(Ricardo v. Lloyd) Morphology Methods Cell and Mo (B-Ok - Xyz)
(Ricardo v. Lloyd) Morphology Methods Cell and Mo (B-Ok - Xyz)
(Ricardo v. Lloyd) Morphology Methods Cell and Mo (B-Ok - Xyz)
METHODS
Cell and Molecular Biology
Techniques
Edited by
RICARDO V. LLOYD, MD, PhD
HUMANA PRESS
Morphology Methods
MORPHOLOGY
METHODS
Cell and Molecular
Biology Techniques
Edited by
Foreword by
Ronald A. DeLellis, MD
Weill Medical College of Cornell University and
New York Presbyterian Hospital, New York, NY
Humana P ress
Totowa, New Jersey
© 2001 Humana Press Inc.
999 Riverview Drive, Suite 208
Totowa, New Jersey 07512
www.humanapress.com
All rights reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or
by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise without written permission
from the Publisher.
The content and opinions expressed in this book are the sole work of the authors and editors, who have warranted due
diligence in the creation and issuance of their work. The publisher, editors, and authors are not responsible for errors
or omissions or for any consequences arising from the information or opinions presented in this book and make no
warranty, express or implied, with respect to its contents.
ANSI Z39.48-1984 (American Standards Institute) Permanence of Paper for Printed Library Materials.
Cover illustrations: Backround is figure 18 and inset is figure 5 from Chapter 4, “In Situ Detection of Infectious Agents,”
by Gary W. Procop and Randall Hayden.
For additional copies, pricing for bulk purchases, and/or information about other Humana titles, contact
Humana at the above address or at any of the following numbers: Tel: 973-256-1699; Fax: 973-256-8341;
E-mail: [email protected], or visit our Website: http://humanapress.com
The past several decades have witnessed an impressive array of conceptual and techno-
logical advances in the biomedical sciences. Much of the progress in this area has
developed directly as a result of new morphology-based methods that have permitted
the assessment of chemical, enzymatic, immunological, and molecular parameters at
the cellular and tissue levels. Additional novel approaches including laser capture
microdissection have also emerged for the acquisition of homogeneous cell popula-
tions for molecular analyses. These methodologies have literally reshaped the
approaches to fundamental biological questions and have also had a major impact in
the area of diagnostic pathology.
Much of the groundwork for the development of morphological methods was estab-
lished in the early part of the 19th century by Francois-Vincent Raspail, generally
acknowledged as the founder of the science of histochemistry. The earliest work in the
field was primarily in the hands of botanists and many of the approaches to the under-
standing of the chemical composition of cells and tissues involved techniques such as
microincineration, which destroyed structural integrity. The development of aniline
dyes in the early 20th century served as a major impetus to studies of the structural
rather than chemical composition of tissue. Later in the century, however, the focus
returned to the identification of chemical constituents in the context of intact cell and
tissue structure. Ultimately, it became possible to localize with great precision major
classes of nucleic acids, proteins, and individual amino acids, lipids, and carbohydrates
on the basis of specific chemical reactions that had been adapted for histological and
cytological preparations.
The development of the immunofluorescence technique for the localization of pneu-
mococcal antigens in 1942 by Albert Coons and his colleagues provided the foundation
for the subsequent development of enzyme-based immunohistochemical procedures.
The advantages of these methods were related to the facts that the reaction products
could be visualized directly in the light microscope and that the multistep staining
procedures possessed sufficient sensitivity to localize even the small amounts of antigen
that survived formalin fixation and paraffin embedding. The subsequent development
of increasingly more sensitive immunohistochemical staining sequences and the avail-
ability of increasing numbers of monoclonal antibodies together with the development
of antigen retrieval technology has made immunohistochemistry clearly one of the most
important research and diagnostic tools in the armamentarium of the investigator.
v
vi Foreward
vii
Contents
Foreword ................................................................................................................. v
Preface .................................................................................................................. vii
Contributors ........................................................................................................ xiii
Color Plates .......................................................................................................... xv
1 Introduction to Molecular Methods
Ricardo V. Lloyd .................................................................................................... 1
2 Laser Capture Microdissection: Principles and Applications
Ricardo V. Lloyd .................................................................................................. 11
PROTOCOL: H&E Staining 23
PROTOCOL: Operating the PIX Cell System 24
PROTOCOL: DNA Extraction from Collected Cells 24
PROTOCOL: RNA Extraction from Collected Cells 25
ix
x Contents
16 Application of Immunohistochemistry
in the Diagnosis of Lymphoid Lesions
Paul J. Kurtin .................................................................................................... 277
17 Applications of Immunohistochemistry
in the Diagnosis of Undifferentiated Tumors
Mark R. Wick and Lisa A. Cerilli .................................................................... 323
Contents xi
18 Applications of Immunohistochemistry
in the Diagnosis of Endocrine Lesions
Ricardo V. Lloyd ................................................................................................ 361
19 Ultrastructural Immunohistochemistry
Sergio Vidal, Eva Horvath, and Kalman Kovacs ........................................... 375
PROTOCOL: Ultrastructural Immunohistochemistry 399
xiii
xiv Contributors
xv
xvi Color Plates
Color Plate 5 Figs. 9 and 10, Chapter 10. • Fig. 9. Three-dimensional projection images of
ACTH and GH cells in normal rat pituitary. (A) ACTH cells are closely associated with the
microvessel networks. (B) ACTH signals show a granular pattern in the cytoplasm. (C) GH
cells are clearly visible three-dimensionally by CLSM reflectance mode. (D) GH signals
appear granular in the cytoplasm three-dimensionally by CLSM reflectance mode. See discus-
sion on page 174 and full caption on page 176. • Fig. 10. Three-dimensional reconstructed
images of bilaterally adrenalectomized and ACTH administered rat pituitary glands. See
discussion on pages 174–175 and full caption on page 177.
Color Plate 6 Figs. 1–5, Chapter 16. • Fig. 1. Immunoperoxidase stains for κ (A) and λ
(B) immunoglobulin light chains on frozen section of a case of B-cell small lymphocytic lym-
phoma. See discussion and full caption on pages 280, 281. • Fig. 2. Immunoperoxidase stains
for κ (A) and λ (B) immunoglobulin light chains on paraffin sections of a case of mantle cell
lymphoma. See discussion on page 280 and full caption on page 282.• Fig. 3. B-cell small
lymphocytic lymphoma. H&E section demonstrating the typical cytologic features (A). Frozen
section immunoperoxidase stains for CD20 (B), CD5 (C), CD3 (D), and CD23 (E). See discus-
sion on page 283 and full caption on page 284. • Fig. 4. Angioimmunoblastic T-cell lymphoma
stained in frozen sections for CD2 (A) and for CD7 (B). See discussion on page 286 and full
caption on page 287. • Fig. 5. Follicular lymphoma stained for bcl-2 paraffin sections. See
discussion on page 288 and full caption on page 289.
Color Plate 7 Figs. 6–13, Chapter 16. • Fig. 6. Mantle cell lymphoma stained for cyclin D1 in
paraffin sections. See discussion and full caption on pages 289, 290. • Fig. 7. CD30-positive ana-
plastic large cell lymphoma stained for p80 paraffin sections. See discussion and full caption on
pages 290, 291. • Fig. 8. Diffuse large B-cell lymphoma involving a lymph node. H&E section (A)
and immunoperoxidase stain (B). See discussion and full caption on pages 292, 293. • Fig. 9.
Angioimmunoblastic T-cell lymphoma involving a lymph node. H&E section (A) and immuno-
peroxidase stain (B) for CD3 performed on a paraffin section of the tumor. See discussion and full
caption on pages 294, 295. • Fig. 10. Immunoperoxidase stain for CD10 performed on a paraffin
section of a follicular lymphoma. See discussion and full caption on pages 296, 297.
• Fig. 11. Immunoperoxidase stain for CD23 performed on a paraffin section of an angio-
immunoblastic T-cell lymphoma. Same case as Fig. 9. See discussion and full caption on pages 298,
299. • Fig. 12. Immunoperoxidase stain performed on a paraffin section for the cytolytic granule
protein Tia-1 in a case of NK/T-cell lymphoma of the nasal type. See discussion and full caption on
pages 298, 299. • Fig. 13. Immunoperoxidase stain for CD30 performed on a CD30-positive ana-
plastic large cell lymphoma. See full caption and discussion on pages 300, 301.
Color Plate 8 Figs. 14–19, Chapter 16 • Fig. 14. Hodgkin's lymphoma, nodular sclerosis
type involving a lymph node. H&E section demonstrating the large neoplastic cells admixed
with small lymphocytes, eosinophils, plasma cells, and macrophages (A). Immunoperoxidase
stains for CD15 (B), CD30 (C), and fascin (D) performed on paraffin sections. See discussion
and full caption on pages 301–303. • Fig. 15. Hodgkin's lymphoma, nodular lymphocyte
predominance type, involving a lymph node. See discussion on pages 302, 304, and full caption
on page 305. • Fig. 16. Acute lymphoblastic leukemia, B-lymphocyte precursor type. See
discussion and full caption on pages 305, 306. • Fig. 17. Acute myelogenous leukemia, FAB
subtype M1. See discussion and full caption on pages 306, 307. • Fig. 18. Langerhans' cell
histiocytosis involving the parotid gland. H&E (A) exhibits the typical cytologic features of
this disorder. Immunoperoxidase stain for CD1a (B) performed in paraffin sections. See discus-
sion and full caption on pages 307, 308. • Fig. 19. Histiocytic sarcoma. Immunoperoxidase
stains for CD45 (A) and CD68 (B), in paraffin sections, and for CD13 on frozen sections of the
tumor (C). See discussion on page 308 and full caption on page 309.
1
Introduction to Molecular Methods
GENE STRUCTURE
The human genome contains 23 pairs of chromosomes with 3.5 × 109 nucleotide
base pairs (bp) in length. With the near completion of the sequencing of the human
gene, the best estimate of functional genes appears to be between 30,000 and 40,000
(1–3). Approximately 90% of the human DNA has no known coding function, but a
significant amount of DNA, especially in regions close to functional genes, has regula-
tory roles in cells. However, most of the functions of much of the DNA remain unknown.
