, the genotype (allelic
the 2-bp-sequence motif AG at a locus might be tandemly composition) of different individuals in the population
repeated five times in one allele (AGAGAGAGAG) but three at each SNP can be determined. DNA microarrays are a
times in another (AGAGAG) (see Figure 18-1). widely used technology for this purpose (Figure 18-2). The
microarrays used for SNP assays can contain thousands
of probes corresponding to known SNPs. Biotechnologists
Single nucleotide polymorphisms (SNPs)
have developed several different methods to detect SNP
SNPs are the most prevalent types of polymorphism in variants using microarrays. In one method, DNA from
most genomes. Most SNPs have just two alleles—for exam-
ple, A and C. SNPs are usually considered common SNPs
in a population if the less common allele occurs at a fre-
quency of about 5 percent or greater. SNPs for which the A microarray is used to
less common allele occurs at a frequency below 5 percent detect variation in SNPs
are considered rare SNPs. For humans, there is a common
SNP about every 300 to 1000 bp in the genome. Of course,
there are a far greater number of rare SNPs.
SNPs occur within genes, including within exons,
introns, and regulatory regions. SNPs within protein-coding
regions can be classified into one of three groups: synony-
mous if the different alleles encode the same amino acid,
nonsynonymous if the two alleles encode different amino
acids, and nonsense if one allele encodes a stop codon and
the other an amino acid. Thus, it is sometimes possible to
associate a SNP with functional variation in proteins and
an associated change in phenotype. SNPs located outside
of coding sequences are called noncoding SNPs (ncSNPs).
If ncSNPs have no effect on gene function and phenotype,
Introduction to Genetic Analysis, 11e
they
Figureare called
18.01 silent.
#1807
To study SNP variation in a population, one can
07/08/14
first determine
Dragonfly which nucleotide sites in the genome are
Media Group
variable—that is, constitute a SNP. This first step is called
SNP discovery. SNPs are often discovered by sequencing the
genomes of a small sample of individuals of a species, then
comparing these sequences. For example, SNP discovery
FIGURE 18-2 Detecting variation in DNA: SNPs. View of a small
in humans began by partially sequencing the genomes of a portion of a microarray used to scan a single individual’s genome.
discovery panel of 48 individuals from around the world. Each dot represents one SNP, with red and green for the homozygous
Variable nucleotide sites were discovered by comparing the classes and yellow for heterozygous.
an individual is labeled with fluorescent tags and hybrid- located outside of coding sequences, and variation in the
ized to the microarray. Each spot (SNP) on the microar- number of repeats is not associated with differences in
ray will fluoresce red for one homozygous class, green for phenotype.
the other homozygote, and yellow for a heterozygote (see Once a microsatellite and its flanking sequences have
Figure 18-2). The entire procedure has been enhanced with been identified, DNA samples from a set of individuals in
robotics to allow rapid genotyping, or assignment of geno- the population can be analyzed to determine the number of
types (for example, A/A versus A/C) on a large-scale basis. repeats that are present in each individual. To carry out the
Direct-to-consumer genotyping services like 23andMe use analysis, oligonucleotide primers are designed that match
SNP microarrays, as do some large-scale medical research the flanking sequences for use in PCR. If the primers are
projects. labeled with a fluorescent tag, then the sizes of the PCR
More recently with the dramatic reduction in the products can be determined on the same apparatus used
cost of DNA sequencing, it has become possible to col- to determine the sequence of DNA molecules (Figure 18-3).
lect the genome sequences of many individuals in a spe- These sizes reveal the number of repeats in a microsatel-
cies and then compare the SNP difference between the lite allele. For example, the PCR product of a microsatellite
multiple genome sequences. For many questions in pop- allele containing seven AG repeats will be 8 bp longer than
ulation genetics, a relatively small sample of genome an allele containing three AG repeats. Heterozygous indi-
sequences of 20 to 100 individuals is adequate. In other viduals will possess products of two different sizes. Since
cases, thousands or even tens of thousands of genome PCR, the sizing of PCR products, and scoring of the alleles
sequences of individuals are being gathered. Such large can all be automated, it is possible to determine the geno-
samples are used when researchers are attempting to types of large samples of individuals for large numbers of
identify regions of the genome that control a phenotype microsatellites relatively rapidly.
such as the risk of having a disease, like type 2 diabetes.
This type of analysis, called a genome-wide association
study, is covered in Chapter 19. For our own species, Haplotypes
the first large-scale multiple genome project was called For some questions in population genetics, it is important to
the 1000 Genomes Project, and it collected that many consider the genotypes of linked loci as a group rather than
genome sequences of people from around the world individually. Geneticists use the term haplotype to refer to
(www.1000genomes.org). the combination of alleles at multiple loci on the same chro-
mosomal homolog. Two homologous chromosomes that
share the same allele at each of the loci under consideration
Microsatellites have the same haplotype. If two chromosomes have differ-
Microsatellites are powerful loci for population genetic ent genotypes at even one of the loci in question, then they
analysis for several reasons. First, unlike SNPs, which typ-
ically have only two alleles per locus and can never have
more than four alleles, the number of alleles at a micro Detecting variation in microsatellites
satellite is often very large (20 or more). Second, they
have a high mutation rate, typically in the range of 10−3 to Individuals
1 2 3
10−4 mutations per locus per generation as compared to
10−8 to 10−9 mutations per site per generation for SNPs.
The high mutation rate means that levels of variation
Locus 1
are higher: more alleles per locus and a greater chance
that any two individuals will have different genotypes.
Third, microsatellites are very abundant in most genomes.
Humans have over a million microsatellites. Discovering Locus 2
Migration
have different haplotypes. If the A locus with alleles A and (I-a, I-b, . . .). The haplotype network shows the relation-
a is linked to the B locus with alleles B and b, then there are ships among the haplotypes, placing each mutation on one
four possible haplotypes for the chromosomal segment on of the branches (Figure 18-4b).
which these two loci are located: What insights can we gain from haplotype analysis?
Population geneticists studying the human Y chromosome
A B
among Asian men discovered one highly prevalent haplo-
A b type, termed the “star-cluster” haplotype (Figure 18-5a).
a B Typically, most men have a rare Y chromosome haplotype,
a b but the “star-cluster” haplotype is present in 8 percent of
Asian men. Using the known mutation rate, the research-
A more complex, but more realistic, example is shown in ers estimated that this common haplotype arose between
Figure 18-4. In Figure 18-4a, there are seven chromosome 700 and 1300 years ago. (Later in this chapter, we will dis-
segments but only six haplotypes because chromosome seg- cuss mutation rates and their use in population genetics.)
ments 5 and 6 have the same haplotype (E). This haplotype is most common in Mongolia, suggesting
Haplotypes are most often used in population genet- that it arose there. The researchers inferred that the “star-
ics for loci that are physically close. For example, the cluster” haplotype traces back to one man in Mongo-
v ariable-nucleotide sites in a single gene can be used to lia about 1000 years ago. Remarkably, the present-day
define haplotypes for that gene. However, the haplotype distribution of this haplotype follows the geographic
concept works for larger regions when there is little or no boundaries of the M ongolian Empire established by
recombination over the region. It can even be applied to the Genghis Khan about 1200 years ago (Figure 18-5b). It
human Y chromosome, most of which does not experience appears that contemporary men with this haplotype are
recombination with the X chromosome. Finally, it is some- all descendants of Genghis Khan (or his male-lineage
times useful to group haplotypes into classes. As shown in relatives).
Figure 18-4a, there are two major classes of haplotypes Y-chromosome star clusters, like the one linked to Geng-
(I and II) that differ at five nucleotide sites plus a micro- his Khan, are a common feature in the genetic history of our
satellite. However, each class contains several subtypes species. Notably, there are multiple star clusters dating to
Haplotype
Haplotype
(a) Haplotypes
Nucleotide position
class
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
Chromosomes from
1 G G C A T C G C G C C G T T A C G T A G A G A G A G A G G T G A A T C A I-a
seven individuals
2 G G C A A C G C G C C G T T A C G T A G A G A G A G A G G T G A A T C B I-b
3 G G G A T C G C G C C G T T A C G T A G A G A G A G A G G T G A A T C C I-c
4 G C C A T C G C T C C G T T A C T T A G A G A G - - - - G T T A G T C D II-a
5 G C C A T C G C T C - - - T A C T T A G A G A G - - - - G T T A G T C E II-b
6 G C C A T C G C T C - - - T A C T T A G A G A G - - - - G T T A G T C E II-b
7 G C C A T C G C T C C G T T A C T T A G A G A G - - - - C T T A G T C F II-c
* * * * * * * *
Indel Microsatellite
C F
SN
P3
P2
SN
A D
SN
l
de
P5
In
Microsatellite
SNP2
SNP9
SNP17
SNP31
SNP33
B
E
FIGURE 18-4 (a) There are a total of six haplotypes (A–F) in the aligned DNA sequences from seven
individual chromosomes from different people. (b) These six haplotypes are joined in a haplotype
network showing the relationships among the haplotypes. Each circle represents one of the six
haplotypes. Any two haplotypes differ at the loci noted on all of the branches connecting them.
The asterisks show the location of SNPs.
(b)
RUSSIA
Ewenki Han
Ossetian MONGOLIA Daur (heilongjiang)
Georgian
Han
Aral (Inner-
Lezgi Sea Mongolian) Chinese
Korean
Armenian Azeri TURKMENISTAN Inner
Kazak Uyghur Mongolian
Japanese
Chinese Mongolian
Uzbek Uyghur Manchu
Korean
Kyrgyz Xibe
Kurd Tajik
Turkmen Chinese Huizu
Kazak
Han
Kalash xingjiang
Han
AFGHANISTAN Balti (Gansu)
CHINA PACIFIC OCEAN
IRAN Hazara
Brahui Baloch
Burusho Shezu
Tibetan Han
Qiangzu (Sichuan)
PAKISTAN
Makrani
Baloch Makrani
n>60
Buyi Yaozu
Negroid NEPAL (Liannan)
Parsi
BHUTAN Han
Hani Yaozu (Guangdong)
(Bamma)
BANGLADESH Lizu
INDIA n=30
Star-cluster
chromosomes
the time when agriculture was invented, suggest that this microsatellites, indels do not contain repeat motifs such as
cultural innovation is associated with differential reproduc- AGAGAGAG.
tive success of the men of that time. Thus far, our discussion of SNP and microsatellites has
focused on the nuclear genome. However, interesting genetic
Other sources and forms of variation variation can also be found in the mitochondrial (mtDNA)
Beyond SNP and microsatellites, any variation in the DNA and chloroplast (cpDNA) genomes of eukaryotes. Both SNP
Introductionoftothe
sequence Genetic Analysis, 11ein a population is amena-
chromosomes and microsatellites are found in these organelle genomes.
Figure 18.05 #1810
ble to population genetic analysis. Variations that can be Since mtDNA and cpDNA are usually maternally inherited,
07/08/14
analyzed
07/23/14 include inversions, translocations, deletions or their analysis can be used to follow the history of female
duplications,
Dragonfly Mediaand the presence or absence of a transpos-
Group lineages. In 1987, a prominent study of the human mito-
able element at a particular locus in the genome. Another chondrial lineage traced the history of the human mtDNA
common form of variation is insertion-deletion polymor- haplotypes and determined that the mitochondrial genomes
phism, or indel for short (see Chapter 15). This type of of all modern humans trace back to a single woman who
polymorphism involves the presence or absence of one lived in Africa about 150,000 years ago (Figure 18-6). She
or more nucleotides at a locus in one allele relative to was dubbed the “mitochondrial Eve” in the popular press.
another. In Figure 18-1, chromosome segments 5 and 6 This study of mtDNA was the first thorough genetic anal-
differ from the other five segments by a 3-bp indel. Unlike ysis to suggest that all modern humans came from Africa.
C, D, G
A
H, J, T, U, Uk, V C, D
A
I, W, X
B
B
N
F
B
L3
L2
M
L1
L0
A
C, D
Q
S B
P
FIGURE 18-6 The haplotype network for human mtDNA haplotype groups drawn onto a world
map. The ancestral L haplotype group appears in Africa, and the derived groups (A, B, and so on) are
dispersed throughout the world. This haplotype network is like the one shown in Figure 18-4, except
here the SNPs are not labeled on the branches. [Data from www.mitomap.org.]
KEY CONCEPT Genomes are replete with diverse types of The gene pool is the sum total of alleles in a population
variation suitable for population genetic analysis. SNPs and
microsatellites are the two most commonly studied types of
polymorphism in population genetics. High-throughput tech- AA AA Aa
nologies allow hundreds of thousands of polymorphisms to
be scored in tens of thousands of individuals.
Aa AA AA Aa Aa
Genotypes AA Aa aa
Perhaps you have watched someone performing a death- Number 5 8 3
defying stunt and thought that they were at risk of eliminat- Alleles A a
ing themselves from the “gene pool.” If so, you were using Number 18 14
a concept, the gene pool, that comes straight out of popula-
tion genetics and has worked its way into popular culture. FIGURE 18-7 A frog gene pool.
The gene-pool concept is a basic tool for thinking about
genetic variation in populations. We can define the gene shows a population of 16 frogs, each of which carries two
pool as the sum total of all alleles in the breeding members alleles at the autosomal locus A. By simple counting, we
of a population at a given time. For example, Figure 18-7 can determine that there are five A /A homozygotes, eight
A /a heterozygotes, and three a /a homozygotes. The size of this section, we will begin to look at how this works. We
the population, usually symbolized by the letter N, is 16, will see how we can use the allele frequencies in the gene
and there are 32 or 2N alleles in this diploid population. pool to make predictions about the genotype frequencies in
With this simple set of numbers, we have described the gene the next generation.
pool with regard to the A locus. The frequency of an allele in the gene pool is equal to
Typically, population geneticists do not care about the the probability that the allele will be chosen when randomly
absolute counts of the different genotypes in a population, picking an allele from the gene pool to form an egg or a
but about the genotype frequencies. We can calculate the sperm. Knowing this, we can calculate the probability that
frequency of the A /A genotype simply by dividing the num- a frog in the next generation will be an A /A homozygote. If
ber of A /A individuals by the total number of individuals in we reach into the frog gene pool (see Figure 18-7) and pick
the population (N) to get 0.31. The frequency of A /a het- the first allele, the probability that it will be an A is p = 0.56,
erozygotes is 0.50, and the frequency of a /a homozygotes is and similarly, the probability that the second allele we pick
0.19. Since these are frequencies, they sum to 1.0. Frequen- is also an A is p = 0.56 . The product of these two proba-
cies are a more practical measurement than absolute counts bilities, or p2 = 0.3136, is the probability that a frog in the
because rarely are population geneticists able to study every next generation will be A /A . The probability that a frog in
individual in a population. Rather, population geneticists the next generation will be a /a is q 2 = 0.44 × 0.44 = 0.1936
will draw a random or unbiased sample of individuals from . There are two ways to make a heterozygote. We might first
a population and use the sample to infer the genotype fre- pick an A with probability p and then pick an a with proba-
quencies in the entire population. bility q, or we might pick the a first and the A second. Thus,
We can make a simpler description of this frog gene pool the probability that a frog in the next generation will be
if we calculate the allele frequencies rather than the genotype heterozygous A /a is pq + qp = 2pq = 0.4928. Overall, the
frequencies (Box 18-1). In Figure 18-7, 18 of the 32 alleles are frequencies (f ) of the genotypes are
A, so the frequency of A is 18/32 = 0.56. The frequency of the
f A /A = p2
A allele is typically symbolized by the letter p, and in this case,
p = 0.56 . The frequency of the a allele is symbolized by the f a /a = q 2
letter q, and in this case, q = 14/32 = 0.44 . Again, since these f A /a = 2pq
are frequencies, they sum to 1.0 : p + q = 0.56 + 0.44 = 1.0.
Finally, as expected, the sum of the probability of being
We now have a description of our frog gene pool using only
A /A plus the probability of being A /a plus the probability
two numbers, p and q.
of being a /a is 1.0:
KEY CONCEPT The gene pool is a fundamental concept p2 + 2pq + q 2 = 1.0
for the study of genetic variation in populations: it is the sum
total of all alleles in the breeding members of a population at This simple equation is the Hardy–Weinberg law, and it is
a given time. We can describe the variation in a population in part of the foundation for the theory of population genetics.
terms of genotype and allele frequencies. The process of reaching into the gene pool to pick an
allele is called sampling the gene pool. Since any individual
As mentioned previously, an important goal of popu- that contributes to the gene pool can produce many eggs
lation genetics is to understand the transmission of alleles or sperm that carry exactly the same copy of an allele, it is
from one generation to the next in natural populations. In possible to pick a particular copy and then reach back into
q = fa /a + 21 fA /a = frequency of a
the gene pool and pick exactly the same copy again. There The latter number predicts that about 6 percent of this pop-
is also an element of chance involved when sampling the ulation are heterozygotes, or carriers of the recessive allele
gene pool. Just by chance, some copies may be picked at OCA2.
more than once and other copies may not be picked at all. When we use the Hardy–Weinberg law to calculate allele or
Later in the chapter, we will look at how these properties genotype frequencies, we make some critical assumptions.
of sampling the gene pool can lead to changes in the gene
• First, we assume that mating is random in the population
pool over time.
with respect to the gene in question. Deviation from ran-
We used the Hardy–Weinberg law to calculate genotype
dom mating violates this assumption, making it inappro-
frequencies in the next generation from the allele frequencies
priate to apply Hardy–Weinberg. For example, a tendency
in the current generation. We can also use the Hardy–Wein-
for individuals who are phenotypically similar to mate
berg law to calculate allele frequencies from the genotype
with each other violates the Hardy–Weinberg law. If albi-
frequencies within a single generation. For example, some
nos mated more frequently with other albinos than with
forms of albinism in humans are due to recessive alleles at the
non-albinos, then the Hardy–Weinberg law would overes-
OCA2 locus. In Africa, a form of albinism called brown ocu-
timate the frequency of the recessive allele.
locutaneous albinism results from a recessive allele of OCA2
(Figure 18-8). Individuals with this condition, who have two • Second, if one of the genotypes has reduced viability such
recessive alleles, are present at frequencies as high as 1 in that some individuals with that genotype die before the
1100 among some ethnic groups in Africa. We can use the genotype frequencies are counted, then the estimate of the
Hardy–Weinberg law to calculate the allele frequencies: gene frequencies will be inaccurate.
• Third, for the Hardy–Weinberg law to apply, the popula-
f a /a = q 2 = 1/1100 = 0.0009
tion must not be divided into subpopulations that are par-
so tially or fully genetically isolated. If there are separate
subpopulations, alleles may be present at different fre-
q = 0.0009 = 0.03
quencies in the different subpopulations. If so, using geno-
and typic counts from the overall population may not give an
accurate estimate of the overall allele frequencies.
p = 1 − q = 0.97
• Finally, the Hardy–Weinberg law strictly applies only to
Using the allele frequencies, we can also calculate the fre- infinitely large populations. For finite populations, there
quency of heterozygotes in the population as will be deviations from the frequencies predicted by the
2pq = 2 × 0.97 × 0.03 = 0.06 Hardy–Weinberg law due to chance when sampling the
gene pool to produce the next generation.
Genotype Gene
frequencies frequencies
Generation A /A A /a a /a A a
t0 0.64 0.32 0.04 0.8 0.2
FIGURE 18-8 Individual of African ancestry with brown oculocutaneous
t1 0.64 0.32 0.04 0.8 0.2
albinism (BOCA), a condition defined by light tan skin and beige to
light brown hair. [Dr. Michele Ramsay Department of Human Genetics,
School of Pathology, the National Health Laboratory Service University of
tn 0.64 0.32 0.04 0.8 0.2
Witwatersrand.]
Here are a few more points about the Hardy–Weinberg law. Male pattern baldness
1. For any allele that exists at a very low frequency, homo-
zygous individuals will only very rarely be found. If
allele a has a frequency of 1 in a thousand (q = 0.001),
then only 1 in a million (q 2 ) individuals will be
homozygous for that allele. As a consequence, recessive
alleles for genetic disorders can occur in the hetero-
zygous state in many more individuals than there are
individuals that actually express the genetic disorder in
question.
2. The Hardy–Weinberg law still applies where there are
more than two alleles per locus. If there are n alleles,
A1 , A2 , . . . An with frequencies p1 , p2 , . . . pn , then the
sum of all the individual frequencies equals 1.0. The
frequencies of each of the homozygous genotypes are
simply the square of the frequencies of the alleles, and FIGURE 18-9 Individual showing male
the frequencies of the different heterozygous classes are pattern baldness, an X-chromosome-linked
two times the product of the frequencies of the first condition. [B2M Productions/Getty Images.]
and second allele. Table 18-1 gives an example with
p1 = 0.5, p2 = 0.3, and p3 = 0.2 . 4. One can test whether the observed genotype frequen-
3. Hardy–Weinberg logic applies to X-linked loci as well. cies at a locus fit Hardy–Weinberg predictions using the
Males are hemizygous for X-linked genes, meaning chi-square test (see Chapter 3). An example is provided
that a male has a single copy of these genes. Thus, for by the human leukocyte antigen gene, HLA-DQA1, of
X-linked genes in males, the genotype frequencies are the major histocompatibility complex (MHC). MHC is
equal to the allele frequencies. For females, genotype a cluster of genes on chromosome 6 that play roles in
frequencies for X-linked genes follow normal Hardy– the immune system. Table 18-2 has genotype frequen-
Weinberg expectations. cies for an SNP (rs9272426) in the HLA-DQA1 for
Male pattern baldness is an X-linked trait 84 residents of Tuscany, Italy. This SNP has alleles A
(Figure 18-9). AR (for androgen receptor) is an X-linked and G. From the genotype frequencies in Table 18-2,
gene involved in male development. There is an AR we can calculate the allele frequencies: f (A) = p = 0.53
haplotype called Eur-H1 that is strongly associated and f (G) = q = 0.47. Next, we can calculate expected
with pattern baldness. Male pattern baldness is com- genotype frequencies under the Hardy–Weinberg law:
mon in Europe, where the Eur-H1 haplotype occurs at p2 = 0.281, 2pq = 0.498, and q 2 = 0.221. Multiplying
a frequency of 0.71, meaning that 71 percent of Euro- the expected genotype frequencies times the sample size
pean men carry it. Using the Hardy–Weinberg law, we (N = 84) gives us the expected number of individuals
can calculate that 50 percent of European women are for each genotype. Now we can calculate the χ2 statistic
Eur-H1 homozygotes and 41 percent are heterozygous. to be 8.29. Using Table 3-1, we see that the probabil-
The inheritance of baldness is complex and is affected ity under the null hypothesis that the observed data fit
by multiple genes, and so not all men who have Eur-H1 Hardy–Weinberg predictions is P < 0.005 with df = 1.
go bald. [We have only one degree of freedom because we have
three genotypic categories and we used two numbers
from the data (N and p) to calculate the expected values
TABLE 18-1 Hardy–Weinberg Genotype Frequencies (3 − 2 leaves 1 degree of freedom). We did not need to
for a Locus with Three Alleles A1, A2, and use q since q = p − 1.] This analysis makes us strongly
A3 with Frequencies 0.5, 0.3, and 0.2, suspect that Tuscans do not conform to Hardy–
Respectively Weinberg expectations with regard to HLA-DQA1.
Genotype Expectation Frequency The Hardy–Weinberg law is part of the foundation of
A1A1 p12 0.25 population genetics. It applies to an idealized population
that is infinite in size and in which mating is random. It
A2 A2 p22 0.09
also assumes that all genotypes are equally fit—that is,
A3 A3 p32 0.04 that they are all equally viable and have the same suc-
A1A2 2 p1p2 0.30 cess at reproduction. Real populations deviate from this
A1A3 2 p1p3 0.20 idealized one. In the rest of the chapter, we will examine
how factors such as nonrandom mating, finite population
A2 A3 2 p2 p3 0.12
size, and the unequal fitness of different genotypes cause
Sum 1.00
deviations from Hardy–Weinberg expectations.
TABLE 18-2 Frequencies of SNP rs9272426 Genotypes in HLA-DQA1 of the MHC Locus for
People from Tuscany, Italy
Genotypes
A/A A/G G/G Sum
Observed number 17 55 12 84
Observed frequency 0.202 0.655 0.143 1
Expected frequency 0.281 0.498 0.221 1
Expected number 23.574 41.851 18.574 84
(Observed–expected)2 /expected 1.833 4.131 2.327 8.29
Source: International HapMap Project.
KEY CONCEPT The Hardy–Weinberg law describes the a plant will not be receptive to pollen that carries either of
relationship between allele and genotype frequencies. This its own two alleles (Figure 18-10). For example, the stigma
law informs us that genetic variation is neither created nor of an S1 /S2 heterozygote will not allow pollen grains carrying
destroyed by the process of transmitting genes from one gen- either an S1 or S2 allele to germinate and fertilize its ovules,
eration to the next. The Hardy–Weinberg law strictly applies although pollen grains carrying the S3 or S4 alleles can do so.
only in infinitely large and randomly mating populations. This mechanism blocks self-fertilization, thereby enforcing
cross-pollination. The S locus violates the Hardy–Weinberg
law since homozygous genotypes at S are not formed.
18.3 MATING SYSTEMS A second example of negative assortative mating is pro-
vided by the major histocompatibility complex (MHC),
LO 18.3 Quantify the effect of inbreeding in a population.
which is known to influence mate choice in vertebrates. MHC
affects body odor in mice and rats, providing a basis for mate
Random mating is a critical assumption of the Hardy– choice. In what are known as the “sweaty T-shirt experiments,”
Weinberg law. The assumption of random mating is met if researchers asked a group of men to wear T-shirts for two days.
all individuals in the population are equally likely as a choice Then they asked a group of women to smell the T-shirts and
when a mate is chosen. However, if a relative, a neighbor, rate them for “pleasantness.” Women preferred the scent of men
or a phenotypically similar individual is a more likely mate whose MHC haplotypes were different from their own. Data
than a random individual, then the assumption of random
mating has been violated. Populations that are not random Self-incompatibility leads to disassortative mating
mating will not exhibit exact Hardy–Weinberg proportions in Brassica
for the genotypes at some or all genes. Three types of bias in (a) Pollen inhibition
mate choice that violate the assumption of random mating
are assortative mating, isolation by distance, and inbreeding.
Assortative mating
Assortative mating occurs if individuals choose mates based
on resemblance or non-resemblance to themselves. Posi-
tive assortative mating occurs when similar types mate; for
example, if tall individuals preferentially mate with other
tall individuals and short individuals mate with other short (b) Pollen-tube growth
individuals. In these cases, genes controlling the difference
in height will not follow the Hardy–Weinberg law. Rather,
we would expect to see an excess of homozygotes for the
“tall” alleles among the progeny of tall mating pairs and an
excess of homozygotes for “short” alleles among the prog-
eny of short mating pairs. In humans, there is positive assor-
tative mating for height.
Negative assortative or disassortative mating occurs
when unlike individuals mate—that is, when opposites
FIGURE 18-10 Disassortative mating caused by the self-
attract. One example of negative assortative mating is pro-
incompatibility locus (S) of the flowering plant genus Brassica. (a) A
vided by the self-incompatibility, or S, locus in plants such self-pollinated S1 /S2 stigma shows no pollen-tube growth. (b) There
as Brassica (broccoli and its relatives). There are numerous is pollen-tube growth for an S1 /S2 stigma cross-pollinated with pollen
alleles at the S locus, S1 , S2 , S3 , and so forth. The stigma of from an S3 /S4 heterozygote. [June Bowman Nasrallah.]
from the human HapMap project have since confirmed that Allele frequency may vary along a gradient
American couples are significantly more heterozygous at the
MHC than expected by chance. The MHC plays a central role (a)
in our immune response to pathogens, and heterozygotes may Frequency of A
be more resistant to pathogens. Therefore, our offspring bene- 1.0
0.5 Kansas City
fit if we mate disassortatively with respect to our MHC geno- 0.0
type. This mechanism may explain why the SNP in the MHC
gene HLA-DQA1 that we discussed earlier does not follow the Hutchinson
Elkhart
Hardy–Weinberg law among residents of Tuscany. Look back
at Table 18-2 and you will notice that there are more heterozy-
(b)
gotes than expected, 55 versus 42. Tuscans appear to be prac-
ticing disassortative mating with respect to this SNP.
Isolation by distance
Another form of bias in mate choice arises from the amount
of geographic distance between individuals. Individuals are
more apt to mate with a neighbor than another member of
their species on the opposite side of the continent—that is,
individuals can show isolation by distance. As a consequence, Frequency of FY null
allele and genotype frequencies often differ between fish in 10–50
separate lakes or between pine trees in different regions of a 50–70
70–75
continent. Species or populations exhibiting such patterning 75–80
of genetic variation are said to show population structure. A 80–85
species can be divided into a series of subpopulations such as 85–90
frogs in different ponds or people in different cities. 90–95
95–100
If a species has population structure, the proportion of
homozygotes will be greater species-wide than expected
FIGURE 18-11 (a) Allele frequency variation across Kansas for a
under the Hardy–Weinberg law. Consider a hypothetical
hypothetical species of wild sunflower. (b) Frequency variation for the
example of a species of wild sunflowers distributed across FY null allele of the Duffy blood group locus in Africa. [Data from P. C.
Kansas with a gradient in the frequency of the A allele from Sabeti et al., Science 312, 2006, 1614–1620.]