Specific restriction enzymes derived from prokaryotes are used to cleave DNA at
specific sequences, which helps in the analysis of the large amount of genomic DNA.
For example, EcoRI, which is derived from Escherichia coli, cleaves DNA at the G↓
AATTC sequence. Because of the specificity of restriction endonucleases, the DNA
can be manipulated by genetic engineering, since DNA from any source cut by EcoRI
will produce identical ends that hybridize to any other source of DNA produced by
digestion with these restriction enzyme (4,5).
1
2 Lloyd
Fig. 1. Structure of molecules making up DNA and RNA. The bases (cytosine, guanine,
thymine, adenine, and unacil) are linked via hydrogen bonds. (A) The purines are linked by
three pairs of hydrogen bonds. (B) The pyrimidines are linked by two pairs of hydrogen bonds.
(C) The structure of unacil, present in RNA in place of thymine. (D) and (E) Structures of the
pentose sugars in DNA and RNA.
represent the diploid number, which represents the two homologous sites of each chro-
mosome.
The sex chromosomes are different in males and females; females have two X
chromosomes, and males have a Y and an X chromosome. The karyotype represents
the complete chromosomal complement of a cell. These karyotypes are usually analyzed
by cytogeneticists. Chromosomes are linked by a centromere with two sister chromatid
portions. Chromosome maps are based on the position of the banding pattern (easily
visualized after staining with specific dyes) and the position of the centromeres. Chromo-
somes are numbered as 1–22 plus the sex chromosomes. The short arm of the chromo-
somes is designated as a p and the long arm as q. The p and q arms are divided into various
sections. Thus a designation of 5q21 would represent the long arm of chromosome 5,
section 2, and band 1 (Fig. 4).
Fig. 3. Structure of the double helix, showing the three-dimensional relationship among the
bases, sugars, and phosphodiester moieties.
4 Lloyd
Fig. 4. Physical map of chromosome 7 after Giemsa staining and G-banding. Giemsa staining
highlights the dark and light bands. The centromere and sister chromatids are shown.
Fig. 6. The genetic code, with the amino acid symbols at the top and codons with the amino
acids they specify at the bottom.
by translation. The RNA in the nucleus is processed by excision of the intron regions
and addition of the 5′CAP site and a 3′ poly(A) tail, which increases the stability of
the mature mRNA. The introns that do not encode proteins are also removed from the
mRNA. The small nuclear RNA assists in the processing of mRNA. The mature mRNA
is transported to the cytoplasm, where it is then translated to protein. The nucleotide
sequence of the RNA (and parent DNA) determines the amino acid sequence of the
newly synthesized proteins.
The genetic code (Fig. 6) which consists of specific codons, specifies one or more
amino acids that are essential for protein synthesis. Because there are four bases in
RNA and 43 (or 64) possible base triplets or codons, the 20 naturally occurring amino
acids can be specified by different codons (Fig. 6). The methionine codon (AUG) is
the start codon for protein synthesis, and UAA, UAG, and UGA are the stop codons.
mRNA translation begins at the 5′ to 3′ end and is carried out on the ribosomes
with the assistance of transfer RNA (tRNA). Codons are recognized by tRNA that
have an anticodon triplet complement. The tRNA molecules associate with specific
amino acids determined by the anticodon sequence. The tRNA then interacts with the
complimentary codon on a message that is being translated, and the ribosome catalyzes
the addition of the amino acid to the protein during chain elongation.
HYBRIDIZATION
Hybridization or annealing involves pairing of complementary strands of nucleic
acids. Hybrids can be formed between complementary strands of nucleic acids (DNA
6 Lloyd
Fig. 7. Schematic of Southern hybridization. After the genomic DNA is digested by specific
restriction endonucleases, the DNA fragments are separated by gel electrophoresis and then
transferred to a nylon membrane and denatured with alkali. The single-stranded DNA is hybrid-
ized with a labeled probe. Hybrids are formed between the probe and complementary DNA
fragments and can be detected with a radioactive or nonradioactive reporter system. Et br,
ethidium bromide.
and RNA). These hybrid strands can be reversibly separated by heat treatment or
alkaline treatment to disrupt the hydrogen bonds between complementary bases.
Hybridization is one of the more powerful tools in molecular analysis. If a tag or
label is placed on one strand of the nucleic acid, it can be used as a probe in the
analysis of the nucleic acids. These types of analyses are designated as Southern
and Northern hybridization for DNA and RNA analyses, respectively, and in situ
hybridization for RNA or DNA tissue analysis (Figs. 7–9). In the typical Southern or
Northern hybridization, the nucleic acids are separated on a gel by electrophoresis.
Because nucleic acids are negatively charged, they move to the positive electrode.
After denaturation and/or transfer to a solid matrix, they can be hybridized with a
known probe to detect the DNA or RNA of interest.
The specificity of the hybridization can be readily controlled by the stringency of
the conditions, i.e., pH, temperature, and salt concentration. High stringency can lead
to detection of small differences or changes such as a change in a single base, as in a
genetic mutation.
Hybridization analyses are used in molecular diagnostics to detect gene mutations
or infectious agents in body fluids or tissues or to study gene expression in tumors or
other diseased tissues.
Fig. 9. In situ hybridization. The probe (DNA or RNA) is labeled with a radioactive or nonradio-
active detection system. The DNA probe is hybridized with the messenger RNA associated with
the rough endoplasmic reticulum. After hybridization, the target cells or tissues of interest can be
detected by autoradiography or with nonisotopic methods such as with a digoxigenin-labeled probe.
Advantages of in situ hybridization include more precise intracellular localization of the target and
direct visualization of the positive cell relative to the adjacent cells or tissues.
8 Lloyd
Fig. 10. Strategy for performing single-strand conformation polymorphism (SSCP) analysis
and sequence analysis from paraffin-embedded tissue sections. The polymerase chain reaction
(PCR) products can be visualized on a gel and then used for sequencing. SSCP also can be
performed with labeled probes and then analyzed on a polyacrylamide gel to detect differences
in migration of the single-stranded DNA compared with a known wild-type DNA. The gene
fragments with a shift in the migration pattern can be analyzed by sequencing to detect the
gene mutation or other genetic changes.
analyzed rapidly by microarray analysis (9–11). The use of fluorescent labeled cDNA
preparations in DNA microarrays allows one to analyze gene expression patterns on
a genomic scale. This procedure uses the basic principles of hybridization. A probe
hybridizes to multiple defined cDNAs or expressed sequence tags. The probe usually
consists of a complex mixture of cDNA fragments derived from mRNA. The probe
fragments are usually fluorescent labeled, and they hybridize to the targets on the glass
slide. A positive sign is detected by a laser system such as an argon-ion laser and then
quantified. Global gene expression can be readily detected with this technique.
Another approach is to use oligonucleotide arrays or DNA chips with synthetic
oligonucleotides, which represent thousand of gene sequences synthesized on the surface
of small areas of a glass slide.
DNA array technology for gene expression profiling has been applied to many areas
of biomedical sciences. A recent dramatic application is the report by Alizadeh et al.
(14), who used gene expression profiling in B-cell lymphomas to identify two molecu-
larly distinct forms of diffuse large B-cell lymphomas that had gene expression patterns
indicative of different stages of B-cell differentiation. Another study of gene-expression
profiles in hereditary breast cancers found functional differences between breast tumors
with BRCA1 mutations and those with BRCA2 mutations (15).
Tissue microarrays for high-throughput molecular profiling of normal and tumor
specimens have been adopted for rapid screening of many tissues (16,17). This new
approach to identify clinically relevant molecular changes uses microarrays from differ-
ent cases of the same type of tissue or tumor, which can range from 50 to a few hundred
sections of tissues on one slide, to check for protein or gene expression in these
microsamples. In the near future, a combination of tissue microarray technology with
pattern recognition software should greatly advance the discovery of new and potentially
useful markers with clinical and diagnostic applications.
REFERENCES
1. International Human Genome Sequencing Consortium (2001) Initial sequencing and analysis
of the human genome. Nature 409, 860–921.
2. Venter, J.C., Adams, M.D., Myers, E.W., et al. (2001) The sequence of the human genome.
Science 291, 1305–1351.
3. Claverie, J.-M. (2001) Gene number. What if there are only 30,000 human genes? Science
291, 1255–1257.
4. Watson, J. D., Hopkins, N. H., Roberts, J. W., Argetsinger Steitz, J., Weiner, A. M. (1987)
Molecular Biology of the Gene, 4th ed. Benjamin/Cumming, San Francisco 1987.
5. Lodish, H., Berk, A., Zipinsky, S. L., Matsudaira, P., Baltimore, D., Darnell, J. (2000)
Molecular Cell Biology. W. H. Freeman, New York.
6. Hensel, C. H., Xiang, R. H., Sakaguchi, A. Y., Naylor, S. L. (1991) Use of single strand
conformation polymorphism technique and PCR to detect p53 gene mutations in small cell
lung cancer. Oncogene 6, 1067–1071.
7. Orita, M., Iwahana, H., Kanazawa, H., Hayaski, K., Sekiya, T. (1989) Detection of polymor-
phisms of human DNA by gel electrophoresis as single-strand conformation polymorphisms.
Proc. Natl. Acad. Sci. USA 86, 2766–2770.
8. Orita, M., Suzuki, Y., Sekiya, T., Hayashi, K. (1989) Rapid and sensitive detection of
point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics
5, 874–879.
10 Lloyd
9. Schena, M., Shalon, D., Heller, R., Chai, A., Brown, P. O., Davis, R. W. (1996) Parallel
human genome analysis: microarray-based expression monitoring of 1000 genes. Proc.
Natl. Acad. Sci. USA 93, 10614–10619.
10. De Risi, J. L., Iyer, V. R., Brown, P. O. (1997) Exploring the metabolic and genetic control
of gene expression on a genomic scale. Science 278, 680–686.
11. Moch, H., Schraml, P., Bubendorf, L., et al. (1999) High-throughput tissue microarray
analysis to evaluate genes uncovered by cDNA microarray screening in renal cell carcinoma.
Am. J. Pathol. 154, 981–986.
12. Xu, J., Stolk, J. A., Zhang, X., et al. (2000) Identification of differentially expressed genes
in human prostate cancer using subtraction and microarray. Cancer Res. 60, 1677–1682.
13. Kurian, K. M., Watson, C. J., Wyllie, A. H. (1999) DNA chip technology. J. Pathol.
187, 267–271.
14. Alizadeh, A. A., Eisen, M. B., Davis, R. E., et al. (2000) Distinct types of diffuse large B-
cell lymphoma identified by gene expression profiling. Nature 403, 503–511.
15. Hedenfalk, I., Duggan, D., Chen, Y., et al. (2001) Gene-expression profiles in hereditary
breast cancer. N. Engl. J. Med. 344, 539–548.
16. Kononen, J., Bubendorf, L., Kallioniemi, A., et al. (1998) Tissue microarrays for high-
throughput molecular profiling of tumor specimens. Nature Med. 4, 844–847.
17. Moch, H., Schraml, P., Bubendorf, L., et al. (1999) High-throughput tissue microarray
analysis to evaluate genes uncovered by cDNA microarray screening in renal cell carcinoma.
Am. J. Pathol. 154, 981–986.
2
Laser Capture Microdissection
Principles and Applications
INTRODUCTION
The acquisition of homogeneous or pure cell populations for cell biologic and
molecular analyses has been a difficult challenge for many decades. A variety of
approaches have been tried, including macroscopic dissection of tissues from frozen
tissue blocks to increase the population of specific cell types (1,2). Some investigators
have used irradiation of manually ink-stained sections to destroy cells that were not
of interest (3,4). Microdissection with the aid of a microscope and needles has been
used by many other investigators (5,6). This latter approach can provide a great deal
of precision in collecting homogenous cells, but it is slow and labor-intensive and
requires a high degree of manual dexterity. Most of these approaches have not provided
the speed, precision, and efficiency needed for research or routine clinical molecular
diagnostic use.
The recent development of a laser capture microdissection (LCM) unit by the National
Cancer Institute (NCI) group led by Liotta and his colleagues (7,8) and by other groups,
mainly in Europe (9,10), has provided a rapid and efficient method to capture pure
cell populations for molecular and other studies.
Principles of LCM
Liotta’s group at the NCI first reported on the development of a rapid and reliable
method of obtaining homogenous population of cells from complex tissues (7,8). The
availability of commercial instruments resulted from a joint venture by the NCI and
Arcturus Engineering (Mountain View, CA).
With LMC, a complex section of tissues or heterogeneous cell populations on a
glass slide is placed on the stage of a specially designed microscope. After the areas
or cells of interest are selected, an ethylene vinyl acetate (EVA) transparent film is
apposed to the section, and an infrared laser beam is directed at the cell of interest.
When the focused laser beam coaxial with the microscope optics is activated, the EVA
film above the targeted area melts, surrounds, and holds the cells of interest, which
are kept in the film after removal from the glass slide. The film can be placed directly
11
12 Lloyd
Fig. 1. Laser capture microdissection (LCM) with the Pix Cell instrument. The laser beam
is focused on the cut section of cells on the glass slide. The transport arm carries the plastic
cap over to align it with the specimen and laser beam. The focused laser beam captures the
cells of interest, which adhere to the ethylene vinyl acetate (EVA) film that binds to the captured
target cells. After collecting the cells of interest, the transport arm carries the cells back to the side.
into DNA or RNA extraction buffers for nucleic acid extraction and analysis or into
other buffers for protein analysis. While in the buffer, the cellular material becomes
detached from the film and can be used for cellular or molecular analyses after extraction
(see Protocols).
Based on numerous reported studies in the literature and our own observations, the
LCM procedure does not lead to significant alterations in the morphologic or apparent
molecular features of the cells of interest.
Technical Aspects
The prototypic instrument used for LCM is the Pix Cell developed by Arcturus
Engineering and the NCI group (Figs. 1 and 2). The instrument consists of a laser
optics deck and illumination tower with a halogen bulb housing, a microscope, and a
slide stage with a vacuum chuck. A joystick with XY control and video camera
attachment are attached to the microscope. The XY control is for fine positioning of
the sample. The laser can be controlled by amplitude and pulse width adjustments
using digital controls on the front panel of the electronic box. When the laser fires, an
emission indicator is lit, and the power supply emits a beeping sound. Typical operating
parameters for the laser for a 30-µm spot is an amplitude of 30 mW and a pulse width
of 5 ms; for a 60-µm spot, an amplitude of 50 mW and a pulse width of 5 ms is used.
When the operator is ready to capture the cells of interest, a plastic cap is transported
with the film carrier and placed on the desired position of the tissue or cells. The laser
Laser Capture Microdissection 13
Fig. 2. The cells on the EVA film are placed in the microfuge tube and immersed in the
appropriate buffer for DNA, RNA, or protein extraction and analysis.
is then activated, and the EVA film binds to the captured target on the tissue. The laser
can be activated as often as is needed to capture the desired numbers of cells. Multiple
clusters of homogenous cells can be accumulated into the same polymer EVA film,
and individual single laser shots can be used to procure specific cell clusters or individual
cells (Fig. 3). In addition, multiple shots can be combined to procure complex but
homogenous tissue structures. Up to 3000 shots can be captured on one transfer film
cap, which may include up to 6000 cells. Since each shot takes less than 1 s to perform,
a large number of cells can be captured in a relatively short period.
Technical Variables
Technical variables with LCM include the types of analyses that will be done with
the tissues relative to fixation and processing, use of frozen versus paraffin tissue
sections, and combination of LCM with other techniques such as immunohistochemistry
(Immuno-LCM).
Tissues fixed in buffered formalin and embedded in paraffin can be used for routine
analysis of DNA and in some cases mRNA as well. Tissues are routinely stained with
hematoxylin and eosin for optimum visualization of cellular details during LCM. These
stains do not affect the integrity of the DNA or RNA. For optimum RNA analysis,
frozen tissue sections are preferred. Surprisingly, the nucleic acids are not adversely
affected by exposing the tissues to xylene for a short period. RNAse-free conditions
are important for obtaining high-quality RNA samples.
The section for frozen tissue should be cut at 10 µm or less for optimum transfer,
because with thicker sections it is more difficult to visualize single cells. Sections have
to be completely dry and not cover-slipped for effective LCM transfer so the final
xylene rinse facilitates efficient LCM transfer. After the frozen sections are cut, the
slides can be used immediately or stained at −80°C until use. We have observed that
14 Lloyd
Fig. 3. LCM capture of cells from normal breast tissue. (A) Groups of normal epithelial
cells were selected for LCM (arrowheads). (B) Arrowheads indicate the space from which the
cells were captured. (C) Captured cells on the cap with the transfer EVA film. (Reproduced
with permission from ref. 13.)
starting with fresh alcohols and xylene for each set of experiments avoids many of the
technical problems associated with difficult transfers during LCM.
The procedure of combining LCM with immunophenotyping for RNA analysis
(immuno-LCM) was first reported from the NCI (11) and subsequently by our groups
and others for analyzing single cells (12) and tissues (13–15) (Fig. 4). The critical
requirements include using RNAse-free condition and using RNAse inhibitors during
the immunostaining procedures. A rapid immunostaining procedure minimizes the time
of tissue exposure to RNAse.
Various studies have shown that precipitating fixatives such as ethanol and acetone
produced better quality reverse transcriptase-polymerase chain reaction (RT-PCR) prod-
uct amplification compared with crosslinking fixatives such as formaldehyde and glutar-
aldehyde (11,12,14). However, precipitating fixatives are less effective in some immuno-
staining procedures compared with crosslinking fixatives (12).
used a low-power laser to create a gap between the cells of interest and adjacent cells;
with an increase in the laser power, they were able to “catapult” the microdissected
cells into the cap of a microfuge tube for further analysis without requiring any direct
contact with the cells of interest. Using this approach, they could amplify simple cells
from archival tissues in the analysis of Ki-ras mutations (16).
Fink and colleagues (17) used an ultraviolet light laser to ablate cells that were not
of interest and subsequently collected the cells of interest with a stent needle under
the control of a micromanipulator (17), which is a modification of the original method of
Shibata et al. (3,4). Other investigators have utilized this approach to study microsatellite
instability in breast cancer at the single cell level (18). A combination of ultraviolet
microbeam microdissection with laser pressure catapulting for the isolation of single
chromosomes for development of chromosome-specific paint probes has been reported
(19). Becker et al. (20) used a combination of ultraviolet laser microbeams with microdis-
section and molecular characterization of individual cells to detect a novel mutation
in the E cadherin gene in single tumor cells (20). A recent study used the PALM Laser-
Microbeam System, which allows the contact-free isolation of single cells or groups
of cells using the laser pressure catapulting technique and real-time PCR to study the
Her-2/Neu gene and topoisomerase IIα gene in breast cancer specimens (21).
cellular details at higher magnifications. This can be partially overcome with a diffusion
filter on the instrument or by using a drop of xylene on the tissues, which provides
wetting and refractive-index matching; the xylene usually evaporates rapidly before
microdissection (22). Another disadvantage for investigators with a tight budget is the
costs of the instrument. Finally, although LCM is faster than manual methods, a great
deal of time is still required to collect cells for an experiment compared with biochemical
analysis, especially when thousands of cells are collected (23).
In a recent study of synovial sarcoma, Kasai et al. (29) used LCM and RT-PCR
with paraffin wax-embedded tissues to show that the SYT-SSX fusion transcript was
present in both the spindle cell and epithelial areas of biphasic synovial sarcomas. This
study also reinforced the idea that RT-PCR can be performed on RNA obtained from
formalin-fixed, paraffin-embedded tissue sections. In a related study, Shibutani et al.
18 Lloyd
Fig. 6. RT-PCR and Southern hybridization examination of prolactin (PRL), prolactin receptor
(PRL-R), and GAPDH RNA obtained by LCM showing representative examples of the analyses.
Lanes 1, 3, and 5 are normal breast tissues. Lanes 2, 4, and 6 are invasive ductal carcinoma.