0.9 near Kansas City to 0.1 near Elkhart (Figure 18-11a). We
sample 100 sunflower plants from each of these two cities
plus 100 from Hutchinson, in the middle of the state, and This allele is rare outside of Africa. Because of this gradi-
we calculate allele frequencies. Each city represents a sub- ent, we cannot use overall allele frequencies in Africa to
population of the sunflowers. For any of the three cities, the calculate genotype frequencies using the Hardy–Weinberg
Hardy–Weinberg law works fine. For example, in Elkhart, law. Later in the chapter and in Chapter 20, we will discuss
we expect Nq 2 = 100 × (0.9)2 = 81 a /a homozygotes, and the relationship between Fy null and malaria.
that is what we observe. However, statewide, we would
predict Nq 2 = 300 × (0.5)2 = 75 a /a homozygotes, yet we KEY CONCEPT Assortative mating and isolation by distance
observed 107. Because of population structure, there are violate the Hardy–Weinberg law and can cause genotype fre-
more homozygous sunflower plants than expected. quencies to deviate from Hardy–Weinberg expectations.
Number of individuals
Inbreeding
N A /A A /a a /a p q
The third type of bias in mating is inbreeding, or mating
Kansas City 100 81 18 1 0.90 0.10
between relatives. Long before anyone knew about del-
Hutchinson 100 25 50 25 0.50 0.50 eterious recessive alleles, some societies recognized that
Elkhart 100 1 18 81 0.10 0.90 disorders such as muteness, deafness, and blindness were
Statewide 300 107 86 107 0.50 0.50 more frequent among the children of marriages between
(observed) relatives. Accordingly, brother–sister and first-cousin mar-
Statewide 300 75 150 75 – – riages were either outlawed or discouraged. Nevertheless,
(expected) many famous individuals have married a cousin, including
Charles Darwin, Albert Einstein, J. S. Bach, Edgar Allan
Here is a real example of population structure from our Poe, Jesse James, and Queen Victoria. As we will see, the
own species. In Africa, the FY null allele of the Duffy blood offspring of marriages between relatives are at higher risk
group shows a gradient with a low frequency in eastern and of having an inherited disorder.
northern Africa, moderate frequency in southern Africa, Progeny of inbreeding are more likely to be homozygous
and high frequency across central Africa (Figure 18-11b). at any locus than progeny of non-inbred matings. Thus, they
are more likely to be homozygous for deleterious recessive Pedigrees show when genes are identical by descent
alleles. For this reason, inbreeding can lead to a reduction in
(a) A
vigor and reproductive success called inbreeding depression.
However, inbreeding can have advantages, too. Many plant
species are highly self-pollinating and highly inbred. These
include the cereal crops rice and wheat, and the model plant
Arabidopsis, a successful weed. Since most plant species bear
male and female organs on the same individual, self-polli-
nation can be accomplished more easily than outcrossing. B C
Another advantage of self-pollination is that when a single
seed is dispersed to a new location, the plant that grows from
the seed has a ready mate—itself, enabling a new population
to be established from a single seed. Finally, if an individual
plant has a beneficial combination of alleles at different loci,
then inbreeding preserves that combination. In selfing plant I
species, benefits such as these offer advantages that outweigh
(b)
the cost associated with inbreeding depression.
A
B C
The inbreeding coefficient
Inbreeding increases the risk that an individual will be
w x
homozygous for a recessive deleterious allele and exhibit
a genetic disease. The amount that risk increases depends I
on two factors: (1) the frequency of the deleterious allele
in the population and (2) the degree of inbreeding. To mea-
FIGURE 18-12 (a) Pedigree for a half-sib mating drawn in the standard
sure the degree of inbreeding, geneticists use the inbreeding format. Small colored balls represent a single copy of a gene. Within
coefficient (F), which is the probability that two alleles in individual A, the pink and blue copies represent the copies of the gene
an individual trace back to the same copy in a common that she inherited from her mother and father, respectively. (b) Pedigree
ancestor. Let’s first consider how to calculate F using pedi- for a half-sib mating drawn in the simplified format used for the analysis
grees and then examine how F can be used to determine the of inbreeding. Only lines connecting parent to offspring are drawn, and
only individuals in the “closed inbreeding loop” are included. w, x, y, and
increase in risk of inheriting a recessive disease condition. z are symbols for the allele transmitted from parent to offspring.
Consider a simple pedigree for a mating between
half-sibs, individuals who have one parent in common
(Figure 18-12a). In the figure, B and C are half-sibs who have y, and z. We use “~” to symbolize IBD. We would like to
the same mother, A, but different fathers; B and C have a calculate the probability that w and x are IBD, but let’s take
daughter, I. Notice that there is a closed loop from I through this calculation step by step. First, what is the probability
B and A and back to I through C. The presence of a closed that x and y are IBD or, symbolically, what is P(x ~ y)? This
loop in the pedigree informs us that I is inbred. The two is the probability that C transmits the copy inherited from
copies of the gene in A are colored blue and pink—the blue A to I, which is 1/2, or P(x ~ y) = 1/2 . Similarly, the proba-
from A’s father and pink from her mother. As drawn, I has bility that B transmits the copy inherited from A to I is 1/2,
inherited the pink copy both through her father (B) and her or P(w ~ z) = 1/2.
mother (C). Since I’s two copies of the gene trace back to the Now we need to calculate the probability that z and y
same copy in her grandmother, her two copies are identical are IBD. There are two ways that z and y can be IBD. The
by descent (IBD). More generally, if the two copies of a gene first way is when z and y are both the same copy (both pink
in an individual trace back to the same copy in an ancestor, or both blue). This happens 1/2 of the time, since 1/4 of
then the copies are IBD. We would like a way to calculate the the time they are both blue and 1/4 of the time both pink.
probability that I’s two alleles will be IBD. This probability The second way is when z and y are different copies (one
is the inbreeding coefficient for I, which in symbol form is FI . pink and the other blue), but individual A was inbred. If
First, since we are interested only in tracing the path of individual A is inbred, then there is a probability that her
IBD alleles, we can simplify the pedigree to contain only the two copies of the gene are IBD. The probability that A’s two
individuals in the closed loop and still follow the transmis- copies are IBD is the inbreeding coefficient of A, FA . The
sion of any IBD alleles (Figure 18-12b). Also, since the sex of probability that z and y are different copies (one pink, the
the individual does not matter, we use circles for both sexes. other blue) is 1/2. So, the probability that z and y are differ-
The alleles transmitted with each mating are labeled w, x, ent copies that are IBD is 1/2 multiplied by the inbreeding
coefficient (FA ) to give 12 FA . Altogether, the probability that violated. However, Hardy–Weinberg can be modified to cor-
z and y are IBD is the probability that they are the same rect the predicted genotypic proportions for different degrees
copy (1/2) plus the probability that they are different copies of inbreeding using F, the mean inbreeding coefficient for the
that are IBD ( 12 FA ). Symbolically, we write population. The modified Hardy–Weinberg frequencies are
P(z ∼ y) = 1 + 12 FA f A /A = p2 + pqF
2
P(x ~ y), P(w ~ z), and P(z ~ y) are independent probabil- f A /a = 2pq − 2pqF
ities, so we can use the product rule and put it all together f a /a = q 2 + pqF
to obtain These modified Hardy–Weinberg proportions make intui-
FI = P(x ~ y) × P(w ~ z) × P(z ~ y) tive sense, showing how inbreeding reduces the frequency
of heterozygotes by 2pqF and adds half this amount to each
= 12 × 12 × ( 12 + 12 FA )
of the homozygous classes. With these modified Hardy–
3 Weinberg equations, you’ll also notice that when there is no
= ( 12 ) (1 + FA )
inbreeding (F = 0), you regain standard Hardy–Weinberg
In the analysis of inbred pedigrees, we can substitute the genotypic frequencies, and when there is complete inbreed-
value of FA into the equation above if it is known. Other- ing (F = 1), you get f A /A = p and f a /a = q .
wise, we can assume FA is zero if there is no information to How much does inbreeding increase the risk that off-
suggest that individual A is inbred. In the current example, spring will exhibit a recessive disease condition? Table 18-3
if we assume FA = 0, then shows the inbreeding coefficients for offspring of some dif-
3
FI = ( 21 ) = 1 ferent inbred matings and the predicted number of homozy-
8
gous recessives for different frequencies (q) of the recessive
This calculation tells us that the offspring of half-sib mat- allele. When q = 0.01, there is a 7-fold (7.19/1.0) increase
ings will be homozygous for alleles that are IBD for at least in homozygous recessive offspring for first-cousin matings
1/8 of their genes. It could be more than 1/8 if FA is greater as compared to matings between unrelated individuals. The
than zero. Additional inbred pedigrees and a general for- increase in risk jumps 13-fold (3.36/0.25) when q = 0.005
mula for calculating F can be found in Box 18-2. and 63-fold (0.63/0.01) when q = 0.001. In other words, the
When there is inbreeding in a population, the degree of risk jumps dramatically for rare alleles. Brother–
r andom-mating assumption of Hardy–Weinberg will be sister and parent–offspring matings are the riskiest: when
where n is the number of individuals in the inbreeding loop For pedigrees with multiple inbreeding loops, you sum the
not counting I. Let’s look at another pedigree, one in which contribution over all of the loops where FA is the inbreeding
the grandparents of I are half-sibs: coefficient of the ancestor (A) of the given loop:
n
A FI = ∑ ( 21 ) (1 + FA )
loops
TABLE 18-3 Number of Homozygous Recessives per as the one on the Tristan de Cunha Islands in the South
10,000 Individuals for Different Allele Atlantic, which has fewer than 300 people. Let’s look at the
Frequencies (q) effect of population size on the overall level of inbreeding
Mating F q = 0.01 q = 0.005 q = 0.001 in a population as measured by F.
Consider a population with Ft being the level of inbreed-
Unrelated parents 0.0 1.00 0.25 0.01
ing at generation t. To form an individual in the next gen-
Parent–offspring 1/4 25.75 12.69 2.51
eration t + 1, we select the first allele from the gene pool.
or brother–sister
Suppose the population size is N. After the first allele is
Half-sib 1/8 13.38 6.47 1.26
selected, the probability that the second allele we pick will
First cousin 1/16 7.19 3.36 0.63 be exactly the same copy is 1/2N and the inbreeding coef-
Second cousin 1/64 2.55 1.03 0.17 ficient for this individual is 1.0. The probability that the
second allele we pick will be a different copy from the first
allele is 1 − 1/2N and the level of inbreeding for the result-
q = 0.001, they show a 250-fold (2.51/0.01) greater risk
ing individual would be Ft , the average inbreeding coeffi-
compared to matings between unrelated individuals.
cient for the initial population at generation t. The level of
The impact of inbreeding on the frequency of genetic
inbreeding in the next generation is the sum of these two
disorders in human populations can be seen in Figure 18-13.
possible outcomes, or
Children of marriages of first cousins show about a two-
fold higher frequency of disorders as compared to children 1 1
Ft +1 = 1 + 1 − Ft
of unrelated parents. Historical records suggest that the 2N 2N
risks of inbreeding were understood long before the field of
genetics existed. This equation informs us that F will increase over
time as a function of population size. When N is large, F
KEY CONCEPT The inbreeding coefficient (F) is the probabil- increases slowly over time. When N is small, F increases
ity that two alleles in an individual trace back to the same copy rapidly over time. For example, suppose Ft in the initial
in a common ancestor. population is 0.1 and N = 10, 000 . Then Ft +1 would be
0.10005, just a slightly higher value. However, if N = 10,
then Ft +1 would be 0.145, a much higher value. We can also
Population size and inbreeding use this equation recursively to calculate Ft +2 by using Ft +1
Population size is a major factor contributing to the level of in place of Ft on the right side. The result with N = 10 and
inbreeding in populations. In small populations, individuals Ft = 0.1 would be Ft +2 = 0.188. The effects of population
are more likely to mate with a relative than in large ones. size on inbreeding in populations are further explored in
The phenomenon is seen in small human populations such Box 18-3.
14
Frequency of disorders (%)
12
10
0
United States France Sweden Japan Average
H = fA /a = 2 pq − 2 pqF
1
which can be rewritten as
(1− F ) = H /2 pq
N = 10
Combining these two equations, we obtain
Inbreeding (F )
N = 50
1 0.5
Ht+1 /2 pq = 1− Ht /2 pq
2N
N = 100
and then
1
Ht+1 = 1− Ht N = 500
2N
0
Thus, for each generation, the level of heterozygosity is 0 50 100
reduced by the fraction (1− 1/2N ). The reduction in H over Time in generations
t generations is Increase in inbreeding (F) over time for several different population sizes.
A consequence of the increased inbreeding is that individu- To describe how we can quantify variation, we will use
als in small populations are more likely to be homozygous for data for the glucose-6-phosphate dehydrogenase (G6PD)
deleterious alleles just as the offspring of first-cousin marriages gene from humans. G6PD is an X-linked gene that encodes
are more likely to be homozygous for such alleles. This effect an enzyme that catalyzes a step in glycolysis. The wild-type
is seen in ethnic groups that live in small, reproductively iso- allele (B) of G6PD has full enzyme activity. A second allele
lated communities. For example, a form of dwarfism in which called A− leads to strongly reduced enzyme activity, and
affected individuals have six fingers occurs at a frequency of individuals who carry this allele develop hemolytic anemia.
more than 1 in 200 among a population of about 13,000 However, this allele also confers a 50 percent reduction in
Amish in Lancaster County, Pennsylvania, although its fre- the risk of severe malaria in carriers. In regions of Africa
quency in the general U.S. population is only 1 in 60,000. where malaria is endemic, the A− allele reaches frequencies
near 20 percent, although this allele is absent or rare else-
KEY CONCEPT The inbreeding coefficient (F) increases over where. Another allele (A+ ) leads to only modestly reduced
time as a function of population size (N). For smaller N, the rate enzyme activity. Unlike individuals carrying the A− allele,
at which F grows is faster than for larger values of N. In smaller individuals carrying only the A+ or B alleles do not develop
populations, there is a greater risk of homozygosity for hemolytic anemia.
deleterious recessive alleles.
Figure 18-14 shows SNPs at 18 polymorphic sites in a
5102-bp segment of G6PD from a worldwide sample of
47 men. The remaining 5084 sites were fixed, or invariant:
18.4 GENETIC VARIATION AND ITS that is, only a single allele (nucleotide) exists in the entire
MEASUREMENT sample for each of these sites. By sampling only males, we
observe just one allele and one haplotype for each individ-
LO 18.1 Describe and analyze data to determine how much ual because the gene is X linked. The A+ allele differs from
genetic variation exists within populations. B by a single amino acid change (aspartic acid in place of
asparagine) at SNP3 in Figure 18-14. The A− allele differs
To study the amount and distribution of genetic variation from the B allele at two amino acids: it contains both the
in populations, we need some ways to quantify variation. “aspartic acid in place of asparagine” change found in the
Haplotype
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
A G A C C G C C C C C G G C T C A C
1 Southern African A- G A G G T T C G 1
2 Central African A- G A G G T T C G 1
3 Central African A- G A G G T T C G 1
4 African American A- G A G G T T C G 1
5 African American A- G A G G T T C G 1
6 Central African A- G A G G T T C G 1
7 Central African A+ G G T C G 2
8 Central African A+ G G T C G 2
9 Central African B C G 3
10 Southern African B A C T G 4
11 Southern African B A C T G 4
12 Southern African B T C T G 5
13 Southern African B C T G 6
14 Southern African B T A C G 7
15 Central African B T C G 8
16 European B T C G 8
17 European B T C G 8
18 European B T C G 8
19 Southwest Asian B T C G 8
20 East Asian B C G 3
21 Native American B A T C G 9
22 Southern African B 10
23 Native American B 10
24 Native American B 10
25 Native American B 10
26 Native American B 10
27 Native American B 10
28 Native American B 10
29 Native American B 10
30 Native American B 10
31 Native American B 10
32 European B 10
33 European B 10
34 European B 10
35 European B 10
36 European B 10
37 European B 10
38 Southwest Asian B 10
39 East Asian B 10
40 East Asian B 10
41 East Asian B 10
42 East Asian B 10
43 East Asian B 10
44 East Asian B 10
45 East Asian B 10
46 Pacific Islander B T 11
47 East Asian B T 12
FIGURE 18-14 Nucleotide variation for 5102 bp of the G6PD gene for a worldwide sample of 47
men. Only the 18 variable sites are shown. The functional allele class ( A− , A+, or B) is shown for each
sequence. SNP2 is a nonsynonymous SNP that causes a valine-to-methionine change that underlies
differences in enzyme activity associated with the A− allele. SNP3 is a nonsynonymous SNP that causes
an aspartic-acid-to-asparagine amino acid change. [Data from M. A. Saunders et al., Genetics 162, 2002,
1849–1861.]
A+ allele and a second amino acid difference (methionine in TABLE 18-4 Diversity Data for Glucose-6-Phosphate
place of valine) at SNP2. Dehydrogenase (G6PD) in Humans
How can we quantify variation at the G6PD locus? Total Non-
One simple measure is the number of polymorphic or sample Africans Africans
segregating sites (S). For the G6PD data, S is 18 for the Sample size 47 16 31
total sample, 14 for the African sample, and 7 for the Number of 18 14 7
non-African sample. Africans contain twice the number of segregating sites
segregating sites despite the fact that our sample has fewer
Number of 12 9 6
Africans. Another simple measure is the number of haplo- haplotypes
types (NH). The value of NH is 12 for the total sample, 9
Gene diversity 0.22 0.47 0.00
for the African sample, and 6 for the non-African sample. (GD) at SNP2
Again, the African sample has greater variation. One short-
Nucleotide diversity 0.0006 0.0008 0.0002
coming of measures such as S and NH is that the values we
observe depend heavily on sample size. If one samples more
individuals, then the values of S and NH are apt to increase. for nucleotide diversity for genes are typically very small.
For example, our sample has 16 Africans compared to 31 For G6PD, there are only 18 polymorphic nucleotide sites
non-Africans. Although S is twice as large in Africans as in but 5084 invariant sites. The average nucleotide diversity
non-Africans, the difference would likely be even greater if for the entire G6PD gene sequence is 0.0008 in Africans,
we had an equal number (31) of Africans and non-Africans. 0.0002 in non-Africans, and 0.0006 for the entire sample.
In place of S and NH, we can calculate allele frequen- These values tell us that Africans have four times as much
cies, which are not biased by differences in sample size. For nucleotide diversity at G6PD as non-Africans.
the G6PD data, B, A− , and A+ have worldwide frequen- Figure 18-15 shows the level of nucleotide diversity in
cies of 0.83, 0.13, and 0.04, respectively. However, you will several organisms. Unicellular eukaryotes are the most
note that A− has a frequency of 0.0 outside of Africa and diverse, followed by plants and then invertebrates. Ver-
0.38 in our African sample, which is a substantial differ- tebrates are the least diverse group; however, most verte-
ence. We can use allele frequency data to calculate a statis- brates still possess a lot of nucleotide diversity. For humans,
tic called gene diversity (GD), which is the probability that
two alleles drawn at random from the gene pool will be
different. The probability of drawing two different alleles is
equal to 1 minus the probability of drawing two copies of Nucleotide diversity among diverse organisms
the same allele summed over all alleles at the locus. Thus, 1 Vertebrates
2
GD = 1 − ∑ pi2 3 Invertebrates
4 Plants
= 1 − (p12 + p22 + p32 + pn2 ) 5
Unicellular
6
7 eukaryotes
where pi is the frequency of the ith allele and ∑ is the sum-
Species (see legend)
8
mation sign, indicating that we add the squares of all n 9
10
observed values of p for i = 1, 2, through the nth allele. 11
The value of GD can vary from 0 to 1. It will approach 12
13
1 when there is a large number of alleles of roughly equal 14
frequencies. It is 0 when there is a single allele, and it is 15
16
near 0 whenever there is a single very common allele with 17
a frequency of 0.99 or higher. Table 18-4 shows that gene 18
19
diversity is quite high in Africans (0.47). Since non-Africans 20
have only the B allele, gene diversity is 0.0. 21
22
The value of GD is equal to the expected proportion of 0 0.02 0.04 0.06 0.08 0.1
heterozygotes under Hardy–Weinberg equilibrium, which is
Nucleotide diversity
heterozygosity (H). However, H as a concept applies only to
diploids, and it would not apply to X-linked loci in males. FIGURE 18-15 Levels of nucleotide diversity at synonymous and
Thus, conceptually gene diversity (GD) is more appropriate silent sites in some different organisms. (1) Mus musculus, (2) Homo
even if it is mathematically the same quantity as H for pop- sapiens, (3) Oryza sativa, (4) Plasmodium falciparum, (5) Fugu rubripes,
(6) Strongylocentrotus purpuratus, (7) Anopheles gambiae, (8) Ciona
ulations of diploids under Hardy–Weinberg equilibrium. intestinalis, (9) Arabidopsis thaliana, (10) Caenorhabditis elegans, (11)
Gene diversity can be calculated for a single nucleotide Zea mays, (12) Encephalitozoon cuniculi, (13) Drosophila melanogaster,
site. It can be averaged over all the nucleotide sites in a (14) Leishmania major, (15) Trypanosoma species, (16) Toxoplasma
gene, in which case it is referred to as nucleotide d iversity. gondii, (17) Giardia lamblia, (18) Neurospora crassa, (19) Dictyostelium
Since the vast majority of nucleotides in any two cop- discoideum, (20) Saccharomyces cerevisiae, (21) Cryptosporidium
parvum, (22) Cryptococcus neoformans. [Data from M. Lynch and J. S.
ies of a gene from a species are typically the same, values Conery, Science 302, 2003, 1401–1404.]
nucleotide diversity is about 0.001, meaning that two ran- later and record any new mutations that have occurred.
domly chosen human chromosomes will differ at about The number of observed mutations per genome per genera-
1 bp per thousand. With 3 billion bp in our genome, that tion provides an estimate of the rate. Because one is looking
adds up to a total of about 3 million differences between for rather rare events, it is necessary to sequence billions of
the set of chromosomes inherited from a person’s mother nucleotides to find just a few SNP mutations. In 2009, the
and the set inherited from a person’s father for non-inbred SNP mutation rate for a part of the human Y chromosome
individuals. was estimated by this approach to be 3.0 × 10−8 mutations/
nucleotide/generation, or about one mutation every 30 mil-
KEY CONCEPT The rich genetic variation in biological pop- lion bp. If we extrapolate to the entire human genome (3
ulations can be quantified by different statistics, such as num- billion bp), then each of us has inherited 100 new muta-
ber of segregating sites, number of haplotypes, gene diversity, tions from each of our parents. Luckily, the vast majority of
and nucleotide diversity, to compare levels of variation among
mutations are not detrimental since they occur in regions of
populations and species.
the genome that are not critical.
Table 18-5 lists the mutation rates for SNPs and micro-
satellites in several model organisms. The SNP mutation
18.5 THE MODULATION OF rate is several orders of magnitude lower than the micro-
satellite rate. Their higher mutation rate and greater varia-
GENETIC VARIATION tion make microsatellites particularly useful in population
genetics and DNA forensics. The SNP mutation rate per
LO 18.4 Explain how new alleles enter a population by generation appears to be lower for unicellular organisms
mutation and migration.
than for large multicellular organisms. This difference can
LO 18.5 Measure the degree of linkage disequilibrium
be explained at least partially by the number of cell divi-
between two loci.
sions per generation. There are about 200 cell divisions
LO 18.6 Calculate the effect of genetic drift on gene
frequencies within populations.
from zygote to gamete in humans but only 1 in E. coli. If
the human rate is divided by 200, then the rate per cell divi-
LO 18.7 Calculate the effect of selection on gene
frequencies within populations. sion in humans is remarkably close to the rate in E. coli.
Other than mutation, the only other means for new
What are the forces that modulate the amount of genetic variation to enter a population is through migration or
variation in a population? How do new alleles enter the gene f low, the movement of individuals (or gametes)
gene pool? What forces remove alleles from the gene pool? between populations. Most species are divided into a set
How can genetic variants be recombined to create novel of small local populations or subpopulations. Physical bar-
combinations of alleles? Answers to these questions are riers such as oceans, rivers, or mountains may reduce gene
at the heart of understanding the process of evolution. In flow between subpopulations, but often some degree of
this section, we will examine the roles of mutation, migra- gene flow occurs despite such barriers. Within subpopula-
tion, recombination, genetic drift (chance), and selection in tions, an individual may have a chance to mate with any
sculpting the genetic composition of populations. other member of the opposite sex; however, individuals
from different subpopulations cannot mate unless there is
migration.
New alleles enter the population: mutation
and migration
Mutation is the ultimate source of all genetic variation. In
TABLE 18-5 Approximate Mutation Rates per
Chapter 15, we discussed the molecular mechanisms that
Generation per Haploid Genome
underlie small-scale mutations such as point mutations,
SNP mutations
indels, and changes in the number of repeat units in micro-
Organism (per bp) Microsatellite
satellites. Population geneticists are particularly interested
in the mutation rate, which is the probability that a copy Arabidopsis 7 × 10 −9
9 × 10−4
of an allele changes to some other allelic form in one gen- Maize 3 × 10−8 8 × 10−4
eration. The mutation rate is typically symbolized by the E. coli –
5 × 10−10
Greek letter µ. As we will see in this section, if we know
the mutation rate and the number of nucleotide differences Yeast 5 × 10 −10
4 × 10−5
between two sequences, then we can estimate how long ago C. elegans 3 × 10−9 4 × 10−3
the two sequences diverged. Drosophila 4 × 10−9 9 × 10−6
Geneticists can estimate mutation rates by starting with
a single homozygous individual and following the pedigree Mouse 4 × 10−9 3 × 10−4
of its descendants for several generations. Then they can Human 3 × 10−8 6 × 10−4
compare the DNA sequence of the founding individual to Note: Microsatellite rate is for di- or trinucleotide repeat microsatellites.
the DNA sequences of the descendants several generations Source: Data from multiple published studies.
Isolated subpopulations tend to diverge as each accu- gained or lost; rather, recombination creates new haplo-
mulates its own unique mutations. Gene flow limits genetic types. Let’s look at how this works. Consider linked loci
divergence between subpopulations. One of the genetic A and B. There could be a population in which only two
consequences of migration is genetic admixture, the mix haplotypes are found at generation t0 : AB and ab. Suppose
of genes that results when individuals have ancestry from an individual in this population is heterozygous for these
more than one subpopulation. This phenomenon is com- two haplotypes:
mon in human populations. It is readily observed in South
A B
Africa, where migrants from around the world were
brought together. As shown in Figure 18-16, the genomes of a b
South Africans of mixed ancestry are complex and include If a crossover occurs during meiosis in this individual, then
parts from the indigenous people of southern Africa plus gametes with two new haplotypes, Ab and aB, could be
contributions of migrants from western Africa, Europe, formed and enter the population in generation t1.
India, East Asia, and other regions.
A b a B
KEY CONCEPT Mutation is the ultimate source of all genetic
variation. Migration can add genetic variation to a population Thus, recombination can create variation that takes the
via gene flow from another population of the same species. form of new haplotypes. The new haplotypes can have
unique properties that alter protein function. For example,
Recombination and linkage disequilibrium suppose an amino acid variant in a protein on one haplo-
type increases the enzyme activity of the protein twofold
Recombination is a critical force sculpting patterns of and a second amino acid variant on another haplotype also
genetic variation in populations. In this case, alleles are not increases activity twofold. A recombination event that com-
bines these two variants would yield a protein with four-
Migrants from around the world have contributed to fold higher activity.
the genomes of some South Africans Let’s now consider the observed and expected frequen-
100 cies of the four possible haplotypes for two loci, each with
two alleles. Linked loci, A and B, have alleles A and a and
B and b with frequencies pA , pa , pB , and pb , respectively.
The four possible haplotypes are AB, Ab, aB, and ab with
observed frequencies PAB , PAb , PaB , and Pab . At what fre-
quency do we expect to find each of these four haplotypes?
different continental regions
Percent of genome from
50 that haplotype:
PAB = pA × pB
PAb = pA × pb
PaB = pa × pB
Pab = pa × pb
For example, suppose that the frequency of each of the
alleles is 0.5; that is, pA = pa = pB = pb = 0.5. When we
0 sample the gene pool, the probability of drawing a chro-
Individuals mosome with an A allele is 0.5. If the relationship between
Southern Africa the alleles at locus A and the alleles at locus B is random,
Western Africa then the probability that the selected chromosome has the
Europe
East Asia
B allele is also 0.5. Thus, the probability that we draw a
India chromosome with the AB haplotype is
Other regions
PAB = pA × pB = 0.5 × 0.5 = 0.25
FIGURE 18-16 Graphical representation of genetic admixture for 39 If the association between the alleles at two loci is random
people of mixed ancestry from South Africa. Each column represents one
as just described, then the two loci are said to be at linkage
person’s genome, and the colors represent the parts of their genome
contributed by their ancestors, who came from many regions of the equilibrium. In this case, the observed and expected fre-
world. The figure is based on the population genetic analysis of over 800 quencies will be the same. Figure 18-17a diagrams a case of
microsatellites and 500 indels that were scored for nearly 4000 people two loci at linkage equilibrium.
from around the world, including the 39 of mixed ancestry from South If the association between the alleles at two
Africa. [Data from S. A. Tishkoff et al., Science 324, 2009, 1035–1044.]
loci is nonrandom, then the loci are said to be in
Linkage disequilibrium is the nonrandom association their alleles. The rate of decline in LD depends on the rate
between two loci at which crossing over occurs. The frequency of recombi-
nants (RF) between the two loci among the gametes that
(a) Linkage equilibrium (b) Linkage disequilibrium
form the next generation (see Chapter 4) provides an esti-
A B A B
mate of recombination rate, which in population genetics is
A B A B symbolized by the lowercase letter r. If D0 is the value for
A b A B linkage disequilibrium between two loci in the current gen-
A b A B eration, then the value in the next generation (D1) is given
a B a b by this equation:
a B a b
D1 = D0 (1 − r)
a b a b
a b a b In other words, linkage disequilibrium as measured by D
declines at a rate of (1 − r) per generation. When r is small,
pA = 0.5 PAB = 0.25 pA = 0.5 PAB = 0.5
pa = 0.5 PAb = 0.25 pa = 0.5 PAb = 0.0 D declines slowly over time. When r is at its maximum
pB = 0.5 PaB = 0.25 pB = 0.5 PaB = 0.0 (0.5), then D declines by 1/2 each generation.
pb = 0.5 Pab = 0.25 pb = 0.5 Pab = 0.5 Since LD decays as a function of time and the recom-
bination fraction, population geneticists can use the level
FIGURE 18-17 (a) Linkage equilibrium and (b) linkage disequilibrium of LD between a mutation and the loci surrounding it to
for two loci (A and B). estimate the time in generations since the mutation first
arose in the population. Older mutations have little LD
linkage disequilibrium (LD). In this case, a specific allele with neighboring loci, while recent mutations show a high
at the first locus is associated with a specific allele at the level of LD with neighboring loci. If you look again at Fig-
second locus more often than expected by chance. Figure ure 18-14, you will notice that there is considerable LD
18-17b diagrams a case of complete LD between two loci. between SNP2 in G6PD and the neighboring SNPs. SNP2
The A allele is always associated with the B allele, while the encodes the amino acid change of valine to methionine in
a allele is always associated with the b allele. There are no the A− allele that confers resistance to malaria. Population
chromosomes with haplotypes Ab or aB. In this case, the geneticists have used LD at G6PD to estimate that the A−
observed and expected frequencies will not be the same. allele arose about 10,000 years ago. Malaria is not thought
We can quantify the level of LD between two loci as the to have been prevalent in Africa until then. Thus, the A−
difference (D) between the observed frequency of a haplo- arose by random mutation but was maintained in the pop-
type and the expected frequency given a random associa- ulation because it provided protection against malaria.
tion among alleles at the two loci. If both loci involved have
just two alleles, then KEY CONCEPT Linkage disequilibrium is the outcome of the
fact that new mutations arise on a single haplotype. Linkage
D = PAB − pA pB disequilibrium will decay over time because of recombination.