Lane 7 represents a positive pituitary tissue control, and lane 8 is a negative control without
reverse transcriptase. Lanes 1–3 and 4–6 represent matching normal and tumor tissues, respec-
tively. M, molecular size marker. The top part of each panel represents the RT-PCR results on
the gel after ethidium bromide staining. The bottom part of each panel is the Southern hybridiza-
tion with the internal probes. (Reproduced with permission from ref. 13.)
In a study of papillary and follicular thyroid carcinoma Gillespie et al. (32) used
LCM and loss of heterozygosity (LOH) analysis of highly polymorphic chromosomes
to show that LOH was more common in follicular than papillary carcinoma. In a related
study on practical clinical applications of LCM, a group of investigators analyzed a
primary duodenal carcinoid as well as tumors in the scalp and cervical lymph nodes
from the same patient by LCM with 22 markers. The 3p12 marker showed loss in all
three tumors, and the 3pl4.2 marker showed an identical shift in the three tumors
indicative of a common microsatellite alteration and supporting the notion that the
shared molecular abnormalities indicated a common clonal origin with the duodenal
carcinoid as the primary tumor (33).
overexpression of the wnt and notch growth and differentiation regulatory systems by
these squamous cell carcinomas (41).
Proteomic Analysis
The use of LCM for protein analysis had been reported in the original descriptions
of LCM (7) in which gelatin zymography was used to study gelatinase A (MMP-2) in
frozen sections of prostate cancer. In a more recent study, Banks et al. (42) used LCM
and two-dimensional electrophoresis to examine protein profiles of selected tissue areas.
In the cervix they were able to show enrichment of some proteins compared with the
whole tissue using these combined techniques. Emmert-Buck et al. (43) used normal
and neoplastic esophageal squamous mucosa as a model to study proteins captured by
LCM followed by two-dimensional polyacrylamide gel electrophoresis (2D-PAGE).
They were able to use 50,000 cells to resolve 675 distinct proteins or isoforms with
molecular weights between 10 and 200 kDa and isoelectric points between pH 3 and
10. Although 98% of the proteins in normal and tumor samples were identical, 17 of
the tumors showed tumor-specific alterations. Seven of these proteins were present
only in normal tissues, and 10 of the proteins were present only in the tumors. Cytokeratin
I and annexin I were identified as two of the altered proteins.
In a study of LCM-captured prostate tissue analyzed with a quantitative chemilumi-
nescent assay, Simone et al. (44) studied prostate-specific antigen (PSA) distribution
in normal prostatic intraepithelial neoplasia and carcinoma and found a range of 2 ×
104 to 6.3 × 106 PSA molecules per microdissected tissue cells. These observations
were concordant with the immunohistochemical staining intensity of tissue sections
from these different tissue areas (44). Ornstein et al. (45) used LCM and PAGE (1D
and 2D) to analyze PSA α1-antichymotrypsin (ACT) by Western blotting studies and
showed a 30-kDa band that was the expected molecular weight of unbound PSA.
Binding studies revealed that PSA recovered from LCM-procured cells had the full
ability to bind ACT as in the normal prostate. Electrophoretic studies also showed that
the PSA/ACT complex was stable, indicating that the complex was similar in normal
and tumorous tissues.
FUTURE DIRECTIONS
Future advances in LCM will include technologic advances in instrumentation and
application of LCM to solve difficult and/or challenging problems in cell biology and
biomedical sciences. The further development of laser fields small enough to capture
single cells or specific parts of cells such as the nucleus are currently under way.
Automation of the LCM system for automatic performance of cell capture is also being
developed. A major challenge for the technology is to develop methods to capture live
cells that can be used for cell culture and other in vitro studies with living cells.
Major advances can be anticipated in combining LCM with other techniques such
as flow cytometry, FISH and comparative genomic hybridization. Significant advances
can be anticipated in gene expression studies with microarrays to profile specific human
diseases starting from clearly defined histopathologic cell populations. The technique
of “molecular fingerprinting” of entire organs such as the prostate has been initiated
by one group (38). Cole et al. (38) have proposed a model for integrating in three
Laser Capture Microdissection 21
dimensions the data obtained by LCM and microarrays for the in vivo molecular
anatomy analysis of normal and neoplastic cells. These types of approaches will rely
heavily on LCM for further advances in understanding the molecular aspects of diseases
as they relate to tissue structure and function.
REFERENCES
1. Fearon, E. R., Hamilton, S. R., and Vogelstein, B. (1987) Clonal analysis of human colorectal
tumors. Science 238, 193–197.
2. Shibata, D., Hawes, D., Li, Z. H., Hernandez, A. M., Spruck, C. H., and Nichols, P. W.
(1992) Specific genetic analysis of microscopic tissue after selective ultraviolet radiation
fractionation and the polymerase chain reaction. Am. J. Pathol. 141, 539–543.
3. Shibata, D. (1993) Selective ultraviolet radiation fractionation and polymerase chain reaction
analysis of genetic alterations. Am. J. Pathol. 143, 1523–1526.
4. Emmert-Buck, M. R., Roth, M. J., Zhuang, Z., et al. (1994) Increased gelatinase A (MMP-
2) and cathepsin B activity in invasive tumor regions of human colon cancer samples. Am.
J. Pathol. 145, 1285–1290.
5. Zhuang, Z., Bertheau, P., Emmert-Buck, M. R., et al. (1995) A microdissection technique
for archival DNA analysis of specific cell populations in lesions <1 mm in size. Am. J.
Pathol. 146, 620–625.
6. Noguchi, S., Motomura, K., Inaji, H., Imaoka, S., and Koyama, H. (1994) Clonal analysis
of predominantly intraductal carcinoma and precancerous lesions of the breast by means
of polymerase chain reaction. Cancer Res. 54, 1849–1853.
7. Emmert-Buck, M. R., Bonner, R. F., Smith, P. D., et al. (1996) Laser capture microdissection.
Science 274, 998–1001.
8. Bonner, R. F., Emmert-Buck, M., Cole, K. et al. (1997) Laser capture microdissection:
molecular analysis of tissue. Science 278, 1481–1483.
9. Schütze, K., Lahr, G. (1998) Identification of expressed genes by laser mediated manipula-
tion of single cells. Nat. Biotechnol. 16, 737–742.
10. Fink, L., Seeger, W., Ermert, L., et al. (1998) Real-time quantitative RT-PCR after laser-
assisted cell picking. Nat. Med. 4, 1329–1333.
11. Fend, F., Emmert-Buck, M. R., Chuaqui, R., et al. (1999) Immuno-LCM: laser capture
microdissection of immunostained frozen sections for mRNA analysis. Am. J. Pathol.
154, 61–66.
12. Jin, L., Thompson, C. A., Qian, X., Kuecker, S. J., Kulig, E., Lloyd, R. V. (1999) Analysis
of anterior pituitary hormone mRNA expression in immunophenotypically characterized
single cells after laser capture microdissection. Lab. Invest. 79, 511–512.
13. Kuecker, S. J., Jin, L., Kulig, E., Ondraogo, G. L., Roche, P. C., Lloyd, R. V. (1999)
Analysis of PRL, PRL-R, TGFβ1 and TGFβ-RII gene expression in normal and neoplastic
breast tissues after laser capture microdissection. Appl. Immunohistochem. Mol. Morphol.
7, 193–200.
14. Goldsworthy, S. M., Stockton, P. S., Trempus, C. S., Foley, J. F., Maronpot, R. R. (1999)
Effects of fixation on RNA extraction and amplification from laser capture microdissected
tissue. Mol. Carcinog. 25, 86–91.
15. Fink, L., Kinfe, T., Stein, M. M., et al. (2000) Immunostaining and laser-assisted cell
picking for mRNA analysis. Lab. Invest. 80, 327–333.
16. Schutze, K., Lahr G. (1998) Identification of expressed genes by laser-mediated manipulation
of single cells. Nat. Biotech. 16, 737–742.
17. Fink, L., Seeger, W., Ermert, L., et al. (1998) Real-time quantitative RT-PCR after laser-
assisted cell picking. Nat. Med. 4, 1329–1333.
22 Lloyd
18. Dietmaier, W., Hartmann, A., Wallinger, S., et al. (1999) Multiple mutation analyses in
single tumor cells with improved whole genome amplification. Am. J. Pathol. 154, 83–95.
19. Schermellch, L., Tyhalhammer, S., Iteckl, W., et al. (1999) Laser microdissection and laser
pressure catapulting for the generation of chromosome-specific paint probes. Biotechniques
27, 362–367.
20. Becker, I., Becker, K. F., Rohrl, M. H., Minkus, G., Schütze, K., Höfler, H. (1996) Single-
cell mutation analysis of tumors from stained histologic slides. Lab. Invest. 75, 801–807.
21. Lehmann, V., Glöckner, S., Kleeberger, W., von Wasielewski, H. F., Kreipe, H. (2000)
Detection of gene amplification in archival breast cancer specimens by laser-assisted micro-
dissection and quantitative real-time polymerase chain reaction. Am. J. Pathol. 156, 1855–
1864.
22. Simone, N. L., Bonner, R. F., Gillespie, J. W., Emmert-Buck, M. R., Liotta, L. A. (1998)
Laser-capture microdissection: opening the microscopic frontier to molecular analysis.
Trends Genet. 14, 272–276.
23. Curran, S., McKay, J. A., McLeod, H. L., Murray, G. I. (2000) Laser capture microscopy.
Mol. Pathol. 53, 64–68.
24. Suarez-Quian, C. A., Goldstein, S. R., Pohida, T., et al. (1999) Laser capture microdissection
of single cells from complex tissues. Biotechniques 26, 328–335.
25. Glasow, A., Haidan, A., Hilbers, U., et al. (1998) Expression of Ob receptor in normal
human adrenals: differential regulation of adrenocortical and adrenomedullary function by
leptin. J. Clin. Endocrinol. Metab. 83, 4459–4466.
26. Glasow, A., Haidan, P., Gillespie, J., Kelly, P. A., Chrousos, G. P., Bornstein, S. R. (1998)
Differential expression of prolactin receptor (PRLR) in normal and tumorous adrenal tissues:
separation of cellular endocrine compartments by laser capture microdissection (LCM).
Endocr. Res. 24, 857–862.
27. Fend, F., Quintanilla-Martinez, L., Kumar, S., et al. (1999) Composite low grade B-cell
lymphomas with two immunophenotypically distinct cell populations are true biclonal
lymphomas. A molecular analysis using laser capture microdissection. Am. J. Pathol.
154, 1857–1866.
28. Jin, L., Tsumanuma, I., Ruebel, K. H., Bayliss, J. M., Lloyd, R. V. (2001) Analysis
of homogeneous populations of anterior pituitary folliculostellate cells by laser capture
microdissection and RT-PCR. Endocrinology, in press.
29. Kasai, T., Shimajiri, S., Hashimoto, H. (2000) Detection of SYT-SSX fusion transcripts in
both epithelial and spindle cell areas of biphasic synovial sarcoma using laser capture
microdissection. Mol. Pathol. 53, 107–110.
30. Shibutani, M., Uneyama, C., Miyazaki, K., Toyoda, K., Hirose, M. (2000) Methacarn
fixation: a novel tool for analysis of gene expressions in paraffin-embedded tissue specimens.
Lab. Invest. 80, 199–208.
31. Kohda, Y., Murakami, H., Moe, O. W., Star, R. A. (2000) Analysis of segmental renal
gene expression by laser capture microdissection. Kidney Int. 57, 321–331.
32. Gillespie, J. W., Nasir, A., Kaiser, H. E. (2000) Loss of heterozygosity in papillary and
follicular thyroid carcinoma: a mini review. In Vivo 14, 139–140.
33. Milchgrub, S., Wistuba, II, Kim, B. K, et al. (2000) Molecular identification of metastatic
cancer to the skin using laser capture microdissection: a case report. Cancer 88, 749–754.
34. Jones, C., Foschini, M. P., Chaggar, R., et al. (2000) Comparative genomic hybridization
analysis of myoepithelial carcinoma of the breast. Lab. Invest. 80, 831–836.
35. Shen, C. Y., Yu, J. C., Lo, Y. L., et al. (2000) Genome-wide search for loss of heterozygosity
using laser capture microdissected tissue of breast carcinoma: an implication for mutator
phenotype and breast cancer pathogenesis. Cancer Res. 60, 3884–3892.
36. DiFrancesco, L. M., Murthy, S. K., Luider, J., Demetrick, D. J. (2000) Laser capture
Laser Capture Microdissection 23
PROTOCOLS
H&E Staining
For paraffin sections, start with step 1. For frozen sections that have been previously
fixed, start with step 5.
1. Xylenes to deparaffinize the slides 2 × 5 min each.
2. 100% ethanol 30 s.
3. 100% ethanol 30 s.
4. 95% ethanol 30 s.
5. Purified water rinse.
6. Mayer’s hematoxylin 30 s.
7. Purified water rinse.
8. Eosin Y 30 s.
9. 95% ethanol 30 s.
10. 95% ethanol 30 s.
11. 100% ethanol 30 s.
12. Xylene 2 × 3 min.
13. Shake off excess, wipe the slide carefully, and air-dry for at least 2 min.
24 Protocols
NOTE: Chain reagents (xylenes and alcohols) each day that LCM staining is done
and filter hematoxylin before use.
1. If entire instrument is shut off, turn on power strip located behind the computer monitor.
Also, turn on power for the TV monitor.
2. On computer screen, click mouse on “shortcut to Arc 100.”
3. When instrument serial # is given, click “continue.”
4. Select name or enter new name and click “acquire data.”
5. Highlight study or enter new study and click “select.”
6. Enter slide #, spot size, cap lot, and thickness (usually 10 µm, but this may vary depending
on slides cut), and click “continue.”
7. Enter laser power of approx. 45 and pulse of approx. 55 (to start) length needed (see
procedures for RNA and DNA extractions for this information).
8. Place slide on microscope, and, using TV monitor, find an appropriate starting area that
will be easy to find after finishing the microdissection.
9. Click on “before” to obtain a picture of the area selected.
10. Insert a row of arcturus caps with transfer film into slot on the right side of the microscope.
11. If the laser power chosen is below 60, place the optic beam adjust piece without the filter
in the indentation on the end of the placement “arm.” If laser power chosen is above 60,
place the beam adjust piece with the filter on the arm.
12. Using the placement arm, pick up a cap, move the arm all the way over to the left (this
will put it directly above the slide), and gently release it so that the cap slowly drops onto
the slide.
13. Pick up the white cord with the red button on the end and press the button to get a laser
pulse. A dark circle will appear around the area if it “melted.” (However, just because an
area melts does not necessarily mean that it transferred to the cap.)
14. Use the joystick located to the bottom left of the microscope to move the slide around
and get pulses in different areas.
15. After finishing the microdissection, use the arm to pick up the cap very gently (so as not
to pick up any other tissue), move it over to the right, and place it on a sterile 0.5-mL
microcentrifuge tube.
16. Find the original starting area and clock on “after” to get an image of the completed
microdissection. Then look at the slide under a microscope to make sure that most of the
areas transferred.
17. Click “done” and then “exit” or “continue.”
1. Strain slides with hematoxylin and eosin according to the written procedure for staining.
2. Aliquot 100 µl 0.05 M Trizma buffer, pH 8.3, and 4 µL proteinase K (10 mg/mL) into sterile
0.5-mL Eppendorf microfuge tubes (see Digestion Buffer below).
3. Use LCM to capture the desired number of cells from the slide.
4. Place the film cap containing the captured cells on the top of the microfuge tube and use
the black cap tool to snap the cap in. (It is critical to place the bottom of the film cap just
DNA Extraction 25
inside the lip of the microfuge tube, leaving about a 1-cm gap. Using the black cap tool
ensures the correct spacing needed to prevent leakage during incubation.)
5. Invert the tubes and incubate in a 55°C water bath for 48 h.
6. Remove tubes from water bath and spin down in microcentrifuge. Discard the film cap and
transfer the solution to a new sterile 0.5-mL tube.
7. Boil samples for 8 min at 95°C in a hot block and place on ice. Store DNA at 4°C until
ready to use.
Digestion Buffer
1. 3.03 g Trizma base.
2. 500 mL water.
3. Adjust pH to 8.3.
INTRODUCTION
Recent developments in molecular biology are rapidly expanding the technologic
advances used to study disease processes. In situ hybridization (ISH) represents a unique
technique in which molecular biologic and histochemical techniques are combined to
study gene expression in tissue sections and cytologic preparations. DNA and RNA
can be readily localized in specific cells with this method. The method involves a
hybridization reaction between a labeled nucleotide probe and complementary target
DNA or RNA. Those hybrids can be detected either by autoradiographic emulsion for
radioactively labeled probes or by histochemical chromogen development for nonisoto-
pically labeled probes. ISH localizes gene sequences in situ and visualizes the product
of gene expression while preserving cell integrity within the heterogeneous tissue,
permitting anatomically meaningful interpretations.
ISH analysis of nucleic acids was first described in 1969 (1,2). Since the mid-1980s,
ISH techniques have developed rapidly and have become important tools in basic
scientific research and clinical diagnosis (3–8). The introduction of nonradioactive
probe labeling and detection systems in the late 1970s has made ISH analysis feasible
in the diagnostic pathology laboratory as a molecular diagnostic tool and in microbiology
for the tissue localization of infectious agents. Other recent developments in the applica-
tions of ISH techniques involve localization of multiple target DNA and RNA sequences
and increasing the assay sensitivity by in situ amplification for both the target nucleic
acid and signal detection system.
This chapter discusses the basic principles and approaches of the ISH technique,
new advances in the ISH method, and some application of ISH methods for the study
of specific diseases.
27
28 Jin, Qian, and Lloyd
phosphodiester linkage. Each of the four nucleotides has a base, including adenine (A),
cytosine (C), guanine (G), and thymidine (T). A is bound to T with two hydrogen
bonds, and C is bound to G with three hydrogen bonds. RNA is a single-stranded
nucleotide with the base uridine (U) substituted for T. ISH is the process by which
there is specific annealing of a labeled nucleic acid probe to a target complementary
sequence in fixed tissue or cells. This is followed by visualization of the hybridized
signals with isotopic or colorimetric detection methods. One can have DNA-DNA, DNA-
RNA, and RNA-RNA ISH, depending on various types of probes and targets concerned.
The main advantages of the ISH method are in its specificity for individual cells in a
heterogeneous tissue or cell population, as well as in its sensitivity in detecting low-copy
gene expression in cells or chromosomal gene mapping. ISH consists of multiple steps
including probe preparation and labeling, tissue preparation, hybridization, and signal
detection. Each of these steps must be systematically optimized for each probe and tissue
preparation. When applied appropriately, the methods can also be semiquantitative.
of the amplified probe is precisely defined by the 5′ ends of the PCR primers. All three
DNA labeling methods allow great flexibility with regard to length of the labeled
fragments. In the nick translation reaction, changing the DNAse concentration alters
the fragment length. In the random primed labeling reaction, changing the primer
concentration affects fragment length. In the PCR procedure, the sequence of the PCR
primers controls fragment length. The probes prepared by these three labeling methods
require denaturation to produce single-stranded DNA before hybridization.
Antisense RNA Riboprobes
Antisense RNA riboprobes are generated by in vitro transcription from a linearized
DNA template, incorporating labeled nucleotides. A promoter for RNA polymerases must
be present on the vector DNA containing the template. SP6, T3, or T7 RNA polymerase
is commonly used to synthesize RNA complementary to the DNA substrate (11). The
synthesized transcripts are an exact copy of the sequence (about 300–800 bp) from the
promoter site to the restriction site used for linearization. The size of the probe can be
adjusted by selection of the linearizing restriction enzyme so that the antisense and sense
probes both have the same length. RNA riboprobes are single-stranded and are susceptible
to degradation by RNAses. However, cRNA probes have the advantage that RNA-RNA
hybrids are more stable than DNA-DNA or DNA-RNA hybrids.