In Figure 18-17a, D = 0 since there is no LD, and in Figure
18-17b, D = 0.25, which is greater than 0, indicating the
presence of LD. Genetic drift and population size
How does LD arise? Whenever a new mutation occurs The Hardy–Weinberg law tells us that allele frequencies
at a locus, the mutation appears on a single specific chro- remain the same from one generation to the next in an
mosome and so it is instantly linked to (or associated with) infinitely large population. However, actual populations of
the specific alleles at any neighboring loci on that chromo- organisms in nature are finite rather than infinite in size.
some. Consider a population in which there are just two In finite populations, allele frequencies may change from
haplotypes: AB and Ab. If a new mutation (a) arises at one generation to the next as the result of chance (sampling
the A locus on a chromosome that already possesses the error) when gametes are drawn from the gene pool to form
b allele at the B locus, then a new ab haplotype would be the next generation. Change in allele frequencies between
formed. Over time, this new ab haplotype might rise in fre- generations due to sampling error is called random genetic
quency in the population. Other chromosomes in the pop- drift, or just drift for short.
ulation would possess the AB or Ab haplotypes at these Let’s consider a simple but extreme case—a population
two loci, but no chromosomes would possess aB. Thus, the composed of a single heterozygous (A /a) individual (N = 1)
loci would be in LD. Migration can also cause LD when at generation t0 . We will allow self-fertilization. In this case,
one subpopulation possesses only the AB haplotype and the gene pool can be described as having two alleles, A and
another only the ab haplotype. Any migrants between the a, each present at a frequency of p = q = 0.5. The size of
subpopulations would give rise to LD within the subpopu- the population remains the same, N = 1, in the subsequent
lation that receives the migrants. generation, t1. What is the probability that the allele fre-
LD between two loci will decline over time as cross- quencies will change (“drift”) to p = 1 and q = 0 at gen-
overs between them randomize the relationship between eration t1? In other words, what is the probability that
the population will become fixed for the A allele, so that 499 copies of a (q = 0.499), then there has been genetic drift,
it consists of a single homozygous A /A individual? Since albeit a very modest level of drift. A general formula for cal-
N = 1, we need to draw just two gametes from the gene culating the probability of observing a specific number of
pool to form a single individual. The probability of drawing copies of an allele in the next generation, given the frequen-
two A’s is p2 = 0.52 = 0.25. Thus, 25 percent of the time cies in the current generation, is presented in Box 18-4.
this population will “drift” away from the initial allele fre- When drift is operating in a finite population, one can cal-
quencies and become fixed for the A allele after just one culate the probabilities of different outcomes, but one cannot
generation. accurately predict the specific outcome that will occur. The
What happens if we increase the population size to process is like rolling dice. At any locus, drift can continue
N = 2 and the initial gene pool still has p = q = 0.5? The from one generation to the next until one allele has become
allele frequencies will change to p = 1 and q = 0 in the fixed. Also, in a particular population, the frequency of the A
next generation only if the population consists of two A /A allele may increase from generation t0 to t1 but then decrease
individuals. For this to happen, we need to draw four A from generation t1 to t2 . Drift does not proceed in a specific
alleles, each with a probability of p = 0.5, so the proba- direction toward loss or fixation of an allele.
bility that the next generation will have p = 1 and q = 0.0 Figures 18-18a and 18-18b show computer-simulated
is p4 = (0.5)4 = 0.0625, or just over 6 percent. Thus, an random trials (rolls of the dice) for six populations of size
N = 2 population is less likely to drift to fixation of the A N = 10 and N = 500. Each population starts having two
allele than an N = 1 population. More generally, the proba- alleles at a frequency of p = q = 0.5, then the random trials
bility of a population drifting to the fixation of the A allele proceed for 30 generations. First, notice the randomness of
in a single generation is p2N , and thus this probability gets the process from one generation to the next. For example,
progressively smaller as the population size (N) gets larger. the frequency of A in the population depicted by the yellow
Drift is a weaker force in large populations. line in Figure 18-18a bounces up and down from one gen-
Drift means any change in allele frequencies due to sam- eration to the next, hitting a low of p = 0.15 at t16 but then
pling error, not just loss or fixation of an allele. In a pop- rebounding to p = 0.75 at t30. Second, whether N = 10 or
ulation of N = 500 with two alleles at a frequency of N = 500, notice that no two populations have exactly the
p = q = 0.5, there are 500 copies of A and 500 copies of a. same trajectory. Drift is a random process, and we are not
If the next generation has 501 copies of A (p = 0.501) and likely to observe exactly the same outcome with different
0.5
0.0
0 10 20 30
Generations (t)
(b)
1.0
N = 500
Allele frequency (p)
0.5
0.0
0 10 20 30
Generations (t)
(c)
1.0
N = 10
Allele frequency (p)
0.5
0.0
0 10 20 30
Generations (t)
populations over many generations except when N is very is very close to the expectation that A will go to fixation
small. Third, notice that when N = 10, the populations 10 percent of the time when p = 0.1.
became fixed (either p = 1 or p = 0 ) before generation The fact that the frequency of an allele is equal to its prob-
30 in five of the six trials. However, when N = 500, the ability of fixation means that most newly arising mutations
populations retained both alleles in all six trials even after will ultimately be lost from a population because of drift. The
30 generations. initial frequency of a new mutation in the gene pool is
In addition to population size, the fate of an allele is
1
determined by its frequency in the population. Specifically,
the probability that an allele will drift to fixation in a future 2N
generation is equal to its frequency in the present gener- If N is even modestly large, such as 10,000, then the prob-
ation. An allele that is at a frequency of 0.5 has a 50:50 ability that a new mutation will ultimately reach fixation
chance of fixation or loss from the population in a future is extremely small: 1/2N = 1/20, 000 = 5 × 10−5. The prob-
generation. You can see the effect of allele frequency on ability that a new mutation will ultimately be lost from the
the fate of an allele in Figure 18-18c. For 10 populations population is
with an initial frequency of p = 0.1, eight populations expe-
2N − 1 1
rienced the loss of the A allele, one its fixation, and one = 1−
population retained both alleles after 30 generations. That 2N 2N
(a) 4N
2N
Number of allele copies
0
Time
(b)
4N
Number of allele copies
2N
Time
FIGURE 18-19 (a) Graphical representation of the appearance, loss, and eventual incorporation of
new mutations in a population over time under the action of genetic drift. Black lines show the fate of
most new mutations, which appear and then are lost from the population within a few generations.
Colored lines show the fate of the few “lucky” mutations that continue to rise in frequency until they
reach fixation. (b) A population that is 1/2 the size of the population in part a. In this population, 4N
generations are about 1/2 as long and the lucky new mutations are fixed more rapidly.
which is close to 1.0 in large populations. It is 0.99995 in a or she may pass only the less favorable allele to his or her
population of 10,000. offspring by chance.
Figure 18-19a shows a graphical representation of the In calculating the probabilities of different outcomes
fate of new mutations in a population. The x-axis rep- under genetic drift, we are assuming that the A and a alleles
resents time and the y-axis the number of copies of an do not confer differences in viability or reproductive suc-
allele. The black lines at the bottom of each graph show cess to the individuals that carry them. We assume that
the fate of most new mutations. They appear and then are A /A, A /a, and a /a individuals are all equally likely to sur-
soon lost from the population. The colored lines show the vive and reproduce. In this case, A and a would be termed
few “lucky” new mutations that become fixed. From popu- neutral alleles (or variants) relative to each other. Change
lation genetic theory, it can be shown that the average time in the frequencies of neutral alleles over time due to drift is
required for a lucky mutation to become fixed is 4N gen- called neutral evolution. The process of neutral evolution is
erations. Figure 18-19b shows a population that is 1/2 the the foundation for the molecular clock, the constant rate of
size of the population in Figure 18-19a. Thus, 4N genera- substitution of newly arising allelic variants for preexisting
tions are about 1/2 as long and the lucky new mutations are ones over long periods (Box 18-5). Neutral evolution is dis-
fixed more rapidly. tinct from Darwinian evolution, in which favorable alleles
AnIntroduction
importanttoconsequence of drift
Genetic Analysis, 11e is that slightly dele- rise in frequency because the individuals that carry them
teriousFigure 18.19
alleles can#1825
be brought to fixation or advantageous leave more offspring. We will discuss Darwinian evolution
alleles07/08/14
lost by this random process. Consider a new allele in the next section of this chapter and in Chapter 20.
Dragonfly Media Group
that arises in a population and endows the individual car- Up until now, we have been considering drift in the con-
rying it with a stronger immune system. This individual can text of populations that remain the same size from one gen-
pass the advantageous allele to his or her offspring, but eration to the next. In reality, populations often contract or
those offspring might die before reproducing because of a expand in size over time. For example, a new population
random event such as being struck by lightning. Or if the of much smaller size can suddenly form when a relatively
individual carrying the favorable allele is heterozygous, he small number of the members of a population migrate to a
time for humans and chimps is 6.0 million years ago. These
calculations assume that the substitutions are neutral and
that the rate of substitution has been constant over time.
Species Species
A B
new location and establish a new population. The migrants, of species that have experienced recent bottlenecks because
or “founders,” of the new population may not carry all the of hunting by humans or encroachment by humans on
alleles present in the original population, or they may carry their habitat. The reduction in population size during a
the same alleles but at different frequencies. Genetic drift bottleneck increases the level of drift in a population. As
caused by random sampling of the original population to explained earlier in the chapter, the level of inbreeding in
create the new population is known as the founder effect. populations is also dependent on population size. Thus,
One of many founder events in human history occurred bottlenecks also cause an increase in the level of inbreeding.
when people crossed the Bering land bridge from Asia to The California condor presents a remarkable example
the Americas during the ice age about 15,000 to 30,000 of a bottleneck. This species was once wide ranging but
years ago. As a result, genetic diversity among Native in the 1980s declined to a breeding population of only
Americans is lower than among people in other regions of 14 captive birds. The population is now about 450 indi-
the world (Figure 18-20). viduals, but the average heterozygosity in the genome
Population size can also change within a single loca- decreased by 8 percent during the initial bottleneck. Fur-
tion. A period of one or several consecutive generations thermore, a deleterious recessive allele for a lethal form of
of contraction in population size is known as a popula- dwarfism occurs at a frequency of about 9 percent among
tion bottleneck. Bottlenecks occur in natural populations the surviving animals, presumably as a result of drift from a
because of environmental fluctuations such as a reduction lower frequency in the pre-bottleneck population. To man-
in the food supply or increase in predation. The gray wolf, age these problems, conservation biologists set up matings
American bison, bald eagle, California condor, whooping of captive animals to minimize further inbreeding and to
crane, and many whale species are some familiar examples purge deleterious alleles from the population.
Human populations have different bottleneck explains why our crop plants possess much less
levels of genetic diversity genetic diversity than their wild ancestors.
0.8
KEY CONCEPT Population size is a key factor affecting
Haplotype heterozygosity
0.5
0.5 0.6 0.7 0.8
Microsatellite heterozygosity Selection
Africa East Asia So far, we have considered how new alleles enter a pop-
Europe Oceania ulation through mutation and migration and how these
Middle East Americas alleles can become fixed in (or lost from) a population
Central/
South Asia by random drift. But mutation, migration, and drift can-
not explain why organisms seem so well adapted to their
FIGURE 18-20 Plot of haplotype heterozygosity versus microsatellite environments. They cannot explain adaptations, features
heterozygosity shows genetic diversity for different geographical
of an organism’s form or physiology that allow it to cope
groups of humans. Genetic diversity is lowest for Native Americans
because of the founder effect. [Data from D. F. Conrad et al., Nat. Genet. with the environmental conditions under which it lives. To
38, 2006, 1251–1260.] explain the origin of adaptations, Charles Darwin, in 1859
in his historic book The Origin of Species, proposed that
Box 18-6 discusses the well-characterized bottleneck that adaptations arise through the action of another process,
occurred during the domestication of crop species. This which he called “natural selection.” In this section, we will
explore the role of natural selection in modulating genetic individuals are the same and superior to the fitness of the
variation within populations. Later, in Chapter 20, we will a /a individuals. We are assuming that this population fol-
consider the effects of natural selection on the evolution of lows the Hardy–Weinberg law, with p = 0.1 and q = 0.9.
genes and traits over extended periods.
Let’s define natural selection as the process by which A /A A /a a /a
individuals with certain heritable features are more likely to Average number of 10 10 5
survive and reproduce than are other individuals that lack offspring (W)
these features. As outlined by Darwin, the process works like Relative fitness (w) 1.0 1.0 0.5
this. In each generation, more offspring are produced than Genotype frequency 0.01 0.18 0.81
can survive and reproduce in the environment. Nature has
a mechanism (mutation) to generate new heritable forms or The relative contribution of each genotype to the gene pool
variants. Individuals with particular variants of some fea- is determined by the product of its fitness and its frequency.
tures are more likely to survive and reproduce. Individuals The more fit and the higher the frequency of a genotype,
with features that enhance their ability to survive and repro- the more it contributes.
duce will transmit these features to their offspring. Over
time, these features will rise in frequency in the population. Genotype A /A A /a a /a Sum
Thus, populations will change over time (evolve) as the envi- Relative 1× 0.01 1× 0.18 0.5 × 0.81 0.595
ronment (nature) favors (selects) features that enhance the contribution = 0.01 = 0.18 = 0.405
ability to survive and reproduce. This is Darwin’s theory of
evolution by means of natural selection. The relative contributions do not sum to 1, so we need
Darwinian evolution is often described using the phrase to rescale them by dividing each by the sum of all three
“survival of the fittest.” This phrase can be misleading. An (0.595) to get the expected frequencies of the genotypes
individual who is physically strong, resistant to disease, and that contribute to the gene pool.
lives a long life but has no offspring is not fit in the Darwin-
ian sense. Darwinian fitness refers to the ability to survive Genotype A /A A /a a /a Sum
and reproduce. It considers both viability and fecundity. Genotype frequencies 0.02 0.30 0.68 1.0
One measure of Darwinian fitness is simply the number
of offspring that an individual has. This measure is called Using these expected genotype frequencies and the Hardy–
absolute fitness, and we will symbolize it with an uppercase Weinberg law, we can calculate the frequencies of the alleles
W. For an individual with no offspring, W equals 0, for an in the next generation:
individual with one offspring, W equals 1, for an individual
p′ = 0.02 + ( 12 × 0.3) = 0.17
with two offspring, W equals 2, and so forth. W is also the
number of alleles at a locus that an individual contributes and
to the gene pool.
q′ = 0.68 + ( 12 × 0.3) = 0.83
Absolute fitness confounds population size and differ-
ences in reproductive success among individuals. Popula- The difference between p′ and p (∆p = p ′ − p) is 0.17 −
tion geneticists are primarily interested in the latter, and so 0.1 = 0.07, so we conclude that the A allele has climbed
they use a measure called relative fitness (symbolized by a 7 percent in one generation due to natural selection. Box 18-7
lowercase w), which is the fitness of an individual relative presents the standard equations for calculating changes in
to some other individual, usually the most fit individual allele frequencies over time due to natural selection.
in the population. If individual X has 2 offspring and the We could go through this process recursively, using the
most fit individual, Y, has 10 offspring, then the relative allele frequencies from the first generation to calculate those
fitness of X is w = 2/10 = 0.2 . The relative fitness of Y is in the second generation, then using those from the second
w = 10/10 = 1. For every 10 alleles Y contributes to the to calculate the third, and so forth. If we then plotted p by
next generation, X will contribute 2. time measured in number of generations (t), we would have
The concept of fitness applies to genotypes as well as a picture of the tempo with which allele frequencies change
to individuals. The absolute fitness for the A /A genotype under the force of natural selection. Figure 18-21 shows such
(WA /A ) is the average number of offspring left by individ- a plot for both a favored dominant and a favored recessive
uals with that genotype. If we know the absolute fitnesses allele. The dominant allele rises rapidly to start but then
for all genotypes at a locus, we can calculate the relative hits a plateau and only slowly approaches fixation. Once
fitnesses for each of the genotypes. the favored dominant allele is at a high frequency, the unfa-
Let’s now look at how allele frequencies can change vored recessive allele occurs mostly in heterozygotes and
over time when different genotypes have different fitnesses; rarely as homozygotes with reduced fitness, so selection is
that is, when natural selection is at work. Below are the ineffective at purging it from the population. The favored
fitnesses and genotype frequencies for the three genotypes recessive behaves in the opposite manner—it rises slowly
at the A locus in a population. In this case, A is a favored in frequency at first since a /a homozygotes with enhanced
dominant allele since the fitnesses of the A /A and A /a fitness are rare but proceeds more rapidly to fixation later.
w = pw A + qw a
Np2WA /A + NpqWA /a
p′ =
NW Substituting this expression for w in the formula for ∆p and
remembering that q = 1− p, we obtain
This equation reduces to
pWA /A + qWA /a pq(w A − w a )
p′ = p ∆p =
W w
Since the heterozygous class has reduced fitness, the unfa- new, favorable mutation or allele to a higher frequency.
vored dominant allele can eventually be purged from the This type of selection is at work when new adaptations
population. evolve. A selective sweep occurs when a favorable allele
reaches fixation. Directional selection can also work to
remove deleterious mutations from the population. This
Forms of selection form of selection is called purifying selection, and it pre-
Natural selection can operate in several different ways. vents existing adaptive features from being degraded or
Directional selection, which we have been discussing, lost. Selection does not always proceed directionally until
moves the frequency of an allele in one direction until it loss or fixation of an allele. If the heterozygous class has a
reaches fixation or loss. Directional selection can be either higher fitness than either of the homozygous classes, then
positive or purifying. Positive selection works to bring a natural selection will favor the maintenance of both alleles
Allele frequencies change under the force In Europe, a selective sweep caused a loss of all
of natural selection diversity at the SLC24A5 locus
1.0 0.3
Native Americans
0.1
0.5 East Asians
Europeans
0.0
0 50 150 200
SLC24A5
Favored Distance in kilobase pairs
recessive
FIGURE 18-23 Gene diversity in human continental groups along
a 2-million-bp segment of human chromosome 15 surrounding the
SLC24A5 gene. [Data from Human Genome Diversity Project, www.hagsc
0.0
0 200 400 600 .org/hgdp.]
Time in generations
haplotype in this region and thus no polymorphism. When
FIGURE 18-21 Change in allele frequency of a favored dominant selection is applied to the target site (shown in red), the tar-
allele (red) and a favored recessive allele (blue) driven by natural
selection over the course of 600 generations.
get and neighboring sites can all be swept to fixation before
recombination breaks up the haplotype in which the favor-
able mutation first occurred. The result is lower diversity
in the population. In this case, the locus is under balancing and higher LD near the target. As distance from the target
selection, and natural selection will move the population to increases, there is more opportunity for recombination, and
an equilibrium point at which both alleles are maintained so diversity goes gradually back up.
in the population (see Chapter 20). Figure 18-23 shows the pattern of diversity in the region
The different forms of selection each leave a distinct surrounding the SLC24A5 gene in humans. This gene influ-
signature on the DNA sequence near the target locus in a ences the deposition of melanin in the skin. When peo-
population. For example, positive selection can be detected ple migrated from Africa to Europe, a selective sweep at
in DNA sequences by its effects on genetic diversity and SLC24A5 caused a loss of all diversity at this locus (gene
linkage disequilibrium. Figure 18-22 shows schematic hap- diversity ≅ 0.0). As a consequence, there is a single allele
lotypes before and after an episode of positive selection. and a single haplotype at this locus in Europe. The sin-
In the panel showing the haplotypes before selection, the gle allele that was selected for in Europe produces lighter
bracketed region has many polymorphisms and multiple skin color. Moving away from the gene in either direction,
haplotypes. However, after selection, there is only a single gene diversity rises in European populations since recom-
bination disrupted the linkage disequilibrium between
Positive selection leaves a distinct signature SLC24A5 and more distance sites. Light skin may be
Haplotypes before selection Haplotypes after selection adaptive in northern latitudes. People are able to syn-
thesize vitamin D, but to do so they need to absorb UV
radiation through the skin. In the equatorial latitudes,
people are exposed to high levels of UV light and can
synthesize vitamin D even with heavily pigmented
skin. At greater distance from the equator, people are
exposed to less UV light, and lighter skin color may
facilitate vitamin D synthesis at these latitudes.
Table 18-6 lists a few of the genes that show evidence
for natural selection in modern humans. These genes
A B A B
Selective sweep
fall into a few basic categories. One group strengthens
resistance to pathogens. The genes G6PD, FY null , and
FIGURE 18-22 Schematic of haplotypes found in a population before and Hb (hemoglobin β, the sickle-cell-anemia gene) all help
after a favored allele (red) is swept to fixation. There are 11 loci altogether. There humans adapt to the threat of malaria. Figure 18-11b
are two alleles (red and gray) at the locus that was the target of selection. There
are two alleles (black and gray) at each locus that is linked to the target locus.
shows that the frequency of FY null is highest in central
After selection, the target and some neighboring sites have all been swept to Africa. Central Africa also has the highest prevalence of
fixation. malaria, suggesting that selection has driven FY null to
TABLE 18-6 Some Genes Showing Evidence for Natural Selection in Specific Human Populations
Gene Presumed Trait Population
EDA2R (ectodysplasin A2 receptor) Male pattern baldness Europeans
EDAR (ectodysplasin A receptor) Hair morphology East Asians
null
FY (Duffy antigen) Resistance to malaria Africans
G6PD (glucose-6-phosphate dehydrogenase) Resistance to malaria Africans
Hb (hemoglobin β) Resistance to malaria Africans
KITLG (KIT ligand) Skin pigmentation East Asians and Europeans
LARGE (glycosyltransferase) Resistance to Lassa fever Africans
LCT (lactase) Lactase persistence; ability to digest Africans, Europeans
milk sugar as an adult
LPR (leptin receptor) Processing of dietary fats East Asians
MC1R (melanocortin receptor) Hair and skin pigmentation East Asians
MHC (major histocompatibility complex) Infectious disease resistance Multiple populations
OCA2 (oculocutaneous albinism) Skin pigmentation and eye color Europeans
PPARD (peroxisome proliferator-activated Processing of dietary fats Europeans
receptor delta)
SI (sucrase-isomaltase) Sucrose metabolism East Asians
SLC24A5 (solute carrier family 24) Skin pigmentation Europeans and West Asians
TYRP1 (tyrosinase-related protein 1) Skin pigmentation Europeans
Sources: P. C. Sabeti et al., Science 312, 2006, 1614–1620; P. C. Sabeti et al., Nature 449, 2007, 913–919; B. F. Voight et al., PLoS Biology 4, 2006,
446–458; J. K. Pickrell et al., Genome Research 19, 2009, 826–837.
its highest frequency in the region where selection pressure is genes, which are involved in immune system recognition
greatest. In the 1990s, medical geneticists discovered an allele of (and response to) pathogens. Balancing selection is one
of the gene CCR5 (chemokine receptor 5) that provides resis- hypothesis proposed to explain the high diversity observed
tance to AIDS. This allele is now a target of natural selection. at the MHC. Since heterozygotes have two alleles, they may
As long as there are pathogens, natural selection will con- be resistant to a greater repertoire of pathogen types, giving
tinue to operate in human populations. heterozygotes a fitness advantage.
Another group of selected genes in Table 18-6 adapts Finally, selection can be imposed by an agent other than
people to regional diets. Before 10,000 years ago, all nature. Humans have imposed selection in the process of
humans were hunter–gatherers. More recently, most domesticating and improving cultivated plants and animals.
humans switched to agricultural foods, but there are This form of selection is called artificial selection. In this
regional differences in diet. In northern Europe and parts case, individuals with traits that humans prefer contribute
of Africa, milk products are a substantial part of the diet. more alleles to the gene pool than individuals with unfa-
In most populations, the lactase enzyme for digesting vored traits. Over time, the alleles that confer the favored
milk sugar (lactose) is expressed during childhood but is traits rise in frequency in the population. The many breeds
switched off in adults. In parts of Europe and Africa where of dogs and dairy cows and varieties of garden vegetables
adults drink milk, however, special alleles of the lactase and cereal crops are all the products of artificial selection.
gene that continue to express the lactase enzyme during
adulthood have risen in frequency due to natural selection. Balancing selection can lead to regions of unusually
Finally, Table 18-6 includes some genes for physiological high genetic diversity
adaptations to climate. Among these are the genes for skin
pigmentation such as SLC24A5, discussed earlier. MHC
SNPs/20 kb
KEY CONCEPT Natural selection is a force that can both which can be rewritten as
drive favorable alleles at a locus to fixation or maintain multiple
4Nµ
alleles at a locus in a population. Selection leaves a signature Hˆ =
in the genome in the form of the pattern of genetic diversity 4Nµ + 1
surrounding the target of selection. Population geneticists have This equation gives the equilibrium value of Ĥ when the
identified many genes that have been targets of selection in
loss by drift and gain by mutation are balanced. This equa-
humans.
tion applies only to neutral variation; that is, we are assum-
ing selection is not at work. We are also assuming that each
new mutation yields a unique allele.
Expressions such as this are useful when we have estimates
Balance between mutation and drift for two of the variables and would like to know the third.
We have considered the forces that regulate variation in For example, nucleotide diversity (H at the nucleotide level)
populations individually. Let’s now consider the opposing for noncoding sequences, which are largely neutral, is about
effects of mutation and drift, the former adding variation 0.0013 in humans, and µ for humans is 3 × 10−8 (see Table
and the latter removing it from populations. When these 18-5). Using these values and solving the equation above for
two forces are in balance, a population can reach an equi- N yields an estimate of the human population size of 10,498
librium at which the loss and gain of variation are equal. humans. This estimate is far below the 7.6 billion of us alive
We will use heterozygosity (H) as a measure of varia- today. What’s up? This is an estimate for the equilibrium value.
tion. Remember that H will be near 0 when a population Modern humans are a young group, only about 150,000 years
is near fixation for a single allele (low variation), and H old. Over the last 150,000 years, our population has grown
approaches 1 when there are many alleles of equal fre- dramatically as we filled the globe, but mutation is a slow pro-
quency (high variation). cess, so genetic diversity has not kept up and the human pop-
Let’s use H with a “hat,” Ĥ , as the symbol for the equi- ulation is not at equilibrium. The population size of 10,498
librium value of H. To find Ĥ , we start with two mathe- represents an estimate of our historical size, or how many
matical equations: one equation that relates change in H breeding members there were about 150,000 years ago.
to population size (drift) and another equation that relates
change in H to the mutation rate. We can then set these Balance between mutation and selection
equations equal to each other and solve for Ĥ . Allelic frequencies may also reach a stable equilibrium when
First, we need an equation for the decline in variation the introduction of new alleles by repeated mutation is bal-
(H) between generations as a function of population size anced by their removal by natural selection. This balance
(drift). We developed such an equation in Box 18-3 when probably explains the persistence of genetic diseases as low-
discussing inbreeding: level polymorphisms in human populations. New deleteri-
ous mutations are constantly arising spontaneously. These
1
H ′ = 1 − H mutations may be completely recessive or partly dominant.
2N
Selection removes them from the population, but there is an
This equation applies to the effects of drift as well as those equilibrium between their appearance and removal.
of inbreeding. From this equation, it follows that the change Let’s begin with the simplest case—the frequency for
in H between generations due to drift is a deleterious recessive when an equilibrium is reached
between mutation and selection. For this purpose, it is
1
∆H = H − H ′ = H convenient to express the relative fitnesses in terms of the
2N selection coefficient (s), which is the selective disadvantage
Second, we need an equation for the increase in variation, of (or loss of fitness in) a genotype:
as measured by H, between generations due to mutation. WA /A WA /a w a /a
Any new mutation will increase heterozygosity at a rate 1 1 1− s
proportional to the frequency of homozygotes in the pop-
Then, as shown in Box 18-8 , the equation for equilib-
ulation (1 − H ) times the rate at which mutation converts
rium frequency of a deleterious recessive allele is
them to heterozygotes (2µ) . (The 2 is necessary because
there are two alleles that could mutate in a diploid.) Thus, µ
qˆ =
the change in H between generations due to mutation is s
∆H = 2µ(1 − H ) This equation shows that the frequency at equilibrium
depends on the ratio µ /s. When the mutation rate for A → a
When the population reaches an equilibrium, the loss of
gets larger and the selective disadvantage smaller, then the
heterozygosity by drift will equal the gain from mutation.
equilibrium frequency (qˆ) of a recessive deleterious allele
Thus, we have
will rise. As an example, a recessive lethal allele (s = 1) that
1 ˆ arises by mutation from the wild-type allele at the rate of
H = 2µ(1 − Hˆ )
2N µ = 10−6 will have an equilibrium frequency of 10−3.
∆qmut = µ p ˆ ˆ2
−spq
µ pˆ =
1− sqˆ 2
A simple way to express the fitnesses of the genotypes
in the case of a recessive deleterious allele a is The frequency of a recessive deleterious allele (qˆ ) at equi-
w A /A = w A /a = 1.0 and w a /a = 1− s, where s, the selec- librium will be quite small, so 1− sqˆ 2 ≈ 1, and we have
tion coefficient, is the loss of fitness in the recessive
ˆ ˆ2
µ pˆ = −spq
homozygotes. We now can substitute these fitnesses in
our general expression for allele frequency change (see µ
qˆ =
Box 18-7) and obtain s
Let’s consider the equilibrium between selection and 18.6 BIOLOGICAL AND SOCIAL
mutation for the slightly more complicated case of a par-
tially dominant deleterious allele—that is, an allele with
APPLICATIONS
some deleterious effect in heterozygotes as well as its effect
in homozygotes. We will define h as the degree of domi- LO 18.8 Explain how population genetics informs many
issues facing modern societies.
nance of the deleterious allele. When h is 1, the deleterious
allele is fully dominant, and when h is 0, the deleterious
Just as the principles of physics guide engineers who design
allele is fully recessive. Then, the fitnesses are
bridges and jet airliners, so the principles of population
WA /A WA /a w a /a genetics touch all of our lives in many, if unseen, ways.
1 1− hs 1− s In Chapter 19, you’ll see how population genetics figures
prominently in the search for genes that contribute to
where a is a partially dominant deleterious allele. A deriva- disease risk in people, using concepts such as linkage dis-
tion similar to the one in Box 18-8 gives us equilibrium, described in this chapter. In this final section
µ of the chapter, we will examine three other areas in which
qˆ =
hs the principles of population genetics are being to applied to
issues affecting modern societies.
Here is an example. If µ = 10−6 and the lethal allele is not
totally recessive but causes a 5 percent reduction in fitness
in heterozygotes (s = 1.0, h = 0.05), then Conservation genetics
µ Conservation biologists attempting to save endangered
qˆ = = 2 × 10−5
hs wild species, and zookeepers attempting to maintain small
populations of captive animals, often perform population
This result is smaller by two orders of magnitude than
genetic analyses. Earlier in this chapter, we discussed how
the equilibrium frequency for the purely recessive case
a genetic bottleneck caused a loss of genetic variation in
described earlier. In general, then, we can expect deleteri-
the California condor and an increase in the frequency of
ous, completely recessive alleles to have frequencies much
a lethal form of dwarfism. Bottlenecks may also increase
higher than those of partly dominant alleles because the
the level of inbreeding in a population, perhaps leading to a
recessive alleles are protected in heterozygotes.
decline in fitness through inbreeding depression. The issue
KEY CONCEPTS The amount of genetic variation in pop- is complex, however, because inbreeding is not always asso-
ulations represents a balance between opposing forces: ciated with a decline in fitness. Inbreeding can sometimes
mutation and migration, which add new variation, versus drift help purge deleterious recessive alleles from a population.
and selection, which remove variation. Balancing selection Purifying selection is more effective at eliminating deleteri-
also serves to maintain variation in populations. As a result of ous recessive alleles with inbreeding since the homozygous
these processes, allele frequencies can reach equilibrium val- recessive class becomes more frequent in inbred popula-
ues, explaining why populations often maintain high levels of
tions. Thus, conservation biologists have debated whether
genetic variation.
they should attempt to maximize genetic diversity and
minimize inbreeding or deliberately subject zoo populations that an unaffected individual such as II-3 is a carrier is
to inbreeding with the goal of purging deleterious alleles. 2pq /(1 − q 2 ) = 0.049. If II-3 is a carrier, then there is a 12
To help address this question, researchers looked for chance she will transmit the disease allele to III-1. These
evidence of successful purging among zoo populations. are all independent probabilities, so we can use the product
Let’s define inbreeding depression as delta (δ) rule. The probability that III-1 will have cystic fibrosis is
wf 1 1
δ = 1− × × 0.049 = 0.003
w0 8 2
where w f is the fitness of inbred individuals and w0 the fit- The frequency of cystic fibrosis among Caucasians is
ness of non-inbred individuals. The value of δ will be pos- q 2 = (0.025)2 = 0.000625. These calculations tell us that
itive when there is a decline in fitness with inbreeding but individuals who have a first cousin with cystic fibrosis have
negative when fitness improves with inbreeding. Research- a 0.003 ÷ 0.000625 = 4.9 -fold higher risk of having a child
ers calculated δ for 119 zoo populations, including 88 spe- with the disease than members of the general population.
cies, and they found evidence that purging had improved Here is another application of population genetics for
fitness (negative values for δ) in 14 populations. Still, it assessing disease risk. Sickle-cell anemia, a recessive disease,
is not clear that deliberate inbreeding of zoo animals is has a frequency of about 0.25 percent, or 1 in 400, among
advisable. For one thing, although 14 of the 119 popula- African Americans (see Chapter 5). Applying the Hardy–
tions improved, the majority of the populations declined in Weinberg law, we estimate the frequency of the disease
fitness when inbred. Thus, if one starts with a small zoo allele (HbS ) as 0.05. What would be the expected frequency
population and purposely inbreeds the animals, a decline in of this disease among the offspring of African Americans
fitness is the most likely outcome. who are first cousins? Using the method described in Box
18-2, we calculate that the inbreeding coefficient (F) for the
offspring of first-cousin marriages is 1/16. In the earlier sec-
Calculating disease risks tion on inbreeding, we saw that the frequency of the homo-
In Chapter 2, we saw how alleles for genetic disorders zygotes when there is inbreeding is increased, as shown by
could be traced in pedigrees and we discussed how to cal- this equation:
culate the risk that a couple will have a child who inherits
f a /a = q 2 + pqF
such a disorder. Population genetic principles allow us to
extend this type of analysis. We will consider two examples. Using this equation, we obtain
The disease allele for cystic fibrosis (CF) occurs at a fre-
f (Hb S /Hb S ) = (0.05)2 + (0.05 × 0.95)(1/16) = 0.0055
quency of about 0.025 in Caucasians. In the pedigree for a
Caucasian family below, individual II-2 has a first cousin This represents a 2.2-fold increase in the risk of having a
(II-1) with cystic fibrosis. II-2 is married to an unrelated child with the disease for first-cousin marriages compared
Caucasian (II-3), and they are planning to have a child. to that in a marriage between unrelated individuals.
What is the chance that the child (III-1) will have cystic
fibrosis?
DNA forensics
Criminals can leave DNA evidence at the scene of a crime
I
1 4
in the form of blood, semen, hairs, or even buccal cells
2 3
from someone other than the suspect. This is what statis- Prob(A4 /A6 ) = 2pq = 2 × 0.03 × 0.05 = 0.003
ticians call the “null hypothesis,” or the hypothesis that is Prob(B1 /B7 ) = 2 × 0.01× 0.12 = 0.0024
considered true unless the evidence shows that it is very
unlikely (see Chapter 4). To perform the test, we calculate To combine these two probabilities, we need to make
the probability of observing a match between the evidence one more assumption. We need to assume that the two
and the suspect, given that the suspect and the person who loci are independent; that is, that the loci are at linkage
left the evidence are different individuals. Symbolically, we equilibrium. By making this assumption, we can apply the
write product rule for independent events (see Chapter 2) and
determine that
Prob(match | different individuals)
Prob(match | different individuals)
where “|” means “given.” If this probability is very small,
= Prob(A4 /A6 ) × Prob(B1 /B7 ) = 7.2 × 10−6
then we can reject the null hypothesis and argue in favor of
an alternative hypothesis: The evidence was left by the sus- Thus, the probability under the null hypothesis that the
pect. We never formally prove the suspect left the evidence evidence came from someone other than the suspect is
since there could be alternative hypotheses such as The evi- 7.2 × 10−6, or about 7 in a million. That is a small proba-
dence was left by the suspect’s identical twin. bility, and so the null hypothesis seems unlikely in this case.
To calculate the probability of observing a match However, if Prob(match | different individuals) were 0.1, then
between the evidence and the suspect if the evidence is from 10 percent of the population would be a match and could
a different individual, we need to know the frequencies of have left the evidence. In that case, we would not want to
the microsatellite alleles in the population. reject the null hypothesis.
Two microsatellites do not provide very much power to
A4 0.03
discriminate, so the FBI in the United States uses a set of
A6 0.05 20 microsatellites. Microsatellite loci typically have large
B1 0.01 numbers of alleles (10 to 20 or more); therefore, the num-
ber of possible genotypes based on 20 microsatellites is
B7 0.12
astronomically large. With 10 alleles per locus, there are
Prob(match | different individuals) is the same as the 55 possible genotypes at each locus and 5520, or 6.4 × 1034 ,
probability that the evidence came from a randomly chosen possible multilocus genotypes for 20 loci. The FBI has also
individual. We can calculate this probability using the allele assembled a database called CODIS (Combined DNA Index
frequencies above. First, we will assume that the Hardy– System) that contains the frequencies of different alleles at
Weinberg law applies and calculate the probability of being these loci in the population, including data specific to dif-
A4 /A6 at the first locus and B1 /B7 at the second: ferent ethnic groups and regions of the country.
SUMMARY
Population genetics seeks to understand the laws that gov- with relatives than expected by chance, then there will be
ern the amount of genetic variation within populations and an excess of homozygous genotypes throughout the entire
changes in genetic variation over time. The concept of the genome and the population becomes inbred. Even when
gene pool provides a model for thinking about the trans- local populations of a species conform to Hardy–Weinberg
mission of genetic variation from one generation to the next expectations, those populations are apt to be isolated from
for an entire population. Basic population genetic theory other populations at distant locations. Thus, a species often
starts with an idealized population that is infinite in size consists of a series of genetically distinct subpopulations;
and in which mating is random. In such a population, the that is, species show population genetic structure.
Hardy–Weinberg law defines the relationship between allele Several forces can add new variation to a population or
frequencies in the gene pool and genotype frequencies in the remove existing variation from it. Mutation is the ultimate
population. source of all genetic variation. Population geneticists have
Real populations deviate to various degrees from the determined reasonably precise estimates of the rate at which
Hardy–Weinberg model. One source of deviation comes in new mutations arise in populations. Migration can also
the form of nonrandom or assortative mating. If individuals bring new variation into a population. Migration results
preferentially mate with others who share a similar pheno- in some individuals who are genetically admixed, having
type, then there will be an excess of homozygotes at genes ancestry from multiple populations. Genetic recombina-
controlling that phenotype compared to Hardy–Weinberg tion can also add variation to populations by recombining
expectations. When individuals mate more frequently alleles into new haplotypes.
Two forces control the fate of genetic variation in pop- of genetic variation in populations. In this chapter, we
ulations. First, genetic drift is a random force that can lead have seen how research in population genetics has both
to the loss or fixation of an allele as a result of sampling developed the basic theory and collected a vast amount
error in finite populations. Drift is a strong force in small of data to achieve this goal. Our understanding of the
populations and a weak force in large ones. Second, natural population genetics of our own species is remarkably
selection drives changes in allele frequencies in populations detailed.
over time. Alleles that enhance the fitness of the individuals Finally, the methods and results of population genetics
that carry them will rise in frequency and can become fixed, both inform us about the evolutionary process and have
while deleterious alleles that reduce fitness will be purged practical applications to issues facing modern societies.
from the population. Population genetic theory and analyses play important roles
The fundamental goal of population genetics is to in the management of endangered species, the identification
understand the relative contributions made by mat- of perpetrators of crimes, plant and animal breeding, and
ing systems, mutation, migration, recombination, drift, assessing the risks that a couple will have a child with a
and natural selection to the amount and distribution disease condition.
KEY TERMS
absolute fitness (p. 629) haplotype network (p. 607) neutral allele (p. 626)
adaptation (p. 628) Hardy–Weinberg equilibrium neutral evolution (p. 626)
allele frequency (p. 610) (p. 611) nucleotide diversity (p. 620)
artificial selection (p. 632) Hardy–Weinberg law (p. 610) number of haplotypes (NH) (p. 620)
balancing selection (p. 631) heterozygosity (H) (p. 620) population (p. 604)
bottleneck (p. 627) identical by descent (IBD) (p. 615) population genetics (p. 604)
common SNP (p. 605) inbreeding (p. 614) population structure (p. 614)
Darwinian fitness (p. 629) inbreeding coefficient (F) (p. 615) positive assortative mating (p. 613)
directional selection (p. 630) inbreeding depression (p. 615) positive selection (p. 630)
disassortative mating (p. 613) isolation by distance (p. 614) purifying selection (p. 630)
discovery panel (p. 605) linkage disequilibrium (LD) (p. 623) random genetic drift (p. 623)
fixed (p. 618) linkage equilibrium (p. 622) rare SNP (p. 605)
founder effect (p. 627) locus (p. 604) relative fitness (p. 629)
gene diversity (GD) (p. 620) microsatellite (p. 605) segregating sites (S) (p. 620)
gene flow (p. 621) migration (p. 621) selection coefficient (s) (p. 633)
gene pool (p. 609) molecular clock (p. 626) single nucleotide polymorphism
genetic admixture (p. 622) mutation rate (μ) (p. 621) (SNP) (p. 604)
genotype frequency (p. 610) natural selection (p. 629)
haplotype (p. 606) negative assortative mating (p. 613)
SOLVED PROBLEMS
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
WORKING WITH THE FIGURES class would you expect to observe under Hardy–
(The first 6 questions require inspection of text figures.) Weinberg equilibrium?
1. Which individual in Figure 18-3 has the most hetero- 11. In a randomly mating laboratory population of
zygous loci, and which individual has the fewest? Drosophila, 4 percent of the flies have black bod-
ies (encoded by the autosomal recessive b ), and
2. Suppose that the seven chromosomes in Figure 18-4a
96 percent have brown bodies (the wild type,
represent a random sample of chromosomes from a
encoded by B ). If this population is assumed to be
population.
in Hardy–Weinberg equilibrium, what are the allele
a. Calculate gene diversity (GD) separately for the frequencies of B and b and the genotypic frequencies
indel, the microsatellite locus, and the SNP at of B /B and B /b?
position 3.
12. In a population of a beetle species, you notice that
b. If the sequence was shortened so that you had data there is a 3:1 ratio of shiny to dull wing covers. Does
only for positions 1 through 24, how many haplo- this ratio prove that the shiny allele is dominant?
types would there be? (Assume that the two states are caused by two alleles
c. Calculate the linkage disequilibrium parameter (D) of one gene.) If not, what does it prove? How would
between the SNPs at positions 29 and 33. you elucidate the situation?
3. Looking at Figure 18-6 , can you count how many 13. Cystic fibrosis (CF) is an autosomal recessive disor-
mitochondrial haplotypes were carried from Asia into der that occurs relatively frequently among people of
the Americas? European descent. In an Amish community in Ohio,
medical researchers reported the occurrence of cys-
4. In Figure 18-13, the “unrelated” (blue) column for
tic fibrosis (CF) as being 1/569 live births. Using the
Japan is higher than the “unrelated” column for
Hardy–Weinberg rule, estimate the frequency of carri-
France. What does this tell you?
ers of the disease allele in this Amish population.
5. In Figure 18-14, some individuals have unique SNP
14. The relative fitness values of three genotypes are
alleles—for example, the T allele at SNP4 occurs only
w A/ A = 1.0, w A /a = 1.0, and w a /a = 0.7.
in individual 12. Can you identify two individuals each
of whom have unique alleles at two SNPs? a. If the population starts at the allele frequency
p = 0.5, what is the value of p in the next
6. Looking at Figure 18-20, do people of the Middle East
generation?
tend to have higher or lower levels of heterozygosity
compared to the people of East Asia? Why might this b. What is the predicted equilibrium allele frequency
be the case? if the rate of mutation of A to a is 2 × 10−5?
15. A /A and A /a individuals are equally fertile. If 0.1 per-
BASIC PROBLEMS cent of the population is a /a , what selection pressure
exists against a /a if the A → a mutation rate is 10−5?
7. What are the forces that can change the frequency of
Assume that the frequencies of the alleles are at their
an allele in a population, and what effect does each
equilibrium values.
have on variation in a population?
16. When alleles at a locus act in a semidominant fash-
8. What assumptions are made when using the Hardy–
ion on fitness, the relative fitness of the hetero-
Weinberg formula to estimate genotypic frequencies
zygote is midway between the two homozygous
from allele frequencies?
classes. For example, genotypes with semidominance
9. In a population of mice, there are two alleles of the at the A locus might have these relative fitnesses:
A locus (A1 and A2). Tests showed that, in this popu- w A /A = 1.0, w A /a = 0.9, and w a /a = 0.8.
lation, there are 384 mice of genotype A1 /A1 , 210 of
a. Change one of these fitness values so that a
A1 /A2 , and 260 of A2 /A2. What are the frequencies of
becomes a deleterious recessive allele.
the two alleles in the population?
b. Change one of these fitness values so that A
10. In a natural population of Drosophila melanogaster,
becomes a favored dominant allele.
the alcohol dehydrogenase gene has two alleles called
F (fast) and S (slow) with frequencies of Adh-F at 0.75 17. If the recessive allele for an X-linked recessive disease
and Adh-S at 0.25. In a sample of 480 flies from this in humans has a frequency of 0.02 in the population,
population, how many individuals of each genotypic what proportion of individuals in the population will
have the disease? Assume that the population is 50:50 a. If the inbreeding coefficient for the common ances-
male:female. tor (A) in Figure 18-12 is 1/2, what is the inbreed-
18. Red-green color blindness is an X-linked recessive dis- ing coefficient of I?
order in humans caused by mutations in one of the b. If the inbreeding coefficient of individual I in Fig-
genes that encodes the light-sensitive protein, opsin. If ure 18-12 is 1/8, what is the inbreeding coefficient
the mutant allele has a frequency of 0.08 in the pop- of the common ancestor, A?
ulation, what proportion of females will be carriers? 27. Consider 10 populations that have the genotype fre-
Assume that the population is 50:50 male:female. quencies shown in the following table:
19. Is a new neutral mutation more likely to reach fixation
in a large or small population? Population A /A A /a a /a
20. It seems clear that inbreeding causes a reduction in fit- 1 1.0 0.0 0.0
ness. Can you explain why? 2 0.0 1.0 0.0
21. In a population of 50,000 diploid individuals, what is 3 0.0 0.0 1.0
the probability that a new neutral mutation will ulti-
4 0.50 0.25 0.25
mately reach fixation? What is the probability that it
will ultimately be lost from the population? 5 0.25 0.25 0.50
22. Inbreeding in a population causes a deviation from 6 0.25 0.50 0.25
Hardy–Weinberg expectations such that there are 7 0.33 0.33 0.33
more homozygotes than expected. For a locus with 8 0.04 0.32 0.64
a rare deleterious allele at a frequency of 0.04, what
9 0.64 0.32 0.04
would be the frequency of homozygotes for the delete-
rious allele in populations with inbreeding coefficients 10 0.986049 0.013902 0.000049
of F = 0.0 and F = 0.125?
a. Which of the populations are in Hardy–Weinberg
23. Sickle-cell anemia is a recessive autosomal disorder
equilibrium?
that is caused by an amino acid substitution in the
β-hemoglobin protein. The DNA mutation underly- b. What are p and q in each population?
ing this substitution is a SNP that alters a GAG codon c. In population 10, the A → a mutation rate is dis-
for the amino acid glutamate to a GTG that codes a covered to be 5 × 10−6. What must be the fitness of
valine. The frequency of sickle-cell anemia among Afri- the a /a phenotype if the population is at equilib-
can Americans is about 1/400. What is the frequency rium?
of this GTG codon in the β-hemoglobin gene among
d. In population 6, the a allele is deleterious; further-
African Americans?
more, the A allele is incompletely dominant; so
24. You have a sample of 10 DNA sequences of 100 bp in A /A is perfectly fit, A /a has a fitness of 0.8, and a /a
length from a section of highly conserved gene from has a fitness of 0.6. If there is no mutation, what
10 individuals of a species. The 10 sequences are almost will p and q be in the next generation?
entirely identical; however, each sequence carries one
28. The hemoglobin β gene (Hb) has a common allele (A)
unique SNP not found in any of the others. What is the
of a SNP (rs334) that encodes the Hb A form of (adult)
nucleotide diversity for this sample of sequences?
hemoglobin and a rare allele (T) that encodes the sick-
ling form of hemoglobin, HbS. Among 571 residents of
CHALLENGING PROBLEMS
a village in Nigeria, 440 were A /A and 129 were A /T ,
25. Figure 18-14 presents haplotype data for the G6PD and 2 were T /T individuals were observed. Use the chi-
gene in a worldwide sample of people. square test to determine whether these observed geno-
a. Draw a haplotype network for these haplotypes. typic frequencies fit Hardy–Weinberg expectations.
Label the branches on which each SNP occurs. 29. A population has the following gametic frequencies at
b. Which of the haplotypes has the most connections two loci: AB = 0.4, Ab = 0.1, aB = 0.1, and ab = 0.4.
to other haplotypes? If the population is allowed to mate at random until
linkage equilibrium is achieved, what will be the
c. On what continents is this haplotype found? expected frequency of individuals that are heterozy-
d. Counting the number of SNPs along the branches gous at both loci?
of your network, how many differences are there 30. Two species of palm trees differ by 50 bp in a 5000-bp
between haplotypes 1 and 12? stretch of DNA that is thought to be neutral. The
26. Figure 18-12 shows a pedigree for the offspring of a mutation rate for these species is 2 × 10−8 substitutions
half-sib mating. per site per generation. The generation time for these
species is five years. Estimate the time since these spe- d. Using a spreadsheet computer software pro-
cies had a common ancestor. gram, make a graph of the decline in D over
31. Color blindness in humans is caused by an X-linked 10 generations.
recessive allele. Ten percent of the males of a large 36. Allele B is a deleterious autosomal dominant. The
and randomly mating population are color-blind. A frequency of affected individuals is 4.0 × 10−6 . The
representative group of 1000 people from this popu- reproductive capacity of these individuals is about
lation migrates to a South Pacific island, where there 30 percent that of normal individuals. Estimate µ ,
are already 1000 inhabitants and where 30 percent the rate at which b mutates to its deleterious allele B.
of the males are color-blind. Assuming that Hardy– Assume that the frequencies of the alleles are at their
Weinberg equilibrium applies throughout (in the two equilibrium values.
original populations before the migration and in the 37. What is the equilibrium heterozygosity for a SNP in
mixed population immediately after the migration), a population of 50,000 when the mutation rate is
what fraction of males and females can be expected to 3 × 10−8?
be color-blind in the generation immediately after the
arrival of the migrants? 38. Of 31 children born of father–daughter matings,
6 died in infancy, 12 were very abnormal and died in
32. Using pedigree diagrams, calculate the inbreeding childhood, and 13 were normal. From this informa-
coefficient (F) for the offspring of (a) parent–offspring tion, calculate roughly how many recessive lethal genes
matings; (b) first-cousin matings; (c) aunt–nephew we have, on average, in our human genomes. (Hint:
or uncle–niece matings; (d) self-fertilization of a If the answer were 1, then a daughter would stand a
hermaphrodite. 50 percent chance of carrying the lethal allele, and the
33. A group of 50 men and 50 women establish a colony probability of the union’s producing a lethal combina-
on a remote island. After 50 generations of random tion would be 1/2 × 1/4 = 1/8. So, 1 is not the answer.)
mating, how frequent would a recessive trait be if it Consider also the possibility of undetected fatalities
were at a frequency of 1/500 back on the mainland? in utero in such matings. How would they affect your
The population remains the same size over the 50 gen- result?
erations, and the trait has no effect on fitness. 39. The B locus has two alleles B and b with frequencies
34. Figure 18-22 shows 10 haplotypes from a popula- of 0.95 and 0.05, respectively, in a population in the
tion before a selective sweep and another 10 haplo- current generation. The genotypic fitnesses at this
types many generations later after a selective sweep locus are wB / B = 1.0, wB / b = 1.0, and wb / b = 0.0.
has occurred for this chromosomal region. There are a. What will the frequency of the b allele be in two
11 loci defining each haplotype, including one with a generations?
red allele that was the target of selection. In the figure,
two loci are designated as A and B. These loci each b. What will the frequency of the b allele be in two gen-
have two alleles: one black and the other gray. Calcu- erations if the fitnesses were wB / B = 1.0, wB / b = 0.0,
late the linkage disequilibrium parameter (D) between and wb / b = 0.0?
A and B, both before and after the selective sweep. c. Explain why there is a difference in the rate of
What effect has the selective sweep had on the level of change for the frequency of the b allele under parts
linkage disequilibrium? a and b of this problem.
35. The recombination rate (r) between linked loci A and 40. The sd gene causes a lethal disease of infancy in
B is 0.10. In a population, we observe the following humans when homozygous. One in 100,000 newborns
haplotypic frequencies: die each year of this disease. The mutation rate from
Sd to sd is 2 × 10−4 . What must the fitness of the het-
AB 0.40 erozygote be to explain the observed gene frequency in
aB 0.10 view of the mutation rate? Assign a relative fitness of
1.0 to Sd/Sd homozygotes. Assume that the population
Ab 0.10 is at equilibrium with respect to the frequency of sd.
Ab 0.40
41. If we define the total selection cost to a population
of deleterious recessive genes as the loss of fitness per
a. What is the level of linkage disequilibrium as mea- individual affected (s) multiplied by the frequency of
sured by D in the present generation? affected individuals (q 2 ), then selection cost = sq 2 .
b. What will D be in the next generation? a. Suppose that a population is at equilibrium
c. What is the expected frequency of the Ab haplo- between mutation and selection for a deleterious
type in the next generation? recessive allele, where s = 0.5 and µ = 10−5. What is
the equilibrium frequency of the allele? What is the alleles (w a ) (see Box 18-7). Set the mean fitness of
selection cost? the A alleles (w A ) equal to the mean fitness of the
b. Suppose that we start irradiating individual mem- a alleles (w a ). Solve the resulting equation for the
bers of the population so that the mutation rate frequency of the A allele. This is the expression for
doubles. What is the new equilibrium frequency of the equilibrium frequency of A (pˆ ).
the allele? What is the new selection cost? b. Using the expression that you just derived, find p̂
c. If we do not change the mutation rate but we lower when s = 0.2 and g = 0.8.
the selection coefficient to 0.3 instead, what hap-
pens to the equilibrium frequency and the selection GENETICS AND SOCIETY
cost? Genome-wide SNP data from companies like 23andMe
42. Balancing selection acts to maintain genetic diversity allow people to compare their genotypes to other people’s
at a locus since the heterozygous class has a greater genotypes to identify distant relatives such as third cous-
fitness than the homozygous classes. Under this form ins. A free Web site (GEDmatch.com) allows you to upload
of selection, the allele frequencies in the population your SNP genotypes and search for relatives in a public
approach an equilibrium point somewhere between 0 database. In 2018, police collected SNP genotypes from
and 1. Consider a locus with two alleles A and a with DNA left by the “Golden State Killer” at one of his crime
frequencies p and q, respectively. The relative geno- scenes. Using GEDmatch, police identified relatives of the
typic fitnesses are shown below, where s and g are the person who left the DNA at the crime scene that led them
selective disadvantages of the two homozygous classes. to identify and arrest Joseph DeAngelo as the Golden State
Killer. Do you see any ethical issues in use of public data-
Genotype A /A A /a a /a bases to track criminals? Do you think people might be
Relative fitness 1– s 1 1– g reluctant to upload their personal data to GEDmatch if they
thought it might lead to the arrest of a relative? Are there
a. At equilibrium, the mean fitness of the A alleles ways in which making your own DNA data public could
(w A ) will be equal to the mean fitness of the a cause unjustified harm to you or your relatives?