Oligonucleotide (Oligo) Probes
Oligonucleotide (oligo) probes consisting of 30–50 bases can be generated with an
automated DNA synthesizer. The oligo-probes probably penetrate cells more readily
and can produce excellent hybridization signals. With the increasing numbers of cloned
genes, oligonucleotide probes can be generated from published cDNA sequences or
gene banks and synthesized rapidly and inexpensively. Labeling oligo-probes for ISH is
commonly done with 3′ end labeling, or “tailing” with relatively few labeled nucleotides
incorporated into each probe. The oligo-probes are relatively less sensitive than longer
cDNA or cRNA probes (4,12), so they are generally considered the most suitable for
detection of relatively abundant gene expression such as hormone mRNAs (13–15).
This disadvantage can be overcome by using a “cocktail” of multiple oligonucleotide
probes that are complementary to different regions of the target molecules (8,16).
Careful selection of oligo-probes with low or no homology with other nucleotide
sequences is most important to ascertain the specificity of ISH (17). By using 3′ end
labeling with deoxydinucleotides, about 10–20 labeled nucleotides are incorporated
into end of probe sequence with extension after incubation at 37°C for 30–45 min.
Oligo-probes may also be generated in the same way as standard riboprobes using a
short DNA template or by combining single-strand oligonucleotides with a bacterio-
phage promoter.
Tissue Preparation
Tissue processing including storage and fixation should be optimized to detect
intracellular nucleic acids. ISH has been applied to cells (smears, cytospins, or cell
pellets), tissue sections (frozen, paraffin, semithin, and ultrathin plastic section), or
whole-mount embryos. Nucleic acids are better preserved in frozen tissue than in
paraffin-embedded tissues. Intact cells are ideal for ISH, as they have less damaged
30 Jin, Qian, and Lloyd
nucleotide sequences. Both fresh frozen tissues (stored at −70°C) and formalin-fixed
archival tissues can be used for ISH after storage for several years. Ideal fixation for
ISH should preserve both RNA/DNA and tissue morphology but still allow penetration
of probes. Crosslinking fixatives such as 4% paraformaldehyde, 4% formaldehyde, and
1% glutaraldehyde, are most commonly used. Generally, fixation for 20 min to 1 h is
sufficient for frozen sections on glass slides. After a brief wash in 2× standard saline
citrate (SSC) to remove residual fixative and dehydration of the slides through a graded
ethanol series, the slides can be stored at −70°C until hybridized. The slides used for
ISH should be pretreated with a suitable coating solution, such as 3-aminoprpyltrime-
thoxysilane or poly-L-lysine.
A series of pretreatment steps before hybridization increases the efficiency of hybrid-
ization and reduce nonspecific background staining. Some steps are routinely used in
most ISH protocols.
Protease treatment (e.g., proteinase K) is considered one of the most important steps
to increase the accessibility of the target nucleic acid, especially for paraffin sections
with nonisotopic probes. The concentrations of proteinase K (1–2 µg/mL for frozen
sections and 20–50 µg/mL for paraffin sections) and the length of treatment (5–30
min) depend on tissue type and length of fixation. Prolonged incubation results in
overdigestion, resulting in loss of signal and morphologic integrity (18). A brief post-
fixation in paraformaldehyde is commonly recommended.
Pretreatment of slides with microwaving to increase ISH sensitivity, especially for
old paraffin-embedded tissues, has been reported (18). For microwaving, sections are
immersed in 10 mM citrate buffer, pH 6.0, and microwaved at full power in an 800-W
microwave for 5–10-min periods. It should be noted that if microwaving treatment is
added, the length of proteinase K treatment should be reduced.
Acetylation of sections using 0.25% acetic anhydride/0.1 M triethanolamine can
reduce charged probe binding to tissues and nonspecific binding. Some investigators
have noted that acetic anhydride prevents the nonspecific binding of unrelated digoxi-
genin-labeled probes to neuroendocrine cells (15). Adding fresh dithiothreitol (DTT)
to the hybridization solution containing 35S-labeled probes can protect the sulfur from
oxidation. The background artifacts with 35S-labeled probes can be dramatically reduced
by increasing the DTT concentration (19). Incubation of sections with prehybridization
buffer, normally containing 50% formamide, is generally considered useful for mainte-
nance of the hybridization stringency, even though some investigators prefer to perform
ISH in a formamide-free condition (20,21). The presence of endogenous biotin or
alkaline phosphatase should be anticipated in some tissues when nonisotopic probes
are used. Endogenous alkaline phosphatase can be inhibited by treatment of sections
with 0.2 N HCL and levamisole, and endogenous biotin can be inhibited by biotin-
blocking agents.
Hybridization
Hybridization between the labeled probe and target DNA or RNA is defined by
hydrogen bonding and the hydrophobic interactions in equilibrium. Annealing and
separation of the two hybrid strands depend on various factors, including temperature,
salt concentrations, pH, the nature of the probes and target molecules, and the composi-
tion of the hybridization and washing solution. The “melting” temperature (Tm) of
Detection of DNA and RNA by ISH 31
hybrids is the point at which 50% of the double-stranded nucleic acid chains are
separated. The optimal temperature for hybridization is 15–25°C below the Tm. There
are various formulas for calculating Tm, depending on probe length and type of hybrids
(5). For a DNA-DNA hybrid with probes longer than 22 bases, the following formula
can be used:
Tm = 81.5 + 16.5 log(Na) + 0.41 (%GC) − 0.62 (% formamide)
− 500/length of base pairs of probe
where Na = sodium concentration and GC = guanine and cytosine content of the probe.
RNA-RNA hybrids are generally 10–15°C more stable than DNA-DNA or DNA-
RNA and therefore require more stringent conditions for hybridization and posthybrid-
ization washing.
Hybridization buffer contain reagents to maximize nucleic acid duplex and inhibit
nonspecific binding of probe to tissues. Generally, 20–30 µL of hybridization buffer
containing labeled probe is enough for each slide. The concentration of probe must be
optimized for each probe and for each tissue. Hybridization is done on a cover-slipped
slide in a humidified chamber in an oven at 42–45°C for oligo-probes or 50–55°C for
cDNA or RNA riboprobes.
After an overnight incubation, careful removal of the cover slips using 2X SSC is
followed by posthybridization washing with 2X, 1X, and 0.5X SSC for 10–30 min
each. The final washing can be done at 42–45°C (for oligo-probes) or 50–55°C (for
DNA or RNA probes) to increase the stringency. If riboprobes are used, treatment with
ribonuclease A (RNAse) after hybridization (generally 0.1–1 ng/mL for paraffin sections
or 0.1–1 µg/mL for frozen sections) will reduce nonspecific background signal, since the
enzyme digests single-stranded, but not double-stranded, RNA hybrids. After additional
washing in phosphate-buffered saline (PBS) for 15–30 min, the slides are ready for
signal detection.
Evaluation
It is important to use various controls to verify the specificity of the ISH signal. A
variety of controls can be used: 1) pretreatment of tissues with RNAse or DNAse,
depending on the target being tested; 2) omission of the specific probes in hybridization
Detection of DNA and RNA by ISH 33
Recent Advances
A variety of new ISH techniques have been developed over the past few years. These
advances can be divided two categories: 1) combined use of various ISH approaches for
simultaneous detection of multiple target nucleic acids at the light microscope (LM)
and electron microscope (EM) levels; and 2) increasing the sensitivity of ISH signals
by in situ amplification for target DNA or RNA or by signal detecting systems.
Multiple-Target Localization
Double-labeling ISH methods involve the simultaneous detection of multiple-target
nucleic acids such as two mRNAs, DNA, and mRNA, or nucleic acids and proteins
for combination ISH with IHC techniques (28–32). In performing double ISH, the two
probes may be hybridized simultaneously or sequentially. For combined ISH/IHC,
ISH is generally performed before IHC because it reduces the chances of RNAse
contamination. A variety of combinations of ISH methods have been reported, as
discussed below.
A combination of nonradioactive and radioactive ISH methods is mainly used for
detection of two mRNAs in the same tissue sections (19,29,30). After histochemical
34 Jin, Qian, and Lloyd
detection for nonisotopic signals, the slides are subject to an autoradiographic approach
for radioactive signals. Some investigators have reported that the use of Ilford emulsion,
rather than Kodak NTB2 emulsion, can reduce silver grain background by chemographic
artifact when digoxigenin and 35S-labeled oligonucleotide probes for double ISH detec-
tion are used simultaneously (31).
Several investigators have successfully used a combination of two nonisotopic labeled
probes, mainly biotin and digoxigenin, conjugated to different enzymes (AP or HPO)
or fluorescence followed by respective detective systems for a simultaneous localization
of multiple mRNA and genomic DNA in the same tissue section, or even in the same
cells (3,32–34). Using probes labeled with different fluorochrome molecules and with
different excitation and emission characteristics allows simultaneous analysis of differ-
ent probe signals on chromosomal and cytogenic preparations (5,33). An alternate
method is the combination of nonisotopic ISH with fluorescence (F)ISH: the two signals
can be visualized by using brightfield and darkfield microscopy (32).
Ultrastructural ISH is used mainly for investigations attempting to correlate molecular
function and ultrastructural morphology. Nonradioactive ISH methods with colloidal
gold are most commonly used for pre- and/or postembedding EM techniques (35–37).
Colloidal gold with different nonoverlapping particle sizes (5–30 nm) greatly facilitates
multiple labeling for ISH analysis at the EM level. Bienz and Egger (35) reported the
simultaneous localization of P1 and P2 virus genome regions in poliovirus-infected cells
by double-EM ISH: a digoxenin labeled P1 probe and a biotin-labeled P2 probe were
detected through simultaneous direct immunodetection with 10-nm gold-labeled anti-
digoxigenin and 6-nm gold-labeled antibiotin antibodies. A combination of ISH with
the terminal deoxynucleotidyl transferase dUTP nick end-labeling technique on pre-
embedded free-floating sections has been used to analyze specific gene expression and
apoptosis at both LM and EM levels (38).