L
ook at almost any large group of men or women analysis is inadequate for the study of complex traits. If
and you’ll notice a considerable range in their progeny cannot be sorted into categories with expected
heights—some are short, some tall, and some about ratios, then the Mendelian approach has little utility for
average. Kareem Abdul-Jabbar, a star basketball center the analysis of complex traits. In response to this problem,
of the 1970s and 1980s, was a towering 7 feet, 2 inches geneticists developed a set of mathematical models and sta-
tall, whereas Willie Shoemaker, a renowned jockey who tistical methods for the analysis of complex traits. Through
won the K entucky Derby four times, was a mere 4 feet, the application of these analytical methods, geneticists have
11 inches. You might also have noticed that in some fam- made great strides in understanding complex traits. The
ilies, the parents and their adult children are all on the subfield of genetics that develops and applies these meth-
tall side, whereas in other families, the parents and adult ods to understand the inheritance of complex traits is called
children are all fairly short. Such observations suggest that quantitative genetics.
genes play a role in determining our heights. Still, people At the heart of the field of quantitative genetics is the
do not segregate cleanly into tall and short categories as we goal of defining the genetic architecture of complex traits.
saw for Mendel’s pea plants. At first inspection, continu- Genetic architecture is a description of all of the genetic fac-
ous traits, such as height, do not appear to follow Mendel’s tors that influence a trait. It includes the number of genes
laws despite the fact that they are heritable. affecting the trait and the relative contribution of each
Traits such as height that show a continuous range of gene. Some genes may have a large effect on the trait, while
variation and do not behave in a simple Mendelian fash- others have only a small effect. As we will see in this chap-
ion are known as quantitative or complex traits. The term ter, genetic architecture is the property of a specific pop-
complex trait is often preferred because variation for such ulation and can vary among populations of a species. For
traits is governed by a “complex” of genetic and environ- example, the genetic architecture of a trait such as systolic
mental factors. How tall you are is partly explained by the blood pressure in humans differs among different popula-
genes you inherited from your parents and partly by envi- tions. This is because different alleles segregate in different
ronmental factors such as how well you were nourished as populations and different populations experience different
a child. Teasing apart the genetic and environmental con- environments; therefore, different populations are apt to
tributions to an individual phenotype is a substantial chal- have different architectures for many traits.
lenge, but geneticists have a powerful set of tools to meet it. Understanding the inheritance of complex traits is
In the early 1900s, when Mendel’s laws were rediscov- one of the most important challenges facing geneticists
ered, controversy arose about whether these laws were in the twenty-first century. Complex traits are of para-
applicable to continuous traits. A group known as the bio- mount importance in medical and agricultural genetics.
metricians discovered that there are correlations between rel- For humans, blood pressure, body weight, susceptibility to
atives for continuous traits such that tall parents tend to have depression, serum cholesterol levels, and the risk of devel-
tall children. However, the biometricians saw no evidence oping cancer or other disorders are all complex traits. For
that such traits followed Mendel’s laws. Some biometricians crop plants, yield, resistance to pathogens, ability to toler-
concluded that Mendelian loci do not control continuous ate drought stress, efficiency of fertilizer uptake, and even
traits. On the other hand, some adherents of Mendelism flavor are all complex traits. For livestock, milk produc-
thought continuous variation was unimportant and could tion in dairy cows, muscle mass in beef cattle, litter size in
be ignored when studying inheritance. By 1920, this contro- pigs, and egg production in chickens are all complex traits.
versy was resolved with the formulation of the multifactorial Despite the importance of such traits, we know far less
hypothesis. This hypothesis proposed that continuous traits about their inheritance than we do about the inheritance of
are governed by a combination of multiple Mendelian loci, simply inherited traits such as cystic fibrosis or sickle-cell
each with a small effect on the trait, and environmental fac- anemia.
tors. The multifactorial hypothesis brought quantitative traits In this chapter, we will explore the inheritance of com-
into the realm of Mendelian genetics. plex traits. We will begin with a review of some basic sta-
Although the multifactorial hypothesis provided a sensi- tistical concepts. Next, we will develop the mathematical
ble explanation for continuous variation, classic Mendelian model used to connect the action of genes inside the cell
644
with the phenotypes we observe at the level of the whole individual may or may not have type 2 diabetes. However,
organism. Using this model, we will then show how quan- type 2 diabetes does not follow simple Mendelian rules
titative geneticists partition the phenotypic variation in a or produce Mendelian ratios in pedigrees. Rather, there
population into the parts that are due to genetic and envi- are multiple genetic and environmental factors that place
ronmental factors. We will review the methods used by someone at risk of developing this disease. Individuals who
plant and animal breeders to predict the phenotype of off- have a certain number of risk factors will exceed a thresh-
spring from the phenotype of their parents. Finally, we will old and develop the disease. Type 2 diabetes is a form of a
see how a combination of the statistical analysis and molec- categorical trait called a threshold trait. Type 2 diabetes has
ular markers can be used to identify the specific genes that complex inheritance.
control quantitative traits. Another type of trait is a meristic trait, or counting
trait, which takes on a range of discrete values. An example
would be clutch size in birds. A bird can lay 1, 2, 3, or more
19.1 MEASURING QUANTITATIVE eggs, but it cannot lay 2.49 eggs. Meristic traits are quan-
titative, but they are restricted to certain discrete values.
VARIATION They do not take on a continuous range of values. Meristic
traits usually have complex inheritance.
LO 19.1 Understand how quantitative genetics uses Quantitative geneticists seek to understand the inher-
mathematical models and statistics to investigate
itance of traits that show complex inheritance resulting
complex traits.
from a mix of genetic and environmental factors. They may
LO 19.2 Analyze data to assess the amount and distribution
of trait variation in populations. investigate traits that are categorical, meristic, or continu-
ous. The emphasis is on the type of inheritance—complex.
To study the inheritance of quantitative traits, we need For this reason, the term complex trait is often preferred
some basic statistical tools. In this section, we will intro- to continuous or quantitative trait because it includes all
duce the mean (or average), which can be used to describe the types of traits with which quantitative genetics is con-
differences between groups, and the variance, which can be cerned. Any biological phenomenon for which variation
used to quantify the amount of variation that exists within exists may show complex inheritance and can be studied as
a group. We will also discuss the normal distribution, a complex trait. Thus, size and shape of structures, enzyme
which is central to understanding quantitative variation in kinetics, mRNA levels, circadian rhythms, and bird songs
populations. But before discussing the statistical tools, let’s can all be treated as complex traits.
define the different types of complex trait variation that
KEY CONCEPT A complex trait is any trait that does not
can occur in a population.
show simple Mendelian inheritance. A complex trait can be
either a categorical trait such as the presence or absence of
a disease condition, or a continuously variable trait such as
Types of traits and inheritance
height in humans.
A continuous trait is one that can take on a potentially
infinite number of states over a continuous range. Height
in humans is a good example. People can range from about The mean
140 cm to 230 cm in height. If we measured height precisely, When quantitative geneticists study the inheritance of a
then the number of possible heights is infinite. For example, trait, they work with a particular group of individuals, or
a person might be 170 cm tall, or 170.2 cm, or 170.02 cm. population. For example, we might be interested in the
Continuous traits typically have complex inheritance involv- inheritance of height for the population of adult men in
ing multiple genes plus environmental factors. Shanghai, China. Here, we are using “population” to denote
For some traits, the individuals in a population can be a group that shares certain features in common, such as age,
sorted into discrete groups or categories. Such traits are sex, ethnicity, or geographic origin. Since there are more
known as categorical traits. Examples include purple ver- than 5 million adult men in Shanghai, determining each of
sus white flowers or tall versus short stems for Mendel’s their heights would be a herculean task. Therefore, quanti-
pea plants, as seen in Chapter 2. Categorical traits often tative geneticists typically study just a subset or sample of
exhibit simple inheritance such that the progeny of crosses the full population. The sample should be randomly chosen
segregate into standard Mendelian ratios such as 3:1 for a such that each of the 5 million men has an equal chance of
single gene or 15:1 for two genes. The inheritance is simple being included in the sample. If the sample meets this crite-
because only one or two genes are involved and the envi- rion, then we can use measurements made on the sample to
ronment has little or no effect on the phenotype. make inferences about the entire population.
Some categorical traits do not show simple inheritance. Using the example of height for men from Shanghai,
These include many disease conditions in humans. In medi- we can describe the population using the mean or average
cal genetics, individuals can be classified into the categories value for the trait. We select a random sample of 100 men
“affected” or “not affected” by a disease. For example, an from the population and measure their heights. Some of the
men might be 166 cm tall, others 172 cm tall, and so forth. add all n observed values of X for i = 1, 2, through n.
To calculate the mean, we simply sum all the individual (Often, the n above Σ and the i = 1 below Σ are omitted to
measurements and divide the sum by the size of the sample simplify the appearance of equations.)
(n), which in this case is 100. For the data in Table 19-1, the There is a distinction made between the mean of a sam-
result would be 170 cm, or 5 feet, 7 inches. Since we have a ple (X) and the true mean of the population. To learn the
random sample, we can infer that the average height in the true mean for the height of men in Shanghai, we would
entire population is 170 cm. need to determine the height of each and every man. The
Height is a random variable, which means it can take true mean is symbolized by the Greek letter µ , so that
on different values, and when we select someone at random we have different symbols for the sample and population
from the population, the value we observe is governed by means.
an element of chance. Random variables are usually repre- Here is another way to calculate the mean, which is
sented by the letter X in statistics. We have measurements often quite useful. We can add the products of each class of
for X1 , X2 , X3 , . . . X100 for the n = 100 men in the sample. values of X in the data set times the frequency of that class
Symbolically, we can express the mean as in the data set. This operation is symbolized as
1 n k
X= ∑ Xi
n i=1
X = ∑ fi Xi
i =1
where X represents the sample mean. The uppercase Greek where fi is the frequency of the ith class of observations, Xi
letter sigma (Σ) is the summation sign, indicating that we is the value of the ith class, and there are a total of k classes.
For the data in Table 19-1, one man of the 100 (f = 0.01)
is 156 cm tall, two men (f = 0.02) are 157 cm tall, and so
forth, so we can calculate the sample mean as
TABLE 19-1 Simulated Data for the Heights of
100 Men from Shanghai, China X = (0.01× 156) + (0.02 × 157) + ... + (0.02 × 184) = 170
Height (cm) Count Frequency × Height The mean is useful for both describing populations and
156 1 1.56 comparing differences between populations. For example,
157 2 3.14 men in urban areas of China are on average 170 cm tall,
158 1 1.58 while men in rural areas of China are 166 cm tall. These
values were calculated using samples drawn from each
159 2 3.18
region. One question that a quantitative geneticist might
160 1 1.60
ask about the observed difference in height between rural
161 1 1.61 and urban Chinese men is the following: Is the difference
162 2 3.24 due to genetic factors, or is it due to differences in nutri-
164 7 11.48 tion, health care, or other environmental factors? Later in
165 7 11.55 the chapter, we will see how quantitative geneticists tease
apart genetic versus environmental contributions to a trait.
166 1 1.66
Lastly, here is another helpful notation from statistics
167 6 10.02 that can be used to define the mean. The mean of a random
168 9 15.12 variable, X, is the expectation or expected value of that
169 7 11.83 random variable. The expected value is the average of all
170 9 15.30 the values we would observe if we measured X many times.
The expectation is symbolized by E, and we write E(X) to
171 5 8.55
signify “the expected value of X.” Symbolically, we write
172 5 8.60
173 6 10.38 E(X) = X
174 5 8.70 We will use the notation of expectation in several places in
175 6 10.50 this chapter.
176 3 5.28
177 4 7.08 The variance
178 2 3.56 Besides the mean, we also need a measure of how much
179 2 3.58 variation exists in populations. We can create a visual rep-
180 2 3.60 resentation of the variation by plotting the count or fre-
181 2 3.62 quency of each height class. Figure 19-1 shows such a plot
for our simulated height data for 100 men from Shanghai.
184 2 3.68
The x-axis shows different height classes, and the y-axis
Sum 100 170.00 shows the count or frequency of each class. In this figure,
Quantitative traits are described Since deviations with negative values form positive squares,
by a frequency distribution both negative and positive deviations will contribute pos-
itively to the sum of squares. The variance is the mean of
35
the squared deviations (or the sum of squares divided by n).
30 0.3 Symbolically, we express the population variance as
1
∑ (Xi − X)2
Number of individuals
25 VX =
n i
Frequency
20 0.2 1
= ∑ (xi )2
n i
15
where VX denotes the variance of X. The population vari-
10 0.1 ance is sometimes symbolized using the lowercase Greek
letter sigma squared (σ2 ). In statistics, there is also a dis-
5
tinction made between the population variance (σ2 ) and
0 0 the sample variance (s2 ) . The latter is calculated by dividing
the sums of squares by n − 1 rather than n to correct a bias
8
6
15
16
16
17
17
17
18
18
caused by small sample size. For simplicity, we will use the
5–
9–
3–
7–
1–
5–
9–
3–
15
15
16
16
17
17
17
18
Height (cm)
population variance and the formula above throughout this
chapter.
FIGURE 19-1 Frequency histogram of simulated data for the height There are several points to understand about the vari-
of adult men from Shanghai, China. ance. First, the variance provides a measure of dispersion
about the mean. When the variance is high, the individual
values are spread farther apart from the mean; when it is
the men were binned into 4-cm groups, for example, from
low, then the individual values cluster closer to the mean.
155 to 158 cm. This type of graph is called a frequency
Second, the variance is measured in squared units such
h istogram. If the values are clustered tightly around the
that if we measure human height in centimeters, then the
mean, then there is less variation, and if the values are
variance would be in centimeters2. Third, the variance can
spread out along the x-axis, there is greater variation.
range from 0.0 to infinity. Fourth, the variance is equal to
We can quantify the amount of variation in a popu-
the expected value of the squared deviation (x2 ) or E (x2 ).
lation using a statistical measure called the variance. The
The variance of quantitative traits is measured in
variance measures the extent to which individuals in the
squared units. These squared units have desirable mathe-
population deviate from the population mean. If all 100
matical properties as we will see below; however, they do
men in our sample had heights very close to the mean,
not make intuitive sense. If we measure weight in kilo-
then the variance would be small. If their heights deviated
grams, then the variance would be in kilograms2 , which has
greatly from the mean, the variance would be large.
no clear meaning. Therefore, another statistic used to quan-
Since the variance is a measure of deviation from the
tify the extent of deviation from the mean in a population
mean, let’s define deviation mathematically. Knowing
is the standard deviation (σ), which is the square root of the
the mean value for the random variable X, we can calculate
variance:
the deviation of each individual from the mean by subtract-
ing X from the individual observations. We will represent σ = σ2
the deviations by a lowercase x:
The standard deviation is expressed in the same units as the
x = X−X trait itself, so its meaning is more intuitive. We will use the
standard deviation in the description of traits below.
Some individuals will have X values above the mean, and
they will have a positive deviation. Others will have X val-
ues below the mean, and they will have a negative devia-
The normal distribution
tion. For the population overall, the expected value of x is Even if you have never taken a statistics course, you likely
0, or E(x) = 0. have heard of the normal distribution, also known as the
To measure the amount of variation for X in the popula- “bell curve” in popular culture. The normal distribution is
tion, we use the variance, which is the mean of the squared remarkably useful in biology in general and quantitative
deviations. First, we calculate the sum of the squared devia- genetics in particular because the frequency distribution
tions (or sum of squares, for short) as for many biological traits approximates a normal curve.
For this reason, geneticists can take advantage of several
sum of squares = ∑ (Xi − X)2 features of the normal distribution to describe quantitative
i
traits and dissect the underlying genetics.
= ∑ (xi )2 The normal distribution is a continuous frequency dis-
i tribution similar to the frequency histogram shown in
The use of normal curves to describe the distribution Here is an example using height data for 660 women from
of a trait in a population the United States collected by the Centers for Disease Control
and Prevention. The frequency histogram shows the classic
(a)
“bell curve” shape with the peak near the mean value of
50 164.4 cm and the off-mean values distributed symmetrically
Number of individuals
0.06 centage would fall above 177 cm. However, it would still be
true that only 2 percent would be more than 2σ above the
0.04 mean, or 180.4 cm [(164.4 + (2 × 8)].
=8
0.02 KEY CONCEPT The field of quantitative genetics studies the
inheritance of complex traits using some basic statistical tools
0.00 including the mean, variance, standard deviation, and normal
140 150 160 170 180
distribution.
Height (cm)
FIGURE 19-2 (a) Frequency histogram of actual data for the height
of adult women from the United States. The red line represents the
normal curve fit to these data with a mean of 164.4 cm and standard
deviation of 6.18 cm. (b) Normal curve for the height of U.S. 19.2 A SIMPLE GENETIC MODEL
women showing the predicted percentages of women who will
fall within different numbers of standard deviations from the mean.
FOR QUANTITATIVE TRAITS
(c) Normal curves with the same mean (164.4 cm) but different
standard deviations, showing the effect of the standard deviation on LO 19.3 Assess the relative contributions of genetic and
the shape of the curve. environmental factors to phenotypic traits.
Figure 19-1. The normal distribution applies to continuous A mathematical model is a simplified representation of
traits. As mentioned in the previous section, continuous traits a complex phenomenon. Models allow us to describe a
can take on an infinite number of values. A person might be phenomenon in terms of the variables that influence it
170 cm tall, or 170.2 cm, or 170.02 cm, and so forth. For and then to use the model to make predictions about the
such traits, the expected frequency of different trait values is state of the phenomenon under different values for these
better represented by a curve than by a frequency histogram. variables. In this section, we will define the mathematical
For the normal distribution, the shape of the curve is deter- model used by quantitative geneticists to study complex
mined by two factors—the mean and the standard deviation. traits.
The exceptional height of Yao Ming Shanghai. Twenty-one years later, we locate this army of
Yao Ming clones, measure their heights, and determine that
their average height is 212 cm. The expectation of e over
the many environments in which the Yao Ming clones were
reared is 0. In some households, the clones get a positive
environment (+e) and in others a negative environment (−e).
Overall, E(e) = 0. Thus, the mean for the clones minus the
population mean equals Yao Ming’s genotypic deviation,
or g = (212 − 170) = 42 cm. The remaining 17 cm of his
remarkable 59-cm phenotypic deviation is e for the specific
environment in which the real Yao Ming was raised. Plug-
ging these values into the equation, we obtain
229 = 170 + 42 + 17
We conclude that Yao Ming’s exceptional height is mostly
due to exceptional genetics, but he also experienced an
FIGURE 19-3 Former basketball star center Yao Ming, who stands environment that boosted his height.
7 feet, 6 inches, talks with retired golf star Gary Player, who is 5 feet, Although our imaginary experiment of cloning Yao Ming
6 inches. [Power Sport Images/Getty Images.] is far-fetched, many plant species and some animal species
can be clonally propagated with ease. For example, one can
Genetic and environmental deviations use “cuttings” of an individual plant to produce multiple
We will now examine how phenotypes can be decomposed genetically identical individuals. Another way of creating
into their genetic and environmental contributions, using as genetically identical individuals is by producing inbred lines
an example the height of Yao Ming, the former center for or stocks that are homozygous at all loci throughout their
the Houston Rockets basketball team. Yao Ming stands out entire genomes. Like clones, all individuals of an inbred line
at 229 cm, or 7 feet, 6 inches (Figure 19-3). That’s right: Yao are genetically identical to each. By using clones or inbred
Ming is nearly two feet taller than the average man from lines, geneticists can estimate the genetic and environmental
Shanghai, which happens to be Yao Ming’s hometown. As contributions to a trait by rearing the clones in randomly
for all of us, Yao Ming’s height is the combined result of assigned environments. Here is an example.
his genotype and the environment in which he was raised. Table 19-2 (experiment I) shows simulated data for 10
Let’s do an imaginary experiment and see how we can tease inbred strains of maize that were grown in three different
apart the genetic and environmental contributions to Yao environments and scored for the number of days between
Ming’s exceptional height. planting and the time that the plants first shed pollen. The
First, we will define a simple mathematical model that overall mean is 70 days. Let’s consider line A when grown
can be applied to any quantitative trait. The value of a in environment 1. The mean for all lines in environment 1
trait (X) for an individual member of a population can be is 68, or 2 less than the overall mean, so e for environment
expressed in terms of the population mean and deviations 1 is −2. The mean line A over all three environments is 64,
from the mean due to genetic (g) and environmental (e) or 6 less than the overall mean, so g for line A is −6. Putting
factors. these two values together, we decompose the phenotype of
line A when grown in environment 1 as
X = X+g+e
62 = 70 + (−6) + (−2)
We are using lowercase g and e for the genetic and envi-
ronmental deviations, just as we used a lowercase x for the We could do the same calculations for the other nine inbred
deviation of X from the mean. Thus, in Yao Ming’s case, lines, and then we would have a complete description of all
his height can be expressed as the mean value for men the phenotypes in each environment in terms of the extent
from Shanghai (170 cm) plus his specific genetic and envi- to which their deviation from the overall mean is due to
ronmental deviations (g + e = 59 cm). We can simplify the genetic and environmental factors.
equation above by subtracting X from both sides to obtain
x = g+e Genetic and environmental variances
We can use the simple model x = g + e to think further
where x represents the individual’s phenotypic deviation.
about the variance of quantitative traits. Recall that the
For Yao Ming’s height, x = g + e = 59 cm.
variance is a way to measure how much individuals devi-
How can we determine the values of g and e for Yao
ate from the population mean. Under this model, the trait
Ming? One way would be if we had clones of Yao Ming
variance can be partitioned into the genetic and the envi-
(clones are genetically identical individuals). Let’s imag-
ronmental variances:
ine that we cloned Yao Ming and distributed these clones
(as newborns) to a set of randomly chosen households in VX = Vg + Ve
TABLE 19-2 Simulated Data for Days to Pollen Shed for 10 Inbred Lines of Maize Grown in Two Experiments
Experiment I
Inbred lines A B C D E F G H I J Mean
Environment 1 62 64 66 66 68 68 70 70 72 74 68
Environment 2 64 66 68 68 70 70 72 72 74 76 70
Environment 3 66 68 70 70 72 72 74 74 76 78 72
Mean 64 66 68 68 70 70 72 72 74 76 70
Experiment II
Inbred lines A B C D E F G H I J Mean
Environment 4 58 60 62 62 64 64 66 66 68 70 64
Environment 5 64 66 68 68 70 70 72 72 74 76 70
Environment 6 70 72 74 74 76 76 78 78 80 82 76
Mean 64 66 68 68 70 70 72 72 74 76 70
This simple equation tells us that the trait or phenotypic gives inaccurate results. We will discuss this important
variation (VX ) is the sum of two components—the genetic assumption later in the chapter.
(Vg ) variance and the environmental (Ve ) variance. As noted We can use the data in Table 19-2 (experiment I)
in Box 19-1, there is an important assumption behind this to explore the equation for variances. First, let’s use all
equation; namely, that genotype and environment are not 30 phenotypic values for the 10 lines in the three environments
correlated—that is, they are independent. If the best gen- to calculate the variance. The result is VX = 14.67 days2 .
otypes are placed in the best environments and the worst Now, to estimate Vg, we calculate the variance of the means
genotypes in the worst environments, then this equation among the 10 inbred lines. The result is Vg = 12.0 days2 .
VX = E ( x 2 )
Finally, to estimate Ve, we calculate the variance of the is a critical concept in quantitative genetics, as we will see
means among the three environments. The result is throughout this chapter.
Ve = 2.67 days2. Thus, the phenotypic variance (14.67) is To visualize the degree of correlation between two vari-
equal to the genetic variance (12.0) plus the environmental ables, we can construct scatter plots, or scatter diagrams.
variance (2.67). The equation works for these data because Figure 19-4 shows the scatter plots that we would see under
genotype and environment are not correlated. several different strengths of correlation between two vari-
If we calculate the standard deviations for the data in ables. These plots use simulated data for the heights of imag-
Table 19-2 (experiment I), we observe that the phenotypic inary sets of identical adult male twins. The top panel of the
standard deviation (3.83) is not the sum of the genetic figure shows a perfect correlation, which is what we would
(3.46) and environmental (1.63) standard deviations. Vari- see if the height of one twin was exactly the same as that of
ances can be decomposed into difference sources. Standard the other twin for all sets of twins. The middle panel shows
deviations cannot be decomposed in this manner. In Sec- a strong but not perfect correlation. Here, when one twin is
tion 19.3, we will see how this property of the variance is short, the other also tends to be short, and when one is tall,
helpful for quantifying the extent to which trait variation is the other tends to be tall. The bottom panel shows the rela-
heritable versus environmental. tionship we would see if the height of one twin was uncor-
Finally, let’s look at what would happen to the variances if related with that of the other twin of the set. Here, the height
genotype and environment are correlated. To do this, imagine of one twin of each set is random with respect to the other
that we knew the genetic deviations (g) for nine Thorough- twin of the set. In the next section, we will see that the data
bred horses for the time it takes them to run the Kentucky for real twins would look something like the middle panel.
Derby. We also know the environmental deviations (e) that In statistics, there is a specific measure of correlation
their trainers contribute to the time it takes each horse to run called the correlation coefficient, which is symbolized by
this race. We will suppose that besides training, there are no a lowercase r. It is a measure of association between two
other sources of environmental variation. The population variables. The correlation coefficient is related to the cova-
mean for this set of Thoroughbreds is 123 seconds to run the riance, which was introduced in Box 19-1; however, it is
Derby. We assign the best horses to the best trainers and the scaled to vary between −1 and +1. If we symbolize one ran-
worst horses to the worst trainers. By doing this, we have cre- dom variable by X and the other by Y, then the correlation
ated a nonrandom relationship or correlation between horses coefficient between X and Y is
(genotypes) and trainers (environments).
COVX ,Y
Table 19-3 shows the data for this imaginary experi- rX ,Y =
ment. You’ll notice that VX (6.67) is not equal to the sum VXVY
of Vg (2.22) and Ve (1.33). Because genotype and environment The term VXVY is used to scale the covariance to vary
are correlated, we violated the assumption of the equation between −1 and +1. The expanded equation for the correla-
that states VX = Vg + Ve . The equation works only when tion coefficient is
genotype and environment are uncorrelated.
Σ(Xi − X)(Yi − Y )
rX ,Y =
Correlation between variables Σ(Xi − X)2 Σ(Yi − Y )2
Let’s look a little more closely at the concept of correlation, The equation is cumbersome, and in practice, the calcu-
the existence of a relationship between two variables. This lation of correlation coefficients is done with the aid of
TABLE 19-3 Simulated Data for Time in Seconds (X) that Horses Run the Kentucky Derby Decomposed into the
Genetic (g) and Environmental (e) Deviations from the Population Mean
Horse Population mean g Trainer e x X
Secretariat 123 −2 Lucien −2 −4 119
Decidedly 123 −2 Horatio −1 −3 120
Barbaro 123 −1 Mike −1 −2 121
Unbridled 123 −1 Carl 0 −1 122
Ferdinand 123 0 Charlie 0 0 123
Cavalcade 123 1 Bob 0 1 124
Meridian 123 1 Albert 1 2 125
Whiskery 123 2 Fred 1 3 126
Gallant Fox 123 2 Jim 2 4 127
Mean (sec) 123 0 0 0 123
Variance (sec ) 2 2.22 1.33 6.67 6.67
Scatter plots for different strengths of correlation This would best be done with a computer and spreadsheet
software. Use the genetic deviations (g) for the x-axis and
(a)
the environmental deviations (e) for the y-axis. Then cal-
Perfect correlation culate the correlation coefficient between g and e. The
198 r = 1.0 scatter diagram will be similar to the one in Figure 19-4
Height of twin 2 (cm)
1.0. H 2 is called “broad sense” because it encompasses sev- So, to estimate heritability Monozygotic twins are
eral different ways by which genes contribute to variation. in humans, we need to use genetically identical
For example, some of the variation will be due to the con- sets of identical twins who
tributions of individual genes. Additional genetic variation were separated shortly after
can be contributed by the way genes work together, the birth and reared apart by
interactions between genes known as epistasis. unrelated adoptive parents.
In Section 19.2, we showed how we can calculate the The equation for esti-
genetic and environmental variances when we have inbred mating H 2 in studies of
lines or clones. For the imaginary example of days to pollen identical twins who are
shed for maize inbred lines in Table 19-2 (experiment I), we reared apart is relatively
saw that Vg is 12.0 and VX is 14.67. Using these values, the simple. It makes use of the
FIGURE 19-5 A set of identical
heritability of the trait is 12.0/14.67 = 0.82, or 82 percent. statistical measure called twins. [Barbara Penoyar/Getty
This estimate of H 2 tells us that genes contribute most of the covariance, which was Images.]
the variation and environmental factors contribute a more introduced in Box 19-1. As
modest share of the variation. Thus, we might conclude explained in Box 19-2 , the
that days to pollen shed is a highly heritable trait in maize. covariance between identical twins who are reared apart is
Let’s look at the data for experiment II in Table 19-2. equal to the genetic variance (Vg ). Thus, we can estimate H 2
The genotypes are exactly the same as in experiment I; in humans by using this covariance as the numerator and
these are the genotypes of the inbred lines A through J. In the trait variance (VX ) as the denominator:
this case, however, the lines are reared in more extreme COVX ′,X ′′
environments. If we calculate the variance for the means H2 =
VX
of the inbred line in experiment II, Vg will be 12.0 days2
as in experiment I. Since the genotypes are the same in Here is how it’s done. For each set of twins, let’s designate
both experiments, the genetic variance is the same. If we the trait value for one twin as X ′ and the other as X ′′ . If
calculate the variance for the means of the different envi- we have n sets of twins, then the trait values for the n sets
ronments (Ve ) in experiment II, we will obtain 24.0 days2 , could be designated X1′ X1′′, X2′ X2′′, … Xn′ Xn′′.
which is much larger than the value for Ve in experiment I Suppose we had IQ measurements for five sets of twins
(2.67). Since the environments are more extreme, the envi- as follows:
ronmental variance is larger. Finally, if we calculate H 2 for
experiment II, we obtain Twin
Vg 12 X′ X ′′
H2 = = = 0.33 1 100 110
Vg + Ve 12 + 24
2 125 118
The estimate of H 2 for experiment II is on the small side—
3 97 90
closer to 0 than to 1. Thus, we might conclude that days to
pollen shed is not a highly heritable trait in maize. 4 92 104
The contrast between the estimates of the heritability for 5 86 89
the same set of maize inbred lines reared in different envi-
ronments highlights the point that heritability is the propor- Using these data and the formula for the covariance from
tion of the phenotypic variance (VX ) due to genetics. Since Box 19-1, we calculate that COVX ′,X ′′ is 119.2 points2 .
VX = Vg + Ve, as Ve increases, then Vg will represent a smaller Using the formula for trait variance, we would calculate
part of VX and H 2 will go down. Similarly, if the environ- that the value of VX is 154.3 points2. Thus, we obtain
mental variance is kept to a minimum, then Vg will represent 119.2 points2
a larger part of VX and H 2 will go up. H 2 is a moving target, H2 = = 0.77
154.3 points2
and results from one study may not apply to another.
The points2 in the numerator and denominator cancel out,
and we are left with a unitless measure that is the propor-
Measuring heritability in humans
tion of the total variance that is due to genetics.
using twin studies Box 19-2 provides some additional details about esti-
How can we measure heritability in humans? Although mating H 2 from twin data, including the derivation of
we don’t have inbred lines for humans, we do have geneti- the formula we just used. It also discusses the relation-
cally identical individuals—monozygotic or identical twins ship between the ratio COVX ′, X ′′ / VX and the correlation
(Figure 19-5). In most cases, identical twins are raised in the coefficient. Quantitative geneticists have developed sev-
same household and so experience a similar environment. eral means for estimating heritability using the correlation
When individuals with the same genotypes are reared in the among relatives. Identical twins share 100 percent of their
same environments, we have violated the assumption of our genes, while brothers, sisters, and dizygotic twins share
genetic model that genes and environment are independent. 50 percent of their genes. The strength of the correlation
between different types of relatives can be scaled for the populations and environments. We saw this phenomenon
proportion of their genes that they share and the results above in the case of the days to pollen shed for maize inbred
used to estimate the genetic and environmental contribu- lines. Second, the twin sets used in many studies were sep-
tions to trait variation. arated at birth and placed into adoptive homes. Adoption
Over the last 100 years, there have been extensive agencies do not assign babies randomly to the full range
genetic studies of twins and other sets of relatives. A great of households in a society; rather, they place babies in eco-
deal has been learned about heritable variation in humans nomically, socially, and emotionally stable households. As a
from these studies. Table 19-4 lists some results from twin result, Ve is smaller than in the general population, and the
studies. It may or may not be surprising to you, but there estimate of H 2 will be inflated. Accordingly, the published
is a genetic contribution to the variance for many differ- estimates likely lead us to underestimate the importance of
ent traits, including physique, physiology, personality attri- environment and overestimate the importance of genetics.
butes, psychiatric disorders, and even our social attitudes Third, for twins, prenatal effects could cause a positive cor-
and political beliefs. We readily observe that traits such as relation between genotype and environment. As we saw in
hair and eye color run in families, and we know these traits the earlier case of Thoroughbreds and jockeys, such a cor-
are the manifestation of genetically controlled biochemical, relation violates our model and will bias H 2 upward.
developmental processes. In this context, it is not so sur- Finally, heritability is not useful for interpreting differ-
prising that other aspects of who we are as people also have ences between groups. Table 19-4 shows that the heritabil-
a genetic influence. ity for height in humans can be very high: 0.88. However,
Twin studies and the estimates of heritability that they this high value for heritability does not tell us anything
provide can easily be over- or misinterpreted. Here are a about whether groups with different heights differ because
few important points to keep in mind. First, H 2 is a prop- of genetics or the environment. For example, men in the
erty of a particular population and environment. For this Netherlands today average 184 cm in height, while around
reason, estimates of H 2 can differ widely among different 1800, men in the Netherlands were about 168 cm tall on
B1 /B2 plants have 2 flowers, and B2 /B2 plants have 3 flow- Contribution
ers. In a case like this, when the heterozygote’s trait value to the mean
is midway between those of the two homozygous classes, Genotype Frequency Phenotype (frequency × value)
gene action is defined as additive gene action. In Figure B1 /B1 0.25 1 0.25
19-6b, the heterozygote has 3 flowers, the same as the B1 /B2 0.50 3 1.5
B2 /B2 homozygote. Here, the B2 allele is dominant to the B1
B2 /B2 0.25 3 0.75
allele. In this case, the gene action is defined as dominant
gene action. (We could also define this gene action as reces- Mean = 2.5
sive with the B1 allele being recessive to the B2 allele.) Gene
If the plant breeder selects a group of 3-flowered plants,
action need not be purely additive or dominant but can
2/3 will be B1 /B2 and 1/3B2 /B2 . When the breeder intermates
show partial dominance. For example, if B1 /B2 heterozy-
the selected plants, 0.44 (2/3 × 2/3) of the crosses would be
gotes had 2.5 flowers on average, then we would say that
between heterozygotes, and 1/4 of the offspring from these
the B2 allele shows partial dominance.
crosses would be B1 /B1 and thus 1-flowered. The remainder
of the offspring would be either B1 /B2 or B2 /B2 and thus
Gene action and the transmission of 3-flowered. The overall mean for the offspring would be
genetic variation 2.78, although the mean of their parents was 3.0. Hence,
Let’s work through a simple example to show how the when there is dominance, the phenotype is not fully her-
mode of gene action influences heritability. Suppose a plant itable. Selection as practiced by the plant breeder worked,
breeder wants to create an improved plant population with but not perfectly, because some of the differences among
more flowers per plant. Flower number is controlled by the individuals are due to dominance.
B locus, which has two alleles, B1 and B2, as diagrammed In conclusion, when there is dominance, we cannot
in Figure 19-6a. The frequencies of the B1 and B2 alleles are strictly predict the offspring’s phenotypes from the parents’
both 0.5, and the frequencies of the B1 /B1 , B1 /B2 , and B2 /B2 phenotypes. Some of the differences (variation) among the
genotypes are 0.25, 0.50, and 0.25, respectively. Plants with individuals in the parental generation are due to the dom-
the B1 /B1 genotype have 1 flower, B1 /B2 plants have 2 flow- inance interactions between alleles. Since parents transmit
ers, and B2 /B2 plants have 3 flowers. The mean number of their genes but not their genotypes to their offspring, these
flowers per plant in the population is 2.0. (Remember that dominance interactions are not transmitted to the offspring.
we can calculate the mean as the sum of the products of
frequency of each class times the value for that class.) The additive and dominance effects
Trait value Contribution As described earlier, traits controlled by genes with additive
(no of to the mean gene action will respond very differently to selection than
Genotype Frequency flowers) (frequency × value) those with dominance. Thus, geneticists need to quantify the
B1 /B1 0.25 1 0.25 degree of dominance and additivity. In this section, we will
B1 /B2 0.50 2 1.0 see how this is done. Let’s again consider the B locus that
controls the number of flowers on a plant (see Figure 19-6).
B2 /B2 0.25 3 0.75
The additive effect (A) provides a measure of the degree of
Mean = 2.0 change in the phenotype that occurs with the substitution
Since the heterozygote has a phenotype that is midway of one B2 allele for one B1 allele. The additive effect is calcu-
between the two homozygous classes, gene action is additive. lated as the difference between the two homozygous classes
There are no environmental effects, and the genotype alone divided by 2. For example, as shown in Figure 19-6a, if the
determines the number of flowers, so H 2 is 1.0. If the plant trait value of the B1 /B1 genotype is 1 and the trait value of
breeder selects 3-flowered plants (B2 /B2 ), intermates them, and the B2 /B2 genotype is 3, then
grows the offspring, then all the offspring will be B2 B2, and the XB2 B2 − XB1B1 3 −1
mean number of flowers per plant among the offspring will be A= = =1
2 2
3.0. When gene action is completely additive and there are no
environmental effects, the phenotype is fully heritable. Selec- The dominance effect (D) is the deviation of the hetero-
tion as practiced by the plant breeder works perfectly. zygote (B1 /B2 ) from the midpoint of the two homozygous
Now let’s consider the case diagrammed in Figure 19-6b, classes. As shown in Figure 19-6b, if the trait value of the
in which the B2 allele is dominant to the B1. In this case, the B1 /B1 genotype is 1, of the B1 /B2 genotype, 3, and of the
B1B2 heterozygote is 3-flowered. The frequency of the B1 and B2 /B2 genotype, 3, then
B2 alleles are both 0.5, and the frequencies of the B1 /B1 , B1 /B2 XB B + XB1B1
, and B2 /B2 genotypes are 0.25, 0.50, and 0.25, respectively. D = XB1B2 − 2 2 = 3 − 2 = 1
2
Again, there is no environmental contribution to the differ-
ences among individuals, so H 2 is 1.0. The mean number of If you calculate D for the situation depicted in Figure 19-6a,
flowers per plant in the starting population is 2.5. you’ll find D = 0; that is, no dominance.
The ratio of D/A provides a measure of the degree of The presence or absence of pelvic spines or eyes are
dominance. For Figure 19-6a, D/A = 0.0, indicating pure two traits studied in the fish below
additivity or no dominance. For Figure 19-6b, D/A = 1.0,
(a)
indicating complete dominance. A D/A ratio of −1 would
indicate a complete recessive. (The distinction between
dominance and recessivity depends on how the phenotypes
are coded and is in this sense arbitrary.) Values that are
greater than 0 and less than 1 represent partial dominance,
and values that are less than 0 and greater than −1 repre-
Pelvic spine
sent partial recessivity.
Here is an example of calculating additive and domi-
nance effects at a single locus. Three-spined sticklebacks
(Gasterosteus aculeatus) have marine populations with
long pelvic spines and populations that live near the bot-
toms of freshwater lakes with highly reduced pelvic spines
( Figure 19-7a ). The spines are thought to play a role in
defense against predation. The bottom-dwelling freshwater
populations are derived from the ancestral marine popula-
tions. A change in predation between the marine and fresh-
water environments may explain the loss of spines in the
freshwater environments (see Chapter 20).
Pitx1 is one of several genes that contribute to pelvic-
spine length in sticklebacks. This gene encodes a transcrip-
tion factor that regulates the development of the pelvis in
vertebrates, including the growth of pelvic spines in stick-
lebacks. Michael Shapiro and his colleagues at Stanford
University measured the pelvic-spine length in an F2
population that segregated for the marine or long (l) allele
and freshwater or short (s) allele of Pitx1. They recorded
FIGURE 19-7 (a) Three-spined stickleback (Gasterosteus aculeatus).
the following mean values (in units of proportion of body (b) Blind cave fish (Astyanax mexicanus) (top) and its sighted, surface
length) for pelvic-spine length for the three genotypic relative (bottom). [(a) Masato Yoshizawa and William Jeffery, University of
classes: Maryland; (b) Masato Yoshizawa and William Jeffery, University of Maryland.]
phenotypes from parental phenotypes when there is dom- the frequencies of the B1 and B2 allele in the population. You
inance, although we can do so in cases of pure additivity. will notice that a + d sum to g. The additive (a) and domi-
When predicting the phenotypes of offspring, we need to sep- nance (d) deviations are dependent on the allele frequencies
arate the additive and dominance contributions. To do this, because the phenotype of an offspring receiving a B1 allele
we need to modify the simple model introduced in Section from one parent will depend on whether that allele combines
19.2, x = g + e. with a B1 or B2 allele from the other parent, and that outcome
Let’s begin by looking more closely at the situation depends on the frequencies of the alleles in the population.
depicted in Figure 19-6b. Individuals with the B1 /B2 and The additive deviation (a) has an important meaning in
B2 /B2 genotypes have the same phenotype, 3 flowers. If we plant and animal breeding. It is the breeding value, or the
subtract the population mean (2.5) from their trait value part of an individual’s deviation from the population mean
(3), we see that they have the same genotypic deviation (g): that is due to additive effects. This is the part that is trans-
mitted to its progeny. Thus, if we wanted to increase the
gB1B2 = gB2 B2 = 0.5
number of flowers per plant in the population, the B2 /B2
Now let’s calculate the mean phenotypes of their offspring. individuals have the highest breeding value. Breeding val-
If we self-pollinate a B1 /B2 individual, the offspring will be ues can also be calculated for the genome overall for an
1 B /B , 1 B /B , and 1 B /B , and the mean trait value of
4 1 1 2 1 2 4 2 2 individual. Animal breeders estimate the genomic breeding
these offspring would be 2.75. However, if we self-pollinate a values of individual animals, and these estimates can deter-
B2 /B2 individual, the offspring will all be B2 /B2 , and the mean mine the economic value of the animal.
trait value of these offspring would be 3.0. Even though the We have partitioned the genetic deviation (g) into the addi-
B1 /B2 and B2 /B2 individuals have the same trait value and tive (a) and dominance (d) deviations. Using algebra similar to
the same value for their genotypic deviation (g), they do not that described in Box 19-1, we can also partition the genetic
produce the equivalent offspring because the underlying basis variance into the additive and dominance variances as follows:
of their phenotypes is different. The phenotype of the B1 /B2
Vg = Va + Vd
individual depends on the dominance effect (D), while that of
the B2 /B2 individual does not involve dominance. where Va is the additive genetic variance and Vd is the dom-
We can expand the simple model (x = g + e) to incorpo- inance variance. Va is the variance of the additive deviations
rate the additive and dominance contributions. The geno- or the variance of the breeding values. It is the part of the
typic deviation (g) is the sum of two components: a, the genetic variation that is transmitted from parents to their
additive deviation, which is transmitted to offspring; and d, offspring. Vd is the variance of the dominance deviations.
the dominance deviation, which is not transmitted to off- Finally, we can substitute these terms in the equation for
spring. We can rewrite the simple model and separate out the phenotypic variance presented earlier in the chapter:
these two components as follows:
VX = Vg + Ve
x = g+e
VX = Va + Vd + Ve
x = a+d +e
where Ve is the environmental variance. This equation
The additive deviation is transmitted from parent to off- assumes that the additive and dominance components are
spring in a predictable way. The dominance deviation is not not correlated with the environmental effects. This assump-
transmitted from parent to offspring since new genotypes tion will be true in experiments in which individuals are
and thus new interactions between alleles are created each randomly assigned to environments.
generation. Thus far, we have described models with genetic, envi-
Let’s look at how the genetic deviation is decomposed ronmental, additive, and dominance deviations and vari-
into the additive and dominance deviations for the case ances. In quantitative genetics, the models can get even
shown in Figure 19-6b. more complex. In particular, the models can be expanded
to include interaction between factors. If one factor alters
B1B1 B1B2 B2 B2
the effect of another factor, then there is an interaction.
Trait value 1 3 3 Box 19-3 briefly reviews how interactions are factored into
Genetic deviation (g) −1.5 0.5 0.5 quantitative genetic models.
Additive deviation (a) −1 0 1
Dominance deviation (d) −0.5 0.5 −0.5
KEY CONCEPT The genetic deviation (g) of an individual from
The genotypic deviations (g) are simply calculated by sub- the population mean is composed of the additive deviation (a)
tracting the population mean (2.5) from the trait value for and dominance deviation (d). The additive deviation is known
each genotype. Each genotypic deviation is then decomposed as the breeding value, and it represents the component of an
individual’s phenotype that is transmitted to its offspring. The
into the additive (a) and dominance (d) deviations using for-
genetic variation for a trait in a population (Vg ) can be decom-
mulas that are beyond the scope of this book. These formulas
posed into the additive (Va ) and the dominance (Vd ) variances.
include the additive (A) and dominance (D) effects as well as
3 3 No interaction Interaction
IL1
2 2 B1 /B1 B1 /B2 B2 /B2 B1 /B1 B1 /B2 B2 /B2
1 IL2 1 A1 /A1 0 1 2 A1 /A1 0 1 2
0 0 A1 /A2 1 2 3 A1 /A2 0 1 3
E1 E2 E1 E2
A2 /A2 2 3 4 A2 /A2 0 1 4
With no interaction, the difference between the two inbreds
is 1.0 in both environments, and so the difference between Now look at the right side of the table. Starting with the
the lines averaged over the two environments is 1.0. A1 /A1; B1 /B1 genotype, substituting an A2 allele for an A1
allele has an effect on the trait value only when the gen-
Environment 1: IL1− IL2 = 2 − 1 = 1.0 otype at the B locus is B2 /B2. The effects of alleles at the
Environment 2: IL1− IL2 = 3 − 2 = 1.0 A locus are dependent of those at the B locus. There is an
interaction or epistasis between the genes.
The difference in the overall mean shows that the lines are
The genetic model can be expanded to include an epi-
genetically different. The mean over both environments is
static or interaction term (i):
2.5 for IL1 and 1.5 for IL2.
The graph on the right shows a case of an interaction x = a+d + i +e
between genotype and environment. IL1 does well in Envi-
and
ronment 1 but poorly in Environment 2. The opposite is true
for IL2. The difference in the trait value between the two VX = Va + Vd + Vi + Ve
lines is +1.0 in Environment 1 but −1.0 in Environment 2.
where Vi is the interaction or epistatic variance.
Environment 1: IL1− IL2 = 2 − 1 = +1.0 If the interaction term is not included in the model,
Environment 2: IL1− IL2 = 1 − 2 = −1.0 then there is an implicit assumption that the genes work
independently; that is, there is no epistasis. The interaction
The difference between the lines averaged over the two
variance (Vi ), like the dominance variance, is not transmit-
environments is 0.0, so we might incorrectly conclude that
ted from parents to their offspring, since new genotypes
these inbreds are genetically equivalent if we looked just at
and thus new epistatic relationships are formed with each
the overall mean.
generation.
The simple model can be expanded to include a
genotype–environment interaction term ( g × e ) :
x = g + e + g×e
Narrow-sense heritability to offspring. The results for your class could deviate from
this value for several reasons. First, if your class is small,
We can now define narrow-sense heritability, which is sym-
sampling error can affect the accuracy of your estimate
bolized by a lowercase h squared (h2 ), as the ratio of the
of h2 . Second, you will not be conducting a randomized
additive variance to the total phenotypic variance:
experiment. If parents re-create in their households the
Va Va growth-promoting (or growth-limiting) environments that
h2 = =
VX Va + Vd + Ve they experienced as children, then there will be a correla-
tion between the environments of the parents and their
This form of heritability measures the extent to which
offspring. This correlation of environments violates an
variation among individuals in a population is predictably
assumption of the analysis. Third, the population of stu-
transmitted to their offspring. Narrow-sense heritability
dents in your class may not be representative of the popula-
is the form of heritability of interest to plant and animal
tion in which the 0.8 value was obtained.
breeders because it provides a measure of how well a trait
Figure 19-8 is a scatter plot with the height data for male
will respond to selective breeding.
and female students and their parents. There is a clear correla-
To estimate h2, we need to measure Va, but how can this
tion between the heights of the students and their same-sex
be accomplished? Using algebra and logic similar to what
parent. These data give estimates of narrow-sense heritabil-
we used to show that Vg can be estimated using the covari-
ity of 0.86 for mother–daughter and 0.82 for father–son.
ance between monozygotic twins reared separately (see
Box 19-2), it can be shown that the covariance between a par-
ent and its offspring is equal to one-half the additive variance: The heights of individuals and their same-sex
parent are correlated
1
COVP ,O = Va
2 Female students
The parent–offspring covariance is one-half of Va because 72 h2 = 0.86
Height of students (inches)
Va 2COVP ,O 68
h2 = =
VX VX
To estimate Va using the covariance between parents and
offspring requires controlling environmental factors in 64
experiments. This can be a challenge because parents and
offspring are necessarily reared at different times. Va can
also be estimated using the covariance between half-sibs, in
60
which case all individuals in the experiment can be reared 60 64 68 72
at the same time in the same environment. Half-sibs share Height of parents (inches)
one-fourth of their genes, so Va equals 4 × the covariance
between half-sibs.
Male students
If you compare the equation for h2 to the one for H 2 78
(see Box 19-2), you will see that both involve the ratio of a h2 = 0.82
Height of students (inches)
The results are close to the value of h2 equals 0.8 obtained The additive part is the heritable part that is transmitted to
from studies in which the children were separated at birth the offspring. Let’s look at a set of parents with phenotypic
from their parents and reared in adoptive households. deviations x ′ for the mother and x ′′ for the father. The par-
Here are a few more points about narrow-sense herita- ents’ dominance deviations (d ′ and d ′′ ) are not transmitted
bility. First, when h2 = 1.0 (Va = VX ), the expected value for to their offspring since new genotypes and new dominance
an offspring’s phenotype will equal the mid-parent value. All interactions are created with each generation. Similarly, the
the variation in the population is additive and heritable in the parents do not transmit their environmental deviations (e ′
narrow sense. Second, when h2 = 0.0 (Va = 0), the expected and e ′′) to their offspring.
value of any offspring’s phenotype will be the population Mom Dad
mean. All the variation in the population is due either to dom-
x = a′ + d ′ + e ′
′ x = a′′ + d ′′ + e ′′
′′
inance or to environmental factors, and thus it is not trans-
missible to offspring. Finally, as with broad-sense heritability
(H 2 ) , narrow-sense heritability is the property of the specific
environment and population in which it was measured. An Offspring
estimate from one population and environment may not be a′ + a′′
meaningful for another population or environment. xo = = ap
2
Narrow-sense heritability is an important concept
both in plant and animal breeding and in evolution. For a Thus, the only factors that parents transmit to their off-
breeder, h2 indicates which traits can be improved by arti- spring are their additive deviations (a′ and a′′). Accordingly,
ficial selection. For an evolutionary biologist, h2 is critical we can estimate the offspring’s phenotypic deviation (xo ) as
to understanding how populations will change in response the mean of the additive deviations of its parents ap .
to natural selection imposed by a changing environment. So, to predict the offspring’s phenotype, we need to
Table 19-5 lists estimates of narrow-sense heritability for know its parents’ additive deviations. We cannot directly
some traits and organisms. observe the parents’ additive deviations, but we can esti-
mate them. The additive deviation of an individual is the
Predicting offspring phenotypes heritable part of its phenotypic deviation; that is,
In order to efficiently improve crops and livestock for traits â = h2 x
of agronomic importance, the breeder must be able to pre-
where â signifies an estimate of the additive deviation or
dict an offspring’s phenotype from its parents’ phenotypes.
breeding value. Thus, we can estimate the mean of the par-
Such predictions are made using the breeder’s knowledge of
ents’ additive deviations as the product of h2 times the mean
narrow-sense heritability. An individual’s phenotypic devia-
of their phenotypic deviation and this product will be an esti-
tion (x) from the population mean is the sum of the addi-
mate of the phenotypic deviation of the offspring (xˆ o ):
tive, dominance, and environmental deviations:
x ′ + x ′′
x = a+d +e xˆ o = h2
2
TABLE 19-5 Narrow-Sense Heritability for Some or
Traits in Several Different Species
Trait h2 (%) xˆ o = h2 xp
Agronomic species The offspring will have its own dominance and environ-
Body weight in cattle 65 mental deviations. However, these cannot be predicted.
Since they are deviations, they will be zero on average over
Milk yield in cattle 35
a large number of offspring.
Back-fat thickness in pig 70
Here is an example. Icelandic sheep are prized for the
Litter size in pig 5 quality of their fleece. The average adult sheep in a par-
Body weight in chicken 55 ticular population produces 6 lb of fleece per year. A sire
Egg weight in chicken 50 that produces 6.5 lb per year is mated with a dam that
Natural species produces 7.0 lb per year. The narrow-sense heritability of
fleece production in this population is 0.4. What is the pre-
Bill length in Darwin’s finch 65
dicted fleece production for offspring of this mating? First,
Flight duration in milkweed bug 20 calculate the phenotypic deviations for the parents by sub-
Plant height in jewelweed 8 tracting the population mean from their phenotypic values:
Fecundity in red deer 46
Life span in collared flycatchers 15 Sire 6.5 − 6.0 = 0.5
Source: D. F. Falconer and T. F. C. Mackay, Introduction to Quantitative Dam 7.0 − 6.0 = 1.0
Genetics, Longman, 1996; J. C. Conner and D. L. Hartl, A Primer in
Ecological Genetics, Sinauer, 2004. Parent mean (xp ) (0.5 + 1.0)/2 = 0.75
Now multiply h2 by xp to determine xˆ 0 , the estimated phe- Selection can change the population mean
notypic deviation of the offspring:
(a) Foundation population
0.4 × 0.75 = 0.3
1.25
Finally, add the population mean (6.0) to the predicted
phenotypic deviation of the offspring (0.3) and obtain the
result that the predicted phenotype of the offspring is 6.3 lb The mean of the
of fleece per year. selected plants
1.63
It may seem surprising that the offspring are predicted
to produce less fleece than either parent. However, this
outcome is expected for a trait with a modest heritability
of 0.4. Most (60 percent) of the superior performance of
the parents is due to dominance and environmental factors
that are not transmitted to the offspring. If the heritability Selected plants
were 1.0, then the predicted value for the offspring would (b) Offspring of
be midway between those of the parents. If the heritability selected individuals
1.44
were 0.0, then the predicted value for the offspring would
be at the population mean since all the variation would be
due to nonheritable factors.
The underlying logic of this calculation is that the response selected and bred to form the next generation. Over the 100
represents the heritable or additive part of the selection generations, the average flight speed of the flies in the pop-
differential. ulation increased from 2 to 170 cm/sec, and neither the flies
Over the last century, quantitative geneticists have con- nor the gains made by selection showed any signs of slow-
ducted a large number of selection experiments like this. ing down after 100 generations. In the second experiment,
Typically, these experiments are performed over many gen- mice were selected over 10 generations for the amount of
erations and are referred to as long-term selection studies. “wheel running” they did per day (Figure 19-10b). There
Each generation, the best individuals are selected to produce was a 75 percent increase over just 10 generations. These
the subsequent generation. Such studies have been performed studies and many more like them demonstrate the tremen-
in economically important species such as crop plants and dous power of artificial selection.
livestock and in many model organisms such as Drosophila,
mice, and nematodes. This work has shown that virtually any KEY CONCEPT Narrow-sense heritability ( h2 ) is the propor-
species will respond to selection for virtually any trait. Popu- tion of the phenotypic variance that is attributable to additive
lations contain deep pools of additive genetic variation. effects. This form of heritability measures the extent to which
Here are two examples of long-term selection experi- variation among individuals in a population is predictably
ments. In the first experiment, fruit flies were selected for transmitted to their offspring. Narrow-sense heritability is an
important quantity in plant and animal breeding since it pro-
increased flight speed over a period of 100 generations
vides a measure of how well a trait will respond to selective
(Figure 19-10a). Each generation, the speediest flies were
breeding.
Frequency distributions show the contributions of Table 19-6 shows part of such a data set for just 20 plants and
alleles at a QTL to a complex trait 5 marker loci that are linked on a single chromosome. For
each BC1 plant, we have the weight of its fruit and the gen-
otypes at the marker loci. Notice that trait values for the
0.06
BC1 plants are intermediate between the two parents as
B/b
expected but closer to the Beefmaster value because this is
Frequency
TABLE 19-6 Simulated Fruit Weight and Marker-Locus Data for a Backcross Population between Two Tomato
Inbred Lines—Beefmaster and Sungold
Markers
Plant Fruit wt. (g) M1 M2 M3 M4 M5
Beefmaster 230 B/B B/B B/B B/B B/B
Sungold 10 S/S S/S S/S S/S S/S
BC1-001 183 B/B B/B B/B B/S B/S
BC1-002 176 B/S B/S B/B B/B B/B
BC1-003 170 B/B B/S B/S B/S B/S
BC1-004 185 B/B B/B B/B B/S B/S
BC1-005 182 B/B B/B B/B B/B B/B
BC1-006 170 B/S B/S B/S B/S B/B
BC1-007 170 B/B B/S B/S B/S B/S
BC1-008 174 B/S B/S B/S B/S B/S
BC1-009 171 B/S B/S B/S B/B B/B
BC1-010 180 B/S B/S B/B B/B B/B
BC1-011 185 B/S B/B B/B B/S B/S
BC1-012 169 B/S B/S B/S B/S B/S
BC1-013 165 B/B B/B B/S B/S B/S
BC1-014 181 B/S B/S B/B B/B B/S
BC1-015 169 B/S B/S B/S B/B B/B
BC1-016 182 B/B B/B B/B B/S B/S
BC1-017 179 B/S B/S B/B B/B B/B
BC1-018 182 B/S B/B B/B B/B B/B
BC1-019 168 B/S B/S B/S B/B B/B
BC1-020 173 B/B B/B B/B B/B B/B
Mean of B/B — 176.3 179.6 180.7 176.1 175.0
Mean of B/S — 175.3 173.1 169.6 175.3 176.4
Overall mean 175.7
line have smaller fruits than those that inherited the B allele different do they need to be before we declare that a QTL is
from the Beefmaster line. located near a marker? The statistical details for answering
Figure 19-13 is a graphical representation of QTL- this question are beyond the scope of this text. However,
mapping data for many plants along one chromosome let’s review the basic logic behind the statistics. The statis-
using values from Table 19-6. The phenotypic data for the tical analysis involves calculating the probability of observ-
B/B and B/S genotypic classes are represented as frequency ing the data (the specific fruit weights and marker-locus
distributions so we can see the distributions of the trait val- genotypes for all the plants) given that there is a QTL near
ues. At marker M1, the distributions are fully overlapping the marker locus and the probability of observing the data
and the means for the B/B and B/S distributions are very given that there is not a QTL near the marker locus. The
close. It appears that the B/B and B/S classes have the same ratio of these two probabilities is called the “odds”:
underlying distribution. At marker M3, the distributions
Prob(data|QTL)
are only partially overlapping and the means for the B/B odds =
and B/S distributions are quite different. The B/B and B/S Prob(data|no QTL)
classes around M3 have different underlying distributions The vertical line | means “given,” and the term
similar to the situation in Figure 19-11. Again, we have evi- Prob(data|QTL) reads “the probability of observing the
dence for a QTL near M3. data given that there is a QTL.” If the probability of the
As shown in Figure 19-13, the trait means for the B/B data when there is a QTL is 0.1 and the probability of
and B/S groups at some markers are nearly the same. At the data when there is no QTL is 0.001, then the odds are
other markers, these means are rather different. How 0.1/0.001 = 100. That is, the odds are 100 to 1 in favor of
Distinct distributions for genotypic classes at a marker peaks of various heights as well as stretches that are rela-
locus signal the location of a QTL near the marker tively flat. The peaks represent putative QTL, but how high
does a peak need to be before we declare that it represents
B/B a QTL? As discussed in Chapters 4 and 18, we can set a
M1 statistical threshold for rejecting the “null hypothesis.” In
this case, the null hypothesis is that “there is not a QTL
B/S at a specific position along the chromosome.” Where the
Lod score exceeds the threshold value, then we reject the
null hypothesis in favor of the alternative hypothesis that
M2 a QTL is located at that position. In Figure 19-14, the Lod
score exceeds the threshold value (red line) near marker
locus M3. We conclude that a QTL is located near M3.