In Situ Amplification
An example of in situ target amplification is in situ PCR, a relatively new molecular
technique that combines the high sensitivity of PCR with anatomic localization by ISH
(39–41). ISH localizes gene sequences at the cellular level, but it generally requires
at least 10–20 or more copies of the mRNA of interest in a cell to make the signals
visible. A combination of PCR with ISH can amplify DNA or RNA sequences inside
single cells and increase the copy numbers to levels readily detectable by ISH methods.
There are two main approaches: direct and indirect. In the direct approach a labeled
nucleotide is incorporated into the PCR products (42); in the indirect method ISH is
performed after in situ amplification using a labeled oligonucleotide probe.
Over the past few years, in situ PCR has been mainly applied to detect DNA
sequences that are not easily detected by conventional ISH. In situ PCR to detect low
copy numbers of viral genes, especially the human immunodeficiency virus and hepatitis
C, has led to significant discoveries about viral diseases (43). In situ RT-PCR is the
process of detecting mRNA in which mRNA is reverse transcribed to single strands
of DNA followed by PCR amplification. Application of in situ RT-PCR to detect gene
expression is still limited primarily to cell preparations and frozen sections. Only a few
applications of this procedure in paraffin sections have been reported (44). Successful
amplification of mRNA by in situ RT-PCR includes hormone, receptor, and oncogenes
Detection of DNA and RNA by ISH 35
Infectious Diseases
In the field of infectious diseases, ISH has been used mainly to detect viral infections,
such as human papillomavirus (HPV) and Epstein-Barr virus (EBV) by the presence
of small nuclear RNAs (EBERs). (See also Chapter 4.)
Viral Infections
Viral infections have been more widely investigated, and the technique has proved
useful both in the identification of specific viral infection and in defining the extent of
36 Jin, Qian, and Lloyd
systemic infection. The combined use of broad-spectrum probes and those identifying
specific virus subtypes has confirmed a strong positive association between HPV and
high-grade cervical intraepithelial neoplasia or progressive disease, with evidence that
types 16 and 18 are more likely to be associated with malignant progression (54,55).
The test has given some insight into the epidemiology, in that similar patterns of
infections were identified in two cohorts separated by 25 years (56). ISH may be a
useful adjunct in the investigation of HPV infection of the male genitalia (57). However,
the role played by HPV in other lesions, including Bowen’s disease and squamous cell
papilloma of the bronchus and larynx, is unclear. EBV has been implicated in the
pathogenesis of a wide range of human lymphoid and epithelial tumors (58) (Figs. 1A,
B). It is difficult to detect viral DNA by ISH in latent infection because of the low
copy number per cell. However, EBERs encoded by the virus are highly expressed
and detectable by nonradioactive ISH. EBV has been detected in some cases of
Hodgkin’s disease, in a variety of lymphomas, including T-cell lymphomas, and in
oral hairy leukoplakia (59–61). The presence of EBV in undifferentiated nasopharyngeal
carcinoma has been studied extensively (62). ISH is useful in confirmation of infectious
mononucleosis in atypical cases and of renal disease associated with EBV patients who
have a negative serology (63,64).
Hybridization to viral DNA and mRNAs in the liver has aided our understanding
of the complexity of hepatitis B infection (65). However, these studies have provided
conflicting data about the correlation between viral replication and disease activity.
More recently, ISH has been used to detect hepatitis C virus (HCV) RNA. However,
negative results were found in a significant proportion of HCV-seropositive cases (66).
This could be due to the low level of replication of the virus, and amplification techniques
may increase the sensitivity. Most studies applied these techniques to formalin-fixed,
paraffin-embedded tissue due to its availability and the good preservation of morphology.
Very few studies have applied this technique to frozen tissue. However, most groups
showed that HCV-positive cells were usually isolated, with occasional clustering, and
that usually <20% of cells were positive (67,68). In a few reports, the proportion of
HCV RNA-positive cells was found to be very high (69). Most studies showed the
localization of HCV in hepatocytes, but some studies showed ISH signals in mononu-
clear cells, bile duct epithelium, and sinusoidal cells (70,71). HCV may be directly
involved in the biliary epithelial damage seen in a proportion of patients with chronic
HCV infection (70,71). The level of HCV positivity appeared to correlate with serum
aminotransferase, suggesting that HCV might cause damage to the liver cells directly
in the absence of overt morphologic changes.
ISH has been used for the identification of cytomegalovirus (CMV) in viral encephali-
tis associated with the acquired immunodeficiency syndrome (AIDS) and chronic
encephalitis associated with AIDS or epilepsy. Pulmonary involvement has been
detected in cytologic specimens and the extent of systemic disease documented when
the presentation was that of isolated oophoritis (72). CMV infection has been detected
in gastrointestinal biopsies in cardiac transplant recipients (73). However, with CMV,
there may not be a significantly greater sensitivity with ISH than with histology and
immunohistochemistry.
Detection of DNA and RNA by ISH 37
Gene Expression
ISH methods are powerful tools for the analysis of gene expression and regulation
in normal and pathologic tissues (89). A major advantage of ISH is its ability to
localize mRNA at the cellular level in heterogeneous tissues, thus expanding the
results of other molecular techniques, such as Northern blot hybridization, for
specific gene analysis.
ISH has been used to detect mRNA as a marker of gene expression when levels of
protein storage are low, for example, to confirm an endocrine tumor as the source of
excess hormone production (13,90). ISH methods have been widely used for the study of
mRNA-encoding oncogenes, growth factors and their receptors, hormones and hormone
receptors, cytokines, structural proteins, enzymes, collagenase, and others.
There are many practical applications of ISH methods in tumor pathology (Fig.
1C–F). The correlation of oncogene expression with prognosis is being investigated
in neuroblastomas and epithelial neoplasms such as colon, lung, prostate, and breast
carcinomas. Detection of genes encoding cell structural proteins, including tumor-
associated markers, represents a potential area of application of ISH methods in patho-
logic diagnosis. For example, nonisotopic ISH methods for localization of immunoglob-
ulin light chain in mRNAs in hyperplastic and neoplastic lymphoproliferative disorders
(16) (Fig. 1C), albumin mRNA to distinguish between hepatocellular and metastatic
carcinomas in the liver (Fig. 1D), and chromogranin/secretogranin mRNAs in the
classification of neuroendocrine tumors (8) (Fig. 1E) are used in some diagnostic
pathology laboratories. ISH methods used to identify cells or tumors on the basis of
their specific mRNA content are different from IHC, which is dependent on the protein
content of cells. Thus, ISH identifies the gene products from de novo synthesis, rather
than nonspecific uptake of proteins by cells, which may result in false-positive immuno-
staining results. ISH analysis has also been used extensively in studies of endocrine
cells and tumors (Fig. 1F). For example, a significant number of small cell lung
carcinomas with few secretory granules are commonly negative for chromogranin
proteins, but the mRNAs may be detected by ISH methods (8,14). ISH studies of gene
expression in endocrine tumors (including parathyroid hormones and calcitonin gene-
40 Jin, Qian, and Lloyd
related peptide) have contributed to our understanding of the biology and pathophysiol-
ogy of various endocrine disorders.
REFERENCES
1. Gall, J. G. and Pardue, M. L. (1969) Formation and detection of RNA-DNA hybrid molecules
in cytological preparation. Proc. Natl. Acad. Sci. USA 63, 378–383.
2. John, H. A., Birnstiel, M. L., and Jones, K. W. (1969) RNA-DNA hybrids at the cytological
level. Nature 223, 582–587.
3. Egger, D., Troxler, M., and Bienz, K. (1994) Light and electron microscopic in situ hybridiza-
tion: non-radioactive labeling and detection, double hybridization, and combined hybridiza-
tion-immunocytochemistry. J. Histochem. Cytochem. 42, 815–822.
4. Wilcox J. N. (1993) Fundamental principles of in situ hybridization. J. Histochem. Cytochem.
41, 1725–1733.
5. Wilkinson, D. G., ed. (1993) In Situ Hybridization: A Practical Approach. IRL Press-
Oxford University Press, New York.
6. Panoskaltsis-Mortari, A. and Bucy, R. P. (1995) In situ hybridization with digoxigenin-
labeled RNA probes. Facts and artifacts. Biotechniques 18, 300–307.
7. Poulsom, R., Longcroft, J. M., Jeffery, R. E., Rogers, L. A., and Steel, J. H. (1998) A
robust method for isotopic riboprobe in situ hybridization to localize mRNAs in routine
pathology specimens. Eur. J. Histochem. 42, 121–132.
8. Lloyd, R. V. and Jin, L. (1995) In situ hybridization analysis of chromogranin A and B
mRNAs in neuroendocrine tumors with digoxigenin-labeled oligonucleotide probe cocktails.
Diagn. Mol. Pathol. 4, 143–151.
9. Temsamani, J. and Agrawal, S. (1996) Enzymatic labeling of nucleic acids (Review). Mol.
Biotechnol. 5, 223–232.
10. Mertz, L. M. and Rashtchian, A. (1994) Nucleotide imbalance and polymerase chain reac-
tion: effects on DNA amplification and synthesis of high specific activity radiolabeled DNA
probes. Anal. Biochem. 221, 160–165.
11. Witkiewicz, H., Bolander, M. E., and Edwards, D. R. (1993) Improved design of riboprobes
from pBluescript and related vectors for in situ hybridization. Biotechniques 14, 458–463.
12. Komminoth, P., Merk, F. B., Leav, I., Wolfe, H. J., and Roth, J. (1992) Comparison of
35S and digoxigenin-labeled RNA and oligonucleotide probes for in situ hybridization.
Expression of mRNA of the seminal vesicle secretion protein II and androgen receptor
genes in the rat prostate. Histochemistry 98, 217–228.
13. Lloyd, R. V., Cano, M., Chandler, W. F., Barkan, A. L., Horvath, E., and Kovacs, K.
(1989) Human growth hormone and prolactin secreting pituitary adenomas analyzed by in
situ hybridization. Am. J. Pathol. 134, 605–613.
14. Lloyd, R. V., Jin, L., Kulig, E., and Fields, K. (1992) Molecular approaches for the analysis
of chromogranins and secretogranins. Diagn. Mol. Pathol. 1, 2–15.