In addition to backcross populations, QTL mapping can
be done with F2 populations and other breeding designs. An
M3
advantage of using an F2 population is that one gets esti-
mates of the mean trait values for all three QTL genotypes:
homozygous parent-1, homozygous parent-2, and heterozy-
gous. With these data, one can get estimates of the additive
(A) and dominance (D) effects of the QTL as discussed ear-
M4
lier in this chapter. Thus, QTL mapping enables us to learn
about gene action, whether dominant or additive, for each
QTL.
Here is an example. Suppose we studied an F2 popu-
lation from a cross of Beefmaster and Sungold tomatoes
M5
and we identified two QTL for fruit weight. The mean fruit
weights for the different genotypic classes at the QTL might
look something like this:
160 190 Fruit weights Effects
Fruit weight (g)
B/B B/S S/S A D
FIGURE 19-13 A tomato chromosomal segment with marker loci
M1 through M5. At each marker locus, the frequency distributions for QTL 1 180 170 160 10 0
fruit weight from a BC1 population of a Beefmaster × Sungold cross QTL 2 200 185 110 45 30
are shown. The red distributions are for the homozygous Beefmaster
(B/B) genotypic class at the marker; the gray distributions are for the We can use these fruit weight values for the QTL to calcu-
heterozygous (B/S) genotypic class. Yellow lines represent the mean of
each distribution.
late the additive and dominance effects. QTL 1 is purely
additive (D = 0), but QTL 2 has a large dominance effect.
Also, notice that the additive effect of QTL 2 is more than
there being a QTL. Researchers report the log10 of the odds, 4 times that of QTL 1 (45 versus 10). Some QTL have large
or the Lod score. So, if the odds ratio is 100, then the log10 effects, and others have rather small effects.
of 100, or Lod score, is 2.0.
In addition to testing for QTL at the marker loci where
the genotypes are known, Lod scores can be
calculated for points between the markers.
Lod scores provide statistical evidence for QTL
This can be done by using the genotypes of
the flanking markers to infer the genotypes at 10
points between the markers. For example, in
Lod score
What can be learned from QTL mapping? With the Similar studies have been done on dozens of other disease
most powerful QTL-mapping designs, geneticists can esti- conditions.
mate (1) the number of QTL (genes) affecting a trait, (2)
the genomic locations of these genes, (3) the size of the
effects of each QTL, (4) the mode of gene action for the From QTL to gene
QTL (dominant versus additive), and (5) whether one QTL QTL mapping does not typically reveal the identity of the
affects the action of another QTL (epistatic interaction). In gene(s) at the QTL. At its best, the resolution of QTL map-
other words, one can get a rather complete description of ping is on the order of 1 to 10 cM, the size of a region that
the genetic architecture for the trait. can contain 100 or more genes. To go from QTL to a single
Much has been learned about genetic architecture from gene requires additional experiments to fine-map a QTL. To
QTL-mapping studies in diverse organisms. Here are two do this, the researcher creates a set of genetic homozygous
examples. First, flowering time in maize is a classic quan- stocks (also called lines), each with a crossover near the QTL.
titative or continuous trait. Flowering time is a trait of These stocks or lines differ from one another near the QTL,
critical importance in maize breeding since the plants must but they are identical to one another (isogenic) throughout
flower and mature before the end of the growing season. the rest of their genomes. Lines that are identical throughout
Maize from Canada is adapted to flower within 45 days their genomes except for a small region of interest are called
after planting, while maize from Mexico can require 120 congenic lines or nearly isogenic lines. The isolation of QTL
days or longer. QTL mapping has shown that the genetic in an isogenic background is critical because only the single
architecture for flowering time in maize involves more QTL region differs between the congenic lines. Thus, the use
than 50 genes. Results from one experiment are shown in of congenic lines eliminates the complications caused by hav-
Figure 19-15a; these results show evidence for 15 QTL. QTL ing multiple QTL segregate at the same time.
for maize flowering time generally have a small effect, such Using the tomato fruit weight example provided earlier,
that substituting one allele for another at a QTL alters the chromosome region for a set of such congenic lines is
flowering time by only one day or less. Thus, the difference shown in Figure 19-16. The genes (flc, arf4, . . .) are shown
in flowering time between tropical and temperate maize at the top, and the location for each crossover is indicated
involves many QTL. by the switch in color from red (Beefmaster genotype) to
Second, mice have been used to map QTL for many yellow (Sungold genotype). The mean fruit weight for the
disease-susceptible traits. What one learns about congenic lines carrying these recombinant chromosomes is
disease-susceptibility genes in mice is often true in humans indicated on the right. By inspection of Figure 19-16, you
as well. Figure 19-15b shows the results of a genomic scan will notice that all lines with the Beefmaster allele of kin1
in mice for QTL for bone mineral density (BMD), the trait (a kinase gene) have fruit of ~ 180 g, while those with the
underlying osteoporosis. This scan identified two QTL, one Sungold allele of kin1 have fruit of about ~ 170 g . None
on chromosome 9 and one on chromosome 12. From stud- of the other genes are associated with fruit weight in this
ies such as this, researchers have identified over 80 QTL in way. If confirmed by appropriate statistical tests, this result
mice that may contribute to susceptibility to osteoporosis. allows us to identify kin1 as the gene underlying this QTL.
10
0
0 100 200
centimorgans
(b) QTL for bone mineral density in mice
Lod score
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 x
Chromosome
FIGURE 19-15 Plot of Lod scores from genomic scans for QTL. (a) Results from a scan for flowering time QTL in maize. (b) Results from a scan
for bone-mineral-density QTL in mice. [(a) Data from E. S. Buckler et al., Science 325, 2009, 714–718; (b) Data from N. Ishimori et al., J. Bone Min. Res. 23,
2008, 1529–1537.]
5 171.2
6 180.7
7 181.8
8 169.3
9 170.7
10 171.4
Table 19-7 lists a small sample of the hundreds of genes KEY CONCEPT Quantitative trait locus (QTL) mapping is a
or QTL affecting quantitative variation from different spe- procedure for identifying the genomic locations of the genes
cies that have been identified. The list includes the gene for (QTL) that control variation for complex traits. QTL mapping
maize flowering time, Vgt, that underlies one of the Lod evaluates the progeny of controlled crosses for their geno-
peaks in Figure 19-15a. One notable aspect of this list is the types at molecular markers and for their trait values. If the
diversity of gene functions. There does not appear to be a different genotypes at a marker locus have different mean
values for the trait, then there is evidence for a QTL near the
rule that only particular types of genes can be a QTL. Most,
marker.
if not all, genes in the genomes of organisms are likely to
contribute to quantitative variation in populations.
TABLE 19-7 Some Genes Contributing to Quantitative Variation That Were First Identified Using QTL Mapping
Organism Trait Gene Gene function
Yeast High-temperature growth RHO2 GTPase
Arabidopsis Flowering time CRY2 Cryptochrome
Maize Branching Tb1 Transcription factor
Maize Flowering time Vgt Transcription factor
Rice Photoperiod sensitivity Hd1 Transcription factor
Rice Photoperiod sensitivity CK2α Casein kinase subunit
Tomato Fruit-sugar content Brix9-2-5 Invertase
Tomato Fruit weight Fw2.2 Cell-cell signaling
Drosophila Bristle number Scabrous Secreted glycoprotein
Cattle Milk yield DGAT1 Diacylglycerol acyltransferase
Mice Colon cancer Mom1 Modifier of a tumor-suppressor gene
Mice Type 1 diabetes I-Aβ Histocompatibility antigen
Humans Asthma ADAM33 Metalloproteinase-domain-containing protein
Humans Alzheimer’s disease ApoE Apolipoprotein
Humans Type 1 diabetes HLA-DQA MHC class II surface glycoprotein
Source: A. M. Glazier et al., Science 298, 2002, 2345–2349.
Recombination
hot spot
Haplotype block Haplotype block
Chromosome SNP1 SNP2 SNP3 SNP4 SNP5 SNP6 SNP7 SNP8 SNP9 SNP10
1 A C A G C G T A C G S
2 A C A G C G T A C G S
3 A C A G C G T A C G S
4 A C A G C A C G G A S
5 A C A G C A C G G A S
6 A C A G C A C G G A D
7 G C A G T G T A C G S
8 G C A G T G T A C G S
9 G C A G T G T A C G S
10 G C A G T A C G G A D
11 G C A G T A C G G A D
12 G C A G T A C G G A D
13 A G T G T G T A C G D
14 A G T G T G T A C G S
15 A G T G T G T A C G S
16 A G T G T A C G G A D
17 A G T G T A C G G A D
18 A G T G T A C G G A D
Strong disequilibrium
No disequilibrium
FIGURE 19-17 (top) Diagram of the distribution of SNPs and haplotypes for a chromosomal segment (golden bar) from 18 individuals. Haplotypes
often occur in blocks (regions of lower recombination) separated from one another by recombination hotspots (different colors indicate haplotype
blocks). (The column of S’s and D’s at the right are for Problem 19-4.) SNP8 (bold) controls a difference in trait values. (bottom) You can tell whether
two SNPs show disequilibrium by noting the color of the square where the rows for the markers intersect. Within a haplotype block, SNPs show
strong disequilibrium (red). SNPs in different haplotype blocks show weak or no disequilibrium (white). [Data from David Altshuler et al., Science 322,
2008, 881–888.]
Association mapping finds a gene for body size in dogs chromosome 15, there is a cluster of SNPs above the
threshold line, indicating that the null hypothesis of no
(a)
IGF1 association can be rejected for these SNPs in favor of
10-4 the alternative hypothesis that a gene affecting body size
Significance threshold
in dogs is located at this position. The strong peak on
P value
15
10
APOE
5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122 x
Crohn’s disease
CARD15
15 IL23R ATG16L1 IRGM
5
(P value)
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122 x
Rheumatoid arthritis
–log10
HLA-DRB1
15
PTPN22
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122 x
Type 1 diabetes
HLA-DRB1
15 PTPN22
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 2122 x
FIGURE 19-19 Results from a genome-wide association study of by this analysis are shown in red. [Reprinted by permission from Macmillan
common diseases in humans. The 23 human chromosomes are arrayed Publishers Ltd. from The Wellcome Trust Case Control Consortium, “Genome-
from left to right. The y-axis shows the P value for the statistical test wide association study of 14,000 cases of seven common diseases and 3,000
of an association between the disease and each SNP. Significant test shared controls,” Nature 447, 2007, 661–678, Figure 4. Permission conveyed
results are shown as green dots. The names of some genes identified through Copyright Clearance Center, Inc.]
These data are ushering in a new era of precision medicine, governed by many genes. However, a perplexing result
in which an individual can have his or her genome scanned was that the 180 genes accounted for only 10 percent of
to determine their genotype at genes known to increase dis- the genetic variance in height. This falls far short of the
ease risk. Although this science is relatively young, it is pos- roughly 80 percent value for broad-sense heritability for
sible to identify individuals who have a 10-fold higher risk height. The difference between 10 percent and 80 percent
Introduction
for certain to Geneticthan
diseases Analysis,
other11emembers of the population. has been dubbed the missing heritability. For disease risk,
Figureinformation
Such 19.19 #1919 can be used to initiate preventative mea- there is also much missing heritability. For example, GWA
08/27/14
sures andMedia
Dragonfly changes in lifestyle (environment) that contribute
Group studies have succeeded in explaining only 10 percent of the
to disease risk. Some companies offer direct-to-consumer genetic variation for Crohn’s disease and only 5 percent of
genotyping service including information on whether you the genetic variation for type 2 diabetes.
carry risk alleles for diseases like breast cancer, Alzheimer’s It has come as a surprise to many geneticists that
disease, macular degeneration, and more. GWA studies with hundreds of thousands of SNPs blan-
Since height in humans is a classic quantitative trait, keting the genome and samples of over 10,000 individu-
quantitative geneticists had great interest in performing als should be able to account for only a tiny fraction of
GWA studies for this trait. GWA studies have identified the heritable variation. Currently, it is unknown why this
over 180 genes affecting height. Each of these genes has a is the case, although there are some possible explanations.
small additive effect (∼1 to 4 mm), as expected for a trait For example, even with tens of thousands of individuals,
GWA studies do not have enough statistical power to detect are now targets for the development of new therapies.
genes of very small effect. Thus, one hypothesis is that sus- Beyond humans, GWA studies have advanced our under-
ceptibility for many common diseases (or height variation) standing of the inheritance of quantitative traits in Arabi-
is largely controlled by very many genes but each with a dopsis, Drosophila, yeast, and maize.
very small effect.
Despite the inability of GWA studies to explain all the KEY CONCEPT GWAS is a method for identifying statistical
heritable variation for traits, this approach has provided a associations between molecular markers and phenotypic vari-
major advance in understanding quantitative genetic vari- ation for complex traits. Linkage disequilibrium in a population
ation. Hundreds of new genes contributing to quantitative between the marker locus and a functional variant in a gene
can cause the association.
variation for disease risk have been identified. These genes
SUMMARY
Quantitative genetics seeks to understand the inheritance quantitative variation, the genetic deviation (g) is decom-
of complex traits—traits that are influenced by a mix of posed into the additive (a) and dominance (d) deviations.
genetic and environmental factors and do not segregate in Only the additive deviation is transmitted from parents to
simple Mendelian ratios. Complex traits can be categorical offspring. The additive deviation represents the heritable
traits, threshold traits, counting (meristic) traits, or continu- part of the phenotype in the narrow sense. The additive part
ously variable traits. of the variance in a population is the heritable part of the
The genetic architecture of a trait is the full description variance. Narrow-sense heritability (h2 ) is the ratio of the
of the number of genes affecting the trait, their relative con- additive variance to the phenotypic variance. Narrow-sense
tributions to the phenotype, the contribution of environ- heritability provides a measure of the degree to which the
mental factors to the phenotype, and an understanding of phenotypes of individuals are determined by the genes they
how the genes interact with one another and with environ- inherit from their parents.
mental factors. To decipher the genetic architecture of com- A knowledge of the narrow-sense heritability of a trait
plex traits, quantitative geneticists have developed a simple is fundamental to understanding how a trait will respond to
mathematical model that decomposes the phenotypes of selective breeding or the force of natural selection. Plant and
individuals into differences that are due to genetic factors animal breeders use their knowledge of narrow-sense herita-
(g) and those that are due to environmental factors (e). bility for traits of interest to guide plant and animal improve-
The differences in trait values among members of a ment programs. Narrow-sense heritability is used to predict
population can be summarized by the variance, a statisti- the phenotypes of offspring and estimate the breeding value
cal measure of the extent to which individuals deviate from of individual members of the breeding population.
the population mean. The variance for a trait can be parti- The genetic loci underlying variation in complex traits
tioned into a part that is due to genetic factors (the genetic are known as quantitative trait loci, or QTL for short. There
variance) and a part that is due to environmental factors are two experimental methods for characterizing QTL and
(the environmental variance). A key assumption behind par- determining their locations in the genome. First, QTL map-
titioning the trait variance into genetic and environmental ping looks for statistical correlations between the genotypes
components is that genetic and environmental factors are at marker loci and trait values in populations with known
uncorrelated or independent. pedigrees such as a BC1 population. Subsequently, the QTL
The degree to which variation for a trait in a population can be fine-mapped to the underlying gene. Second, associ-
is explained by genetic factors is measured by the broad- ation mapping looks for statistical correlations between the
sense heritability (H 2 ) of the trait. H 2 is the ratio of the genotypes at marker loci and trait values in random-mating
genetic variance to the phenotypic variance. Broad-sense populations. Genome-wide association (GWA) studies can
heritability expresses the degree to which the differences in directly identify the gene that underlie a QTL.
phenotype among the individuals in a population are deter- Most traits of importance in medicine, agriculture, and
mined by differences in their genotypes. The measurement evolutionary biology show complex inheritance. Exam-
of H 2 in humans has revealed that most traits have some ples include disease risk in humans, yield in soybeans, milk
genetic influences. production in dairy cows, and the full spectrum of pheno-
Parents transmit genes but not genotypes to their off- types that differentiate all the species of plants, animals, and
spring. At each generation, new dominance interactions microbes on earth. Quantitative genetic analyses are at the
between the alleles at a locus are created. To incorpo- forefront of understanding the genetic basis of these critical
rate this phenomenon into the mathematical model for traits.
KEY TERMS
additive effect (A) (p. 656) dominance effect (D) (p. 656) nearly isogenic line (p. 667)
additive gene action (p. 656) dominant gene action (p. 656) normal distribution (p. 647)
additive genetic variance (p. 658) environmental variance (p. 650) partial dominance (p. 656)
association mapping (p. 669) fine-map (p. 667) precision medicine (p. 672)
breeding value (p. 658) frequency histogram (p. 647) population (p. 645)
broad-sense heritability (H 2 ) gene action (p. 655) QTL mapping (p. 663)
(p. 652) genetic architecture (p. 644) quantitative genetics (p. 644)
candidate gene (p. 669) genetic variance (p. 650) quantitative trait (p. 644)
categorical trait (p. 645) genome-wide association (GWA or quantitative trait loci (QTL)
complex inheritance (p. 645) GWAS) (p. 669) (p. 663)
complex trait (p. 644) inbred line or stock (p. 649) sample (p. 645)
congenic line (p. 667) isogenic (p. 667) selection differential (S) (p. 662)
continuous trait (p. 645) mean (p. 645) selection response (R) (p. 662)
correlation (p. 651) meristic trait (p. 645) simple inheritance (p. 645)
correlation coefficient (p. 651) multifactorial hypothesis (p. 644) standard deviation (p. 647)
covariance (p. 650) narrow-sense heritability (h2 ) threshold trait (p. 645)
deviation (p. 647) (p. 655) variance (p. 647)
SOLVED PROBLEMS
and broad heritability is What seems unexpected about these results? How could the
unexpected results be explained?
46
H2 = = 0.75 (75%)
61 SOLUTION
Broad-sense heritability is the ratio of the total genetic vari-
SOLVED PROBLEM 3 ance (Vg ) to the phenotypic variance (VX ). The total genetic
In an experimental population of Tribolium (flour bee- variance includes both the additive (Va ) and the dominance
tles), body length shows a continuous distribution with a (Vd ) variance
mean of 6 mm. A group of males and females with a mean Vg V + Vd
body length of 9 mm are removed and interbred. The body h2 = = a
VX VX
lengths of their offspring average 7.2 mm. From these data,
calculate the heritability in the narrow sense for body length Narrow-sense heritability is the ratio of the additive vari-
in this population. ance (Va ) to the phenotypic variance (VX ).
SOLUTION Va
h2 =
The selection differential (S) is 9 − 6 = 3 mm, and the selec- VX
tion response (R) is 7.2 − 6 = 1.2 mm. Therefore, the herita-
Thus, all other variables being equal, H 2 should be greater
bility in the narrow sense is
than or equal to h2 . It will be equal to h2 when Vd is 0.0. It
R 1.2 is unexpected that h2 should be greater than H 2. However,
h2 = = = 0.4(40%)
S 3.0 the two research teams studied different populations—in
Iceland and in the United States. Estimates of heritability
SOLVED PROBLEM 4 apply only to the population and environment in which
One research team reports that the broad-sense heritability they were measured. Estimates made in one population can
for height in humans is 0.5 based on a study of identical be different from those made in another population because
twins reared apart in Iceland. Another team reports that the the two populations may segregate for different alleles at
narrow-sense heritability for human height is 0.8 based on numerous genes and the two populations experience differ-
a study of parent–offspring correlation in the United States. ent environments.
PROBLEMS
Visit SaplingPlus for supplemental content. Problems with the icon are available for review/grading.
or SNP6 show evidence for an association with the a. What are the heights of very tall and very short
growth phenotype? Explain. women?
5. Figure 19-18a shows a plot of P values (represented b. In a population of 10,000 women, how many are
by the dots) along the chromosomes of the dog expected to be very tall and how many very short?
genome. Each P value is the result of a statistical test of 12. A bean breeder is working with a population in which
association between a SNP and body size. Other than the mean number of pods per plant is 50 and the vari-
the cluster of small P values near IGF1, do you see ance is 10 pods2. The broad-sense heritability is known
any chromosomal regions with evidence for a sig- to be 0.8. Given this information, can the breeder be
nificant association between a SNP and body size? assured that the population will respond to selection
Explain. for an increase in the number of pods per plant in the
6. Figure 19-19 shows plots of P values (represented next generation?
by the dots) along the chromosomes of the human 13. The table below shows the number of piglets per litter
genome. Each P value is the result of a statistical test for a group of 60 sows. What is the mean number of
of association between a SNP and a disease condition. piglets per litter? What is the relative frequency of lit-
There is a cluster, or spike, of statistically significant ters with at least 12 piglets?
P values (green dots) at the gene HLA-DRB1 for two
diseases. Why might this particular gene contribute to Number of litters Piglets/litter
susceptibility for the autoimmune diseases rheumatoid 1 6
arthritis and type 1 diabetes?
3 7
7 8
BASIC PROBLEMS
12 9
7. Distinguish between continuous and discontinuous 18 10
variation in a population, and give some examples of
20 11
each.
17 12
8. What are the central assumptions of the multifactorial
hypothesis? 14 13
6 14
9. The table below shows a distribution of bristle number
in a Drosophila population. Calculate the mean, vari- 2 15
ance, and standard deviation for these data.
14. A chicken breeder is working with a population in
Bristle number Number of individuals which the mean number of eggs laid per hen in one
1 1 month is 28 and the variance is 5 eggs2 . The nar-
2 4 row-sense heritability is known to be 0.8. Given this
information, can the breeder expect that the popula-
3 7 tion will respond to selection for an increase in the
4 31 number of eggs per hen in the next generation?
5 56 a. No, applying selection is always risky and a
6 17 breeder never knows what to expect.
7 4 b. No, a breeder needs to know the broad-sense heri-
tability to know what to expect.
10. Suppose that the mean IQ in the United States is c. Yes, since the narrow-sense heritability is close to
roughly 100 and the standard deviation is 15 points. 1(0.8), then we would expect selective breeding
People with IQs of 145 or higher are considered could lead to increased egg production in the next
“geniuses” on some scales of measurement. What per- generation.
centage of the population is expected to have an IQ of
145 or higher? In a country with 300 million people, d. Yes, since the variance is greater than 0.
how many geniuses are there expected to be? e. Both c and d are correct.
11. In a sample of adult women from the United States, 15. The narrow-sense heritability of the number of peas
the average height was 164.4 cm and the standard per pod in a population of sugar snap peas is 0.5. The
deviation was 6.2 cm. Women who are more than mean of the population is 6.2 peas per pod. A plant
2 standard deviations above the mean are considered breeder selects one plant with 6.8 peas per pod and
very tall, and women who are more than 2 standard crosses with a second plant that has 8.0 peas per pod.
deviations below the mean are considered very short. What is the expected numbers of peas per pod among
Height in women is normally distributed. the offspring of this cross?
16. QTL mapping and GWA (association) mapping are the ability of the finches to eat large seeds. The mean
two different methods used to identify genes that affect bill depth for the population is 9.6 mm. A male with a
complex traits. For each of the following statements, bill depth of 10.8 mm is mated with a female with a bill
choose whether it applies to QTL mapping, associa- depth of 9.8 mm. What is the expected value for
tion mapping, or both. bill depth for the offspring of this mating pair?
Statement QTL GWA Both 19. Two inbred lines of laboratory mice are intercrossed.
This method requires that the In the F1 (which have identical genotypes at all loci),
experimenter make crosses the variance in adult weight is measured at 3 g2 . The
between different strains to F1 animals are intercrossed to create an F2 in which the
produce a mapping population. variance in adult weight is 16 g 2 . Estimate the broad-
sense heritability of adult weight in the F2 population
This method can scan the entire
of this experiment. (The environments in which the F1
genome to find QTL for a trait.
and F2 animals were reared were equivalent.)
This method can often identify the
20. The table below shows the weights of 100 individual
specific genes that represent the
mice of the same inbred strain reared on different diets.
QTL.
For an individual mouse that weights 27 g, how much
This method may sample a large of its weight is due to its genetics and how much to
number of individuals from a the specific diet it was fed (environment)? (Other than
random-mating population that diet, the mice were reared in equivalent environments.)
has variation for the trait being
studied. Number of mice Weight (g)
This method typically tests two 5 21
alleles that differ between the 13 22
two parents of the mapping
18 23
population.
21 24
22 25
CHALLENGING PROBLEMS
16 26
17. In a large herd of cattle, three different characters
5 27
showing continuous distribution are measured, and
the variances in the following table are calculated:
21. The table below contains measurements of total serum
Characters cholesterol (mg/dl) for 10 sets of monozygotic twins
Shank Neck Fat who were reared apart. Calculate the following: over-
Variance length length content all mean, overall variance, covariance between the
twins, and broad-sense heritability (H 2 ) .
Phenotypic 310.2 730.4 106.0
Environmental 248.1 292.2 53.0 X′ X ′′
Additive genetic 46.5 73.0 42.4 228 222
Dominance genetic 15.6 365.2 10.6 186 152
204 220
a. Calculate the broad- and narrow-sense heritabili-
142 185
ties for each character.
226 210
b. In the population of animals studied, which char-
acter would respond best to selection? Why? 217 190
207 226
c. A project is undertaken to decrease mean fat con-
tent in the herd. The mean fat content is currently 185 213
10.5 percent. Animals with a mean of 6.5 percent 179 159
fat content are interbred as parents of the next gen- 170 129
eration. What mean fat content can be expected in
the descendants of these animals? 22. The table on the next page contains the height in centi-
18. In a species of the Darwin’s finches (Geospiza fortis), meters for 10 sets of adult women twins. Calculate the
the narrow-sense heritability of bill depth has been correlation coefficient (r) between the heights of the
estimated to be 0.79. Bill depth is correlated with sisters for the twin pairs.
Twin 1 Twin 2 The mean trait values (pulses per second) at the three
genotypic classes in the F2 for each QTL are shown
158 163 in the table below, where P indicates the L. paranigra
156 150 allele and K indicates the L. kohalensis allele.
172 173
QTL P/P P/K K/K
156 154
1 1.54 1.89 2.10
160 163
2 1.75 1.87 1.94
159 153
3 1.72 1.88 1.92
170 174
4 1.70 1.82 2.02
177 174
5 1.67 1.80 2.13
165 168
6 1.57 1.88 2.19
172 165
a. Calculate the additive (A) and dominance (D)
23. Population A consists of 100 hens that are fully effects and the D/A ratio for each of the six QTL.
isogenic and that are reared in a uniform environment.
b. Which of these QTL shows the greatest amount of
The average weight of the eggs they lay is 52 g, and the
dominance?
variance is 3.5 g2. Population B consists of 100 genet-
ically variable hens that produce eggs with a mean c. Which of these has the largest additive effect?
weight of 52 g and a variance of 21.0 g2. Population d. The mean pulse rate for L. kohalensis is 3.72, and
B is raised in an environment that is equivalent to that it is 0.71 for L. paranigra. Do all six QTL act in
of Population A. What is the environmental variance the expected direction with the L. kohalensis allele
(Ve ) for egg weight? What is the genetic variance in conferring a higher pulse rate than the L. paranigra
Population B? What is the broad-sense heritability in allele?
Population B?
27. Question 26 refers to QTL on the cricket autosomes.
24. Maize plants in a population are on average 180 cm For the sex chromosomes, female crickets are XX
tall. Narrow-sense heritability for plant height in this and male crickets are XO, having just one X chromo-
population is 0.5. A breeder selects plants that are some but no Y chromosome. Can QTL for pulse rate
10 cm taller on average than the population mean be mapped on cricket X chromosomes? If the song is
to produce the next generation, and the breeder con- sung only by male crickets, can the dominance effects
tinues applying this level of selection for eight gener- of QTL on the X be estimated?
ations. What will be the average height of the plants
28. GWA studies reveal statistical correlations between the
after eight generations of selection? Assume that h2
genotypes at marker loci in genes and complex traits.
remains 0.5 and Ve does not change over the course of
Do GWA studies prove that allelic variation in a gene
the experiment.
actually causes the variation in the trait? If not, what
25. In a population of Drosophila melanogaster reared in experiments could prove that allelic variants in a gene
the laboratory, the mean wing length is 0.55 mm and in a population are responsible for variation in a trait?
the range is 0.35 to 0.65. A geneticist selects a female
29. The ocular albinism-2 ( OCA2 ) gene and the
with wings that are 0.42 mm in length and mates her
melanocortin-1-receptor ( MC1R ) gene are both
with a male that has wings that are 0.56 mm in length.
involved in melanin metabolism in skin cells in
a. What is the expected wing length of their offspring humans. To test whether variation at these genes con-
if wing length has a narrow-sense heritability of tributes to sun sensitivity and the associated risk of
1.0? being afflicted with skin cancer, you perform associ-
b. What is the expected wing length of their offspring ation analyses. A sample of 1000 people from Iceland
if wing length has a narrow-sense heritability of were asked to classify themselves as having tanning or
0.0? burning (nontanning) skin when exposed to the sun.
The individuals were also genotyped for a SNP in each
26. Different species of crickets have distinct songs, and
gene (rs7495174 and rs1805007). The table shows the
they use these songs for mate recognition. Research-
number of individuals in each class.
ers crossed two species of Hawaiian crickets (Laupala
paranigra and L. kohalensis) whose songs are distin- OCA2 (rs7495174) MC1R (rs1805007)
guished by pulse rate (the number of pulses per second;
A/A A/G G/G C/C C/T T/T
Shaw et al., Molecular Ecology 16, 2007, 2879–2892).
Then, they mapped QTL in the F2 population derived Burning 245 56 1 192 89 21
from this cross. Six autosomal QTL were detected. Tanning 555 134 9 448 231 19
a. What are the frequencies of tanning and burning GENETICS AND SOCIETY
phenotypes in Iceland? Bioethicists have expressed concern that consumers are
b. What are the allelic frequencies at each locus not prepared to properly evaluate the results from direct-
(SNP)? to-consumer genetic testing for disease risk alleles because
consumers lack adequate knowledge to interpret the results
c. Using the χ2 test (see Chapter 3) and Table 3-1,
appropriately and may react to test results by taking an
test the null hypothesis that there is no association
action that is damaging to their health. Should society out-
between these SNPs and sun-sensitive skin. Does
law such direct-to-consumer testing, require that a genetic
either SNP show evidence for an association?
counselor be consulted, or allow such testing?
d. If you find evidence for an association between the
gene and the trait, what is the mode of gene action?
e. If the P value is greater than 0.05, does that prove
that the gene does not contribute to variation for
sun sensitivity? Why?
C
harles Darwin (1809–1882) arrived in the Darwin’s insight was that species might change. This
G alápagos Islands in 1835, well into the fourth was not what he had learned at Cambridge University.
year of what was supposed to be a two-year voy- This was heresy. Although Darwin decided to keep such
age. One might think that these islands, now inextricably dangerous thoughts to himself, he was gripped by the idea.
linked with Darwin’s name, were the young naturalist’s After arriving home in England, he filled a series of note-
paradise. Far from it. Darwin found the islands hellishly books with thoughts about species changing. Within a year
hot, their broken black volcanic rock scorching under the he had convinced himself that species arise naturally from
hot sun. In his diary he observed that “the stunted trees preexisting species, as naturally as children are born from
show little signs of life . . . the plants also smell unpleas- parents and parents from grandparents. He then pondered
antly. . . . The black lava rocks on the beach are frequented how species change and adapt to their particular circum-
by large (2–3 ft.) most disgusting clumsy lizards. . . . They stances. In 1838, just two years after the conclusion of his
assuredly well become the land they inhabit.”1 Other than voyage and before he had yet turned 30, he conceived his
the lizards and the tortoises, the animal life on the islands answer—natural selection. In this competitive process,
was scant and unimpressive. He could not wait to leave the individuals bearing some relative advantage over others live
place. The 26-year-old explorer did not know that his five longer and produce more offspring, which in turn inherit
weeks in the Galápagos would inspire a series of radical the advantage.
ideas that, some 24 years later with the publication of his Darwin knew that to convince others of these two
On the Origin of Species (1859), would change our percep- ideas—the descent of species from ancestors and natural
tion of the world and our place in it. selection—he would need more evidence. He spent the next
Several months after leaving the islands, on the last leg two decades marshaling all of the facts he could from bot-
of the voyage home to England, Darwin had his first flash any, zoology, embryology, and the fossil record.
of insight. He had begun to organize his copious field notes He received crucial information from experts who
from his nearly five years of exploration and collecting. His helped to sort out and characterize his collections. For
plan was for experts back in England to lead the study of example, ornithologist John Gould pointed out to Darwin
his collections of fossils, plants, animals, and rocks. Turn- that what the young naturalist thought were blackbirds,
ing to his observations on the birds of the Galápagos, he grosbeaks, and finches from the Galápagos were actually
recalled that he had found slightly different forms of mock- 13 new and distinct species of ground finches (Figure 20-1).
ingbirds on three different islands. Now, there was a puzzle. We now know that the Galápagos species, though clearly
The prevailing view of the origin of species in 1835, held by finches, exhibit an immense variation in feeding behavior
most of Darwin’s teachers and much of the scientific estab- and in the bill shape that corresponds to their food sources.
lishment, was that species were specially created by God in For example, the vegetarian tree finch uses its heavy bill to
their present form, unchangeable, and placed in the habitat eat fruits and leaves, the insectivorous finch has a bill with
to which they were best suited. Why, then, would there be a biting tip for eating large insects, and, most remarkable of
slightly different birds on such similar islands? Darwin jot- all, the woodpecker finch grasps a twig in its bill and uses it
ted in his ornithology notebook: to obtain insect prey by probing holes in trees.
This diversity of species, Darwin deduced, must have
When I see these Islands in sight of each other and
arisen from an original population of finch that arrived in
possessed of but a scanty stock of animals, tenanted
the Galápagos from the mainland of South America and
by these birds but slightly differing in structure filling
populated the islands. The descendants of the original col-
the same place in Nature, I must suspect they are
onizers spread to the different islands and formed local
only varieties. . . . If there is the slightest foundation
populations that diverged from one another and eventually
for these remarks, the zoology of Archipelagoes
formed different species.
will be well worth examining; for such facts would
The finches illustrate the process of adaptation, in
undermine the stability of species [emphasis added].2
which the characteristics of a species change over time if
1
those traits increase the chance of survival and reproduc-
C. Darwin, Charles Darwin’s Beagle Diary, Ed. R. D. Keynes, Cambridge
University Press, 2001.
tion in the environment in which the species lives. A mea-
2
C. Darwin, Charles Darwin’s Beagle Diary, Ed. R. D. Keynes, Cambridge sure of an organism’s ability to survive and reproduce is
University Press, 2001. called its fitness. Darwin provided one level of explanation
682
Seed eaters
Seed eaters have bills that are adapted
for collecting and crushing seeds.
Bud eater
The bud eater has a heavy bill adapted
for pulling buds from branches.
Vegetarian finch
(Platyspiza crassirostris)
ANCESTOR FINCH
from South American
mainland
Insect eaters
Insect eaters have a variety of bills adapted
to eating different sizes and types of insects,
which they capture in different ways.
Mangrove finch
(C. heliobates) The woodpecker finch has a long
beak it uses to search for insects
Woodpecker finch in crevices in dead wood and bark.
(C. pallidus)
for the process, natural selection, but he could not explain incorporated the role of inheritance. Darwin’s theory of
how traits varied or how they changed with time because evolution by natural selection begins with the variation
he did not understand the mechanisms of inheritance. that exists among organisms within a species. Individu-
Understanding the genetic basis of adaptation has been one als of one generation are qualitatively different from one
of the long-standing goals of evolutionary biology. another. Evolution of the species as a whole results from
A first step toward this goal was taken when Mendel’s the fact that the various types differ in their rates of sur-
work pointing to the existence of genes was rediscovered vival and reproduction. Better-adapted types leave more
two decades after Darwin died. Another key emerged a offspring. When offspring inherit the type of their parents,
half century later, when the molecular basis of inheritance the relative frequencies of the types will change over time.
and the genetic code were deciphered. For many decades Thus, the four critical ingredients to evolutionary change
since, biologists have known that species and traits evolve that Darwin put forth were variation, selection, inheritance,
through changes in DNA sequence. However, elucidating and time. Darwin said:
the specific changes in DNA sequence underlying physio-
Can it, then, be thought improbable . . . that
logical or morphological evolution has been fraught with
variations useful in some way to each being in the
considerable technical challenges. Advances in molec-
great and complex battle of life, should sometimes
ular genetics (Chapters 7–12), developmental genetics
occur in the course of thousands of generations? . . .
(Chapter 13), and comparative genomics (Chapter 14) are
Can we doubt (remembering that many more
now revealing the diverse mechanisms underlying the evo-
individuals are born than can possibly survive)
lution of genes, genomes, traits, and species.
that individuals having any advantage, however
The study of evolution is a very large and expanding
slight, over others, would have the best chance of
discipline. As such, we will not attempt a comprehensive
surviving and of procreating their kind? On the
overview of all facets of evolutionary analysis. Rather, in
other hand, we may feel sure that any variation
this chapter, we will examine the molecular genetic mech-
in the least degree injurious would be rigidly
anisms underlying the variation in and evolution of traits
destroyed. This preservation of favorable variations
and the adaptation of organisms to their environments. We
and the rejection of injurious variations I call
will first examine the evolutionary process in general and
Natural Selection. (On the Origin of Species,
then focus on specific examples for which the genetic and
Chapter IV)3
molecular bases of the phenotypic differences between pop-
ulations or species have been pinpointed. All of the exam- Darwin’s writings and ideas are well known, and jus-
ples will focus on the evolution of relatively simple traits tifiably so, but it is very important to note that he was
controlled by one or a few genes. These relatively simple not alone in arriving at this concept of natural selection.
examples are sufficient to illustrate the fundamental pro- Alfred Russel Wallace (1823–1913), a fellow Englishman
cess of evolution at the DNA level and the variety of ways who explored the jungles of the Amazon and the Malay
in which the evolution of genes affects the gain, loss, and Archipelago for a total of 12 years, reached a very similar
modification of traits, as well as the formation of species. conclusion in a paper that was copublished with an excerpt
from Darwin in 1858:
The life of wild animals is a struggle for existence.
20.1 EVOLUTION BY NATURAL . . . Perhaps all the variations from the typical form
of a species must have some definite effect, however
SELECTION slight, on the habits or capacities of the individuals.
. . . It is also evident that most changes would
LO 20.1 Identify and explain the principles of evolution by
affect, either favourably or adversely, the powers
natural selection.
of prolonging existence. . . . If, on the other hand,
any species should produce a variety having slightly
The modern theory of evolution is so completely identified
increased powers of preserving existence, that
with Darwin’s name that many people think Darwin him-
variety must inevitably in time acquire a superiority
self first proposed the concept that organisms have evolved,
in numbers.4
but that is not the case. The idea that life changed over time
was circulating in scientific circles for many decades before While today Darwin’s name tends to be exclusively
Darwin’s historic voyage. The great question was, How did linked to evolution by natural selection, in their day, the the-
life change? For some, the explanation was a series of special ory was recognized as the Darwin-Wallace theory. Perhaps
creations by God. To others, such as Jean-Baptiste Lamarck
3
(1744–1829), change was caused by the environment acting C. Darwin, On the Origin of Species by Means of Natural Selection,
directly on the organism, and those changes acquired in an 80–81. John Murray, London, 1859.
4
C. Darwin and A. Wallace, “On the Tendency of Species to Form Varie
organism’s lifetime were passed on to its offspring. ties; and on the Perpetuation of Varieties and Species by Natural Means of
What Darwin provided was a detailed explanation of Selection,” Journal of the Proceedings of the Linnean Society of London.
the mechanism of the evolutionary process that correctly Zoology 3, 1858, 45–50.
the current perception is at least in part due to Wallace him- Heritable variation provides the raw material for suc-
self, who was always deferential to Darwin and referred to cessive changes within a species and for the multiplication
the emergent theory of evolution as “Darwinism.” of new species. The basic mechanisms of those changes (as
discussed in Chapter 18) are the origin of new genetic and
KEY CONCEPT Darwin and Wallace proposed a new phenotypic variation in a population by mutation or migra-
explanation to account for the phenomenon of evolution. tion and the change in frequency of alleles and phenotypes
They understood that the population of a given species at within populations by selective and random processes
a given time includes individuals of varying characteristics. (Figure 20-2). From those basic mechanisms, a set of prin-
They realized that the population of succeeding generations
ciples governing changes in the genetic composition of
will contain a higher frequency of those types that most
successfully survive and reproduce under the existing
environmental conditions. Thus, the frequencies of various The evolutionary forces that influence variation
types within the species will change over time.
aa aa aa aa
There is an obvious similarity between the process of evo-
aa Mutation aa
lution as Darwin and Wallace described it and the process by aa aa aa aa
which the plant or animal breeder improves a domestic stock.
aa Aa
The plant breeder selects the highest-yielding plants from the aa
aa
aa
aa
current population and uses them as the parents of the next
generation. If the characteristics causing the higher yield are
heritable, then the next generation should produce a higher
yield. It was no accident that Darwin chose the term natural
selection to describe his model of evolution through differ- AA AA aa aa
ences in the rates of reproduction shown by different variants
Migration aa
in the population. As a model for this evolutionary process in AA AA AA aa aa
the wild, he had in mind the selection that breeders exercise
AA aa
on successive generations of domestic plants and animals. AA AA AA aa
We can summarize the theory of evolution by natural
selection in three principles:
1. Principle of variation. Among individuals within any pop-
ulation, there is variation in morphology, physiology, and
Aa aa
behavior.
aa
2. Principle of heredity. Offspring resemble their parents Genetic drift
Aa Aa AA Aa
more than they resemble unrelated individuals. AA
populations can be derived. The application of these prin- complications by occluding blood vessels and lead to early
ciples of population genetics provides a genetic theory of mortality.
evolution. In 1949, the very year Allison went into the field,
Linus Pauling’s research group demonstrated that patients
KEY CONCEPT Evolution, the change in populations or with sickle-cell anemia had a hemoglobin protein with an
s pecies over time, is the conversion of heritable variation abnormal charge (Hemoglobin S, or HbS) in their blood,
between individuals within populations into heritable compared with the hemoglobin of unaffected individuals
differences between populations in time and in space by (Hemoglobin A, or HbA). This was the first demonstration
population genetic mechanisms.
of a molecular abnormality linked to a complex disease. It
was generally understood at the time that carriers of sickle
20.2 NATURAL SELECTION IN cell were heterozygous and thus had a mixture of HbA
and HbS (denoted AS), whereas affected individuals were
ACTION: AN EXEMPLARY homozygous for the HbS allele (denoted SS).
CASE Allison collected blood specimens from members of the
Kikuyu, Masai, Luo, and other tribes across the very diverse
LO 20.2 Summarize the various processes of evolution and geography of Kenya. While he did not see any particularly
the role they play in the evolution of genes, traits, striking association between ABO or MN blood types
and species. among the tribes, he measured remarkably different fre-
quencies of HbS. In tribes living in arid central Kenya or in
For nearly a century after the publication of On the Origin the highlands, the frequency of HbS was less than 1 percent;
of Species, not one example of natural selection had been however, in tribes living on the coast or near Lake Victoria,
fully elucidated, that is, where the agent of natural selec- the frequency of HbS often exceeded 10 percent and
tion was known, the effect on different genotypes could be approached 30 percent in some locations (Table 20-1).
measured, the genetic and molecular basis of variation was The allele frequencies were surprising for two reasons.
identified, and the physiological role of the gene or protein First, since sickle-cell anemia was usually lethal, why were
involved was well understood. the frequencies of the HbS allele so high? And second, given
The first such “integrated” example of natural selec- the relatively short distances between regions, why was the
tion on a molecular variant was elucidated in the 1950s, HbS frequency high in some places and not in others?
before the genetic code was even deciphered. Remarkably, Allison’s familiarity with the terrain, tribes, and trop-
this trailblazing work revealed natural selection operating ical diseases of Kenya led him to the crucial explanation.
on humans. It still stands today as one of the most detailed Allison realized that the HbS allele was at high frequency
and important examples of evolution by natural selection in low-lying humid regions with very high levels of
in any species. malaria, and nearly absent at high altitudes such as around
The story began when Tony Allison, a Kenyan-born Nairobi. Carried by mosquitoes, the intracellular parasite
Oxford medical student, undertook a field study of blood Plasmodium falciparum, which causes malaria, multiplies
types among Kenyan tribes. One of the blood tests he inside red blood cells ( Figure 20-4). Mosquitoes and the
ran was for sickle cells, red blood cells that form a sickle d isease are prevalent throughout sub-Saharan Africa in
shape on exposure to the reducing agent sodium beta- humid, low-lying regions near bodies of water where the
sulfite or after standing for a few days (Figure 20-3). The mosquitoes reproduce. Allison surmised that the HbS allele
deformed cells are a hallmark of sickle-cell anemia, a dis- might, by altering red blood cells, confer some degree of
ease first described in 1910. These cells cause pathological resistance to malarial infection.
Malarial parasites live within red blood cells balanced by natural selection operating against AA individ-
uals susceptible to malaria and SS individuals who would
succumb to sickle-cell anemia. In other words, the hetero-
zygote has an advantage over either of the homozygotes.
How much of an advantage do AS individuals experi-
ence? This can be calculated by measuring the frequency
of the HbS allele in populations and examining how these
frequencies differ from the frequencies expected under
the assumptions of the Hardy–Weinberg equation (see
Chapter 18). A large-scale survey of 12,387 West Africans
revealed an HbS allele frequency (q) of 0.123 (see Box 18.1
for assistance in calculating allele frequencies). The fre-
quencies calculated from the Hardy–Weinberg equation
(expected phenotype frequencies in Table 20-2) are higher
than what was actually observed for the homozygous
FIGURE 20-4 A blood smear of an individual infected with malarial
phenotypes and lower for the heterozygous phenotype
parasites. A red blood cell sample was treated with Giemsa stain to (Table 20-2). If it is assumed that the AS heterozygote has
reveal parasites within cells (red dots). [CDC/Dr. Mae Melvin.] a fitness of 1.0, then the relative fitness of the two homozy-
gous genotypes can be estimated. For example, the relative
fitness of the homozygous AA genotype compared to the
found HbS frequencies of up to 40 percent in malarial areas AS genotype is 0.88. The AS genotype thus has a selective
and frequencies as low as 0 percent where malaria was advantage of 1.136, or approximately 14 percent, over the
absent. AA genotype.
The link suggested that the HbS allele might affect para- This selective advantage has been well documented by
site levels, so Allison also undertook a study of the level of long-term survival studies of AA, AS, and SS children in
parasites in the blood of heterozygous AS children versus Kenya. These studies have found that AS individuals have a
wild-type AA children. In a study of nearly 300 children, he pronounced survival advantage over AA and SS individuals
found the incidence of malarial parasites was indeed lower in the first few years of life (Figure 20-5).
in AS children (27.9 percent) than in AA children (45.7 per-
cent) and that parasite density was also lower in AS chil- KEY CONCEPT The sickle-cell hemoglobin allele, HbS, is
dren. The results indicated that AS children had a lower under balancing selection in malarial zones and conveys a
incidence and severity of malarial infection and would large survival advantage in heterozygotes over the first few
thus have a selective advantage in areas where malaria was years of life.
prevalent.
The advantage to AS heterozygotes was especially
striking in light of the disease suffered by SS homozygotes.
Allison noted: The molecular origins of HbS
After Allison’s discovery, there was keen interest in deter-
The proportion of individuals with sickle cells in
mining the molecular basis of the difference(s) between HbS
any population . . . will be the result of a balance
and Hb A. Protein sequencing determined that HbS differs
between two factors: the severity of malaria, which
from Hb A by just one amino acid, a valine in place of a
will tend to increase the frequency of the gene, and
glutamic acid residue. This single amino acid change alters
the rate of elimination of the sickle-cell genes in
the charge of hemoglobin and causes it to aggregate into
individuals dying of sickle-cell anaemia.
long rodlike structures within red blood cells. Once the
The sickle-cell mutation was under balancing selec- genetic code was deciphered and methods for sequencing
tion (see Chapter 18) in areas where malaria was present. DNA were developed, HbS was determined to be caused by
Natural selection operating in favor of AS individuals is a single point mutation (CTC → CAC) in the glutamic acid
0.87
HbAA
0.82
0.77 HbSS
0.72
0 30 180 360 540 720 900 1080 1260 1440 1620 1800 2140
Time until death (days)
codon encoding the sixth amino acid of the β-globin sub- Allison composed maps showing these striking correlations
unit within the hemoglobin protein. (Figure 20-6) and inferred that the HbS alleles in different
Interestingly, Allison also noted a high incidence of HbS regions arose independently, rather than through spread-
outside of Africa, including in Italy, Greece, and India. Other ing by migration. Indeed, with the advent of tools for DNA
blood-type markers did not indicate strong genetic relation- genotyping, it is clear that the HbS mutation has arisen inde-
ships among these populations. Rather, Allison observed pendently at least five times and then increased to high fre-
that these were also areas with a high incidence of malaria. quency in particular regions. Based on the limited genetic
The correlation between HbS frequency and the incidence diversity of malarial populations, it is believed that HbS
of malaria held across not only East Africa, but the African mutations arose in just the past several thousand years, once
continent, southern Europe, and the Indian subcontinent. populations began living around bodies of water as part of
the advent of agriculture.
Elucidating the role of HbS in conferring resistance
The geography of sickle-cell hemoglobin and malaria
to malaria illustrated three important facets of the evo-
lutionary process:
1. E
volution can and does repeat itself. The multiple
independent origins and expansions of the HbS
mutation demonstrate that the same mutations can
arise and spread repeatedly. Many examples are
now known of the precise, independent repetition
Distribution of the evolution of adaptive mutations, and we will
of malaria encounter several more in this chapter.
2. F
itness is a relative, conditional status. Whether a
mutation is advantageous, disadvantageous, or nei-
ther depends very much on environmental condi-
Frequency of sickle-cell trait tions. In the absence of malaria, HbS is very rare
and disfavored. Where malaria is present, HbS can
>20% 1–10%
reach high frequencies despite the disadvantages
15–20% <1%
imparted to SS homozygotes. In African Amer-
FIGURE 20-6 These maps show the close correspondence between the icans, the frequency of HbS is declining because
distribution of malaria (left) and the frequency of the sickle-cell trait (right) across there is selection against the allele in the absence of
Africa. [Data from A. C. Allison, Genetics 66, 2004, 1591; redrawn by Leanne Olds.] malaria in North America.
3. Natural selection acts on whatever variation is avail- Proteins differ in substitution rate
able, and not necessarily by the best means imaginable.
220
s
ile
pt
the heterozygous state), also causes a life-threatening
ey
200
al s
s
ep s/re
sh
s/
m tile
al
pr
s te
/fi
m
m
am p
ct r a
es
M s/re
am
condition (in the homozygous state). Over 40 percent
la
se eb
til
p/
180
in ert
rd
ar
Bi
V
of the world’s population lives in areas where malaria 160
is prevalent. In these places, the imperative of combat-
140
ing malaria counterbalances the deleterious effect of the Fibrinopeptides
sickle-cell mutation. 120
100
Hemoglobin
80
natural selection. Moreover, it provided a baseline assump- since their divergence from a common ancestor, the result
tion of how DNA should change over time if no other agent should be a straight line with slope equal to µ . That is,
such as natural selection intervened. evolution should proceed according to a molecular clock
(see Box 18-5) that is ticking at the rate µ . Figure 20-8
KEY CONCEPT The neutral theory of molecular evolution shows such a plot for the β-globin gene. The results are
roposed that most mutations in DNA or amino acid
p quite consistent with the claim that nucleotide substitu-
replacements between species are functionally neutral or nearly tions in this gene have been neutral in the past 500 million
neutral and fixed by random genetic drift. The assumption of years. Two sorts of neutral nucleotide substitutions are
neutrality offers a baseline expectation of how DNA should
plotted: synonymous substitutions, which are from one
change over time when natural selection is absent.
codon to another, making no change in the amino acid,
and nonsynonymous substitutions, which result in an
The rate of neutral substitutions amino acid change. Figure 20-8 shows a much lower slope
for nonsynonymous substitutions than for synonymous
As we saw in Chapter 18 (see Box 18-5), we can calculate
changes, which means that the substitution rate of neutral
the expected rate of neutral changes in DNA sequences
nonsynonymous substitutions is much lower than that of
over time. If µ is the rate of new mutations at a locus per
synonymous neutral substitutions.
gene copy per generation, then the absolute number of
This outcome is precisely what we expect under natural
new mutations that will appear in a population of N dip-
selection. Mutations that cause an amino acid substitution
loid individuals is 2Nµ. The new mutations are subject to
should have a deleterious effect more often than synony-
random genetic drift: most will be lost from the popula-
mous substitutions, which do not change the protein. Such
tion, while a few will become fixed and replace the previ-
deleterious variants will be removed from populations by
ous allele. If a newly arisen mutation is neutral, then there
purifying selection (see Chapter 18). A lower-than-expected
is a probability of 1/(2N) that it will replace the previous
ratio of nonsynonymous to synonymous changes is a signa-
allele because of random genetic drift. Each one of the 2Nµ
ture of purifying selection. It is important to note that these
new mutations that will appear in a population has a prob-
observations do not show that synonymous substitutions
ability of 1/(2N) of eventually taking over that population.
have no selective constraints on them; rather, they show
Thus, the absolute substitution rate k is the mutation rate
that these constraints are, on the average, not as strong as
multiplied by the probability that any one mutation will
those for mutations that change amino acids. So, a synon-
eventually take over by drift:
ymous change, although it has no effect on the amino acid
k = rate of neutral substitution = 2Nµ × 1/(2N ) = µ sequence, does change the mRNA for that sequence and
thus may affect mRNA stability or efficiency at which the
That is, we expect that, in every generation, there
mRNA is translated.
will be µ substitutions in the population, purely from the
genetic drift of neutral mutations.
sites
on DNA sequences
3.0
When measurements of molecular change deviate from
what is expected for neutral changes, that is an important
signal—a signal that selection has intervened. That signal
may reveal that selection has favored some specific change 2.0
or that it has rejected others. We have seen, in the case of
the HbS mutation, how natural selection favors the muta-
tion in the presence of the malarial parasite, but rejects it
1.0
where malaria is absent. The most pervasive influence of
natural selection on DNA is, in fact, to conserve gene func- Nonsynonymous
sites
tion and sequence.
All classes of DNA sequences, including exons, introns,
regulatory sequences, and sequences in between genes, 0 1 2 3 4 5
show nucleotide diversity among individuals within pop- Divergence time (× 108 years)
ulations and between species. The constant rate of neutral FIGURE 20-8 The amount of nucleotide divergence at synonymous
substitutions predicts that, if the number of nucleotide dif- sites is greater than the amount of divergence at nonsynonymous sites
ferences between two species were plotted against the time of the β-globin gene.
Purifying selection is the most widespread, but often its replacement by esterase activity (splits any ester, R–O–R,
underappreciated, facet of natural selection. The “rejec- into an acid and an alcohol). Three-dimensional modeling of
tion of injurious variations,” as Darwin termed it, is per- the molecule indicates that the protein with esterase activity
vasive. Purifying selection explains why we find many gains the ability to bind a water molecule close to the site
protein sequences that are unchanged or nearly unchanged of attachment of the organophosphate. The water molecule
over vast spans of evolutionary time. For example, there then reacts with the organophosphate, splitting it in two.
are several dozen genes that exist in all domains of life—
Archaea, Bacteria, and Eukarya (fungi, plants, protists, and KEY CONCEPT There is no proportionate relation between
animals)—and encode proteins whose sequences have been how much DNA change takes place in evolution and how
largely conserved over 3 billion years of evolution. To pre- much change in function results.
serve such sequences, variants that have arisen at random
in billions of individuals in tens of millions of species have
been rejected by selection over and over again. The signature of positive selection
on DNA sequences
KEY CONCEPT Purifying selection is a pervasive aspect of
natural selection that reduces genetic variation and preserves Evidence for the molecular clock supports the idea that most
DNA and protein sequences over eons of time. nucleotide substitutions that have occurred in evolution were
neutral. However, it does not tell us how much of molecular
evolution has been adaptive change driven by positive selec-
Another prediction of the theory of neutral evolution is tion. One way of detecting the adaptive evolution of a pro-
that different proteins will have different clock rates because tein is by comparing the synonymous and nonsynonymous
the metabolic functions of some proteins will be much more nucleotide polymorphisms within species with the synony-
sensitive to changes in their amino acid sequences. Proteins mous and nonsynonymous nucleotide changes between spe-
that are more sensitive to amino acid changes will have a cies. If all mutations are neutral, the ratio of nonsynonymous
lower rate of neutral mutation because a smaller proportion to synonymous nucleotide polymorphisms within a species
of their mutations will be neutral compared with proteins should be the same as the ratio of nonsynonymous to synon-
that are more tolerant of substitution. Figure 20-7 shows a ymous nucleotide substitutions between species. On the other
comparison of the clocks for fibrinopeptides, hemoglobin, hand, if the amino acid changes between species have been
and cytochrome c. That fibrinopeptides have a much higher driven by positive selection, there ought to be an excess of
proportion of neutral mutations is reasonable because these nonsynonymous changes between species.
peptides are not known to have a function after they are cut One test for detecting positive selection on DNA
out of fibrinogen to activate the blood-clotting reaction. It sequences was developed by John McDonald and Martin
is not obvious why hemoglobins are less sensitive to amino Kreitman. This test involves several logical but simple steps:
acid changes than is cytochrome c.
1. The DNA sequence of a gene is obtained from a num-
KEY CONCEPT The rate of neutral evolution for the amino ber of separate individuals from each of two species.
acid sequence of a protein depends on the sensitivity of the Sequences from ten or more individuals of each species
protein’s function to amino acid changes. would be desirable. The fixed nucleotide differences
between species are then classified into nonsynonymous
(a in the table below) and synonymous (b in the table
Because so much sequence evolution is neutral, there below) differences.
is no simple relation between the amount of change in a 2. The nucleotide differences among individuals within
gene’s DNA sequence and the amount of change, if any, each species (polymorphisms) are then tabulated, and
in the encoded protein’s function. At one extreme, almost classified as either those that result in amino acid
the entire amino acid sequence of a protein can be replaced changes (nonsynonymous polymorphisms; c in the table
while maintaining the original function if those amino below) or those that do not change the amino acid (syn-
acids that are substituted maintain the enzyme’s three-di- onymous polymorphisms; d in the table below).
mensional structure. This is the case for the fibrinopep-
tides, in which nearly any amino acid can be substituted at 3. If the divergence between the species is purely the result
any position of the peptide. of random genetic drift, then we expect a/b to be equal to
In contrast, the function of an enzyme can be changed by c/d. If, on the other hand, there has been selective diver-
a single amino acid substitution. The sheep blowfly, Lucilia gence, there should be an excess of fixed nonsynonymous
cuprina, has evolved resistance to organophosphate insecti- differences, and so a/b should be greater than c/d.
cides used widely to control it. This resistance is the con- Fixed species
sequence of a single substitution of an aspartic acid for a differences Polymorphisms
glycine residue in the active site of a blowfly enzyme that Nonsynonymous a c
is ordinarily a carboxylesterase (splits a carboxyl ester,
Synonymous b d
R–COO–R, into an alcohol and a carboxylate). The muta-
tion causes complete loss of the carboxylesterase activity and Ratio a/b c/d