15. Pagani, A., Cerrato, and Bussolati, G. (1993) Nonspecific in situ hybridization reaction in
neuroendocrine cells and tumors of the gastrointestinal tract using oligonucleotide probes.
Diagn. Mol. Pathol. 2, 125–130.
16. Weiss, L. M., Movahed, L. A., Chen, Y. Y., et al. (1990) Detection of immunoglobulin
light-chain mRNA in lymphoid tissues using a practical in situ hybridization method. Am.
J. Pathol. 137, 979–988.
17. Stahl, W. L., Eakin, T. J., and Baskin D. G. (1993) Selection of oligonucleotide probes for
detection of mRNA isoforms. J. Histochem. Cytochem. 41, 1735–1740.
18. Oliver, K. R., Heavens, R. P., and Sirinathsinghji, D. J. S. (1997) Quantitative comparison
of pretreatment regimens used to sensitize in situ hybridization using oligonucleotide probes
on paraffin-embedded brain tissue. J. Histochem. Cytochem. 45, 1707–1713.
Detection of DNA and RNA by ISH 41
19. Miller, M. A., Kolb, P. E., and Raskind, M. A. (1993) A method for simultaneous detection
of multiple mRNAs using digoxigenin and radioisotopic cRNA probes. J. Histochem.
Cytochem. 41, 1741–1750.
20. Farquharson, M. A., Harvie, R., Kennedy, A., and McNicol, A. M. (1992) Detection of
mRNA by in situ hybridization and in Northern blot analysis using oligodeoxynucleotide
probes labeled alkaline phosphatase. J. Clin. Pathol. 45, 999–1002.
21. Thomas, G. A., Davies, H. G., and Williams, E. D. (1993) Demonstration of mRNA using
digoxigenin labeled oligonucleotide probes for in situ hybridization in formamide free
conditions. J. Clin. Pathol. 46, 171–174.
22. Steel, J. H., Jeffery, R. E., Longcroft, J. M., Rogers, L. A., and Poulsom, R. (1998)
Comparison of isotopic and non-isotopic labelling for in situ hybridisation of various mRNA
targets with cRNA probes. Eur. J. Histochem. 42, 143–150.
23. Singer, R. H. and Ward, D. C. (1982) Actin gene expression visualized in chicken muscle
tissue culture by using in situ hybridization with biotinated nucleotide analogue. Proc. Natl.
Acad. Sci. USA 79, 7331–7335.
24. Szakacs, J. G. and Livingston, S. K. (1994) mRNA in-situ hybridization using biotinylated
oligonucleotide probes: implications for the diagnostic laboratory. Ann. Clin. Lab. Sci.
24, 324–338.
25. Yoshii, A., Koji, T., Ohsawa, N., and Nakome, P. (1995) In situ hybridization of ribosomal
RNAs is a reliable reference for hybridizable RNA in tissue sections. J. Histochem. Cyto-
chem. 43, 321–327.
26. Tongiorgi, E., Righi, M., and Cattaneo, A. (1998) A non-radioactive in situ hybridization
method that does not require RNAse-free conditions. J. Neurosci. Methods. 85, 129–139.
27. Guiot, Y. and Rahier, J. (1997) Validation of nonradioactive in situ hybridization as a
quantitative approach of messenger ribonucleic acid variations. A comparison with northern
blot. Diagn. Mol. Pathol. 6, 261–266.
28. Kriegsmann, J., Keyszer, G., Geiler, T., Gay, R. E., and Gay, S. (1994) A new double
labeling technique for combined in situ hybridization and immunohistochemical analysis.
Lab. Invest. 71, 911–917.
29. Trembleau, A., Roche, D., and Calas, A. (1993) Combination of non-radioactive and
radioactive in situ hybridization with immunohistochemistry: a new method allowing the
simultaneous detection of two mRNAs and one antigen in the same brain tissue section.
J. Histochem. Cytochem. 41, 489–498.
30. Dagerlind, A., Friberg, K., Bean, A. J., and Hokfelt, T. (1992) Sensitive detection using
unfixed tissues: combined radioactive and nonradioactive in situ hybridization histochemis-
try. Histochemistry 98, 39–43.
31. Young, W. S. III and Hsu, A. C. (1991) Observations on the simultaneous use of digoxigenin
and radiolabeled oligodeoxyribonucleotide probes for hybridization histochemistry. Neuro-
peptides 18, 75.
32. Van Wijk, I. J., Van Vugt, J. M. G., Konst, A. A. M., Mulders, M. A. M., Nieuwint,
A. W. M., and Oudejans, C. B. M. (1995) Multiparameter in situ hybridization of trophoblast
cells in mixed cell populations by combined DNA and RNA in situ hybridization. J.
Histochem. Cytochem. 43, 709–714.
33. Gray, J. W., Pinkel, D., and Brown, J. M. (1994) Fluorescence in situ hybridization in
cancer and radiation biology (Review). Radiat. Res. 137, 275–289.
34. Speel, E. J. M. (1999) Detection and amplification systems for sensitive, multiple-target
DNA and RNA in situ hybridization: looking inside cells with a spectrum of colors.
Histochem. Cell. Biol. 112, 89–113.
35. Bienz, K. and Egger, D. (1995) Immunocytochemistry and in situ hybridization in the
electron microscope: combined application in the study of virus-infected cells. Histochemis-
try 103, 328–338.
42 Jin, Qian, and Lloyd
36. Lloyd, R. V., Jin, L., and Song, J. (1990) Ultrastructural localization of prolactin and
chromogranin B messenger ribonucleic acids with biotinylated oligonucleotide probes in
cultured pituitary cells. Lab. Invest. 63, 413–419.
37. Sibon, O. C. M., Cremers, F. F. M, Boonstra, H. J., and Verkeij, A. J. (1995) Localization
of nuclear RNA by pre- and post-embedding in situ hybridization using different gold
probes. Histochem. J. 27, 35–45.
38. Bessert, D. A. and Skoff, R. P. (1999) High-resolution in situ hybridization and TUNEL
staining with free-floating brain sections. J. Histochem. Cytochem. 47, 693–701.
39. Long, A. A. (1998) In situ polymerase chain reaction: foundation of the technology and
today’s options. Eur. J. Histochem. 42, 101–109.
40. Uhlmann, V., Silva, I., Luttich, K., Picton, S., and O’Leary, J. J. (1998) In cell amplification.
J. Clin. Pathol. Mol. Pathol. 51, 119–130.
41. Komminoth, P. and Long, A. A. (1995) In situ polymerase chain reaction and its applications
to the study of endocrine diseases. Endocr. Pathol. 6, 167–171.
42. Nuovo, G. J., Gallery, F., MaConnell, P., Becker, J., and Bloch, W. (1991) An improved
technique for the in situ detection of DNA after polymerase chain reaction amplification.
Am. J. Pathol. 139, 1239–1244.
43. Nuovo, G. J., Lidonnici, K., MacConnel, P., and Lane, B. (1993) Intracellular localization
of polymerase chain reaction (PCR)-amplified hepatitis C cDNA. Am. J. Surg. Pathol.
17, 683–690.
44. Martinez, A., Miller, M. J., Quinn, K., Unsworth, E. J., Ebina, M., and Cuttitta, F. (1995)
Non-radioactive localization of nucleic acids by direct in situ PCR and in situ RT-PCR in
paraffin-embedded sections. J. Histochem. Cytochem. 43, 739–747.
45. Jin, L., Qian, X., and Lloyd, R. V. (1995) Comparison of mRNA expression detected by
in situ polymerase chain reaction and in situ hybridization in endocrine cells. Cell Vision
2, 314–321.
46. Chen, R. H. and Fuggle, S. V. (1993) In situ cDNA polymerase chain reaction. A novel
technique for detecting mRNA expression. Am. J. Pathol. 143, 1527–1534.
47. Patel, V. G., Shum-Siu, A., Heniford, B. W., Wieman, T. J., and Hendler, F. J. (1994)
Detection of epidermal growth factor receptor mRNA in tissue sections from biopsy speci-
mens using in situ polymerase chain reaction. Am. J. Pathol. 144, 7–14.
48. Bagasra, O., Sheshamma, T., Hansen, J., Bobroski, L., Saikumari, P., and Pomerantz, R. J.
(1995) Application of in situ PCR methods in molecular biology. II. Special applications
in electron microscopy. Cytogenics and immunohistochemistry. Cell Vision 2, 61–70.
49. Hofler, H., Putz, B., Mueller, J. D., Neubert, W., Sutter, G., and Gais, P. (1995) In situ
amplification of measles virus RNA by the self-sustained sequence replication reaction.
Lab. Invest. 73, 577–585.
50. Wilkens, L., Komminoth, P., Nasarek, A., Wasielewski, R. van, and Werner, M. (1997)
Rapid detection of karyotype changes in interphase bone marrow cells by oligonucleotide
primed in situ labeling (PRINS). J. Pathol. 181, 368–373.
51. Speel, E. J. M., Hopman, A. H. N., and Komminoth, P. (1999) Amplification methods to
increase the sensitivity of in situ hybridization: play CARD(S). J. Histochem. Cytochem.
47, 281–288.
52. Merz, H., Malisius, R., Mannweiler, S, et al. (1995) Immunomax. A maximized immunohis-
tochemical method for the retrieval and enhancement of hidden antigens. Lab. Invest.
73, 149–156.
53. Wiedorn, K. H., Kuhl, H., Galle, J., Caselitz, J., and Vollmer, E. (1999) Comparison of in
situ hybridization, direct and indirect in situ PCR as well as tyramide signal amplification
for the detection of HPV. Histochem. Cell Biol. 111, 89–95.
54. Pollanen, R., Vuopala, S., and Lehto, V. P. (1993) Detection of human papillomavirus
Detection of DNA and RNA by ISH 43
73. Murray, J., Farquharson, M., Wheatley, D. J., and McPhaden, A. R. (1996) Cytomegalovirus
detection by in situ hybridization in upper gastrointestinal biopsies from cardiac transplant
patients. J. Pathol. 179 (suppl.), 34A.
74. Gaffey, M. J., Benezra, J. M., and Weiss, L. M. (1991) Herpes simplex lymphadenitis. Am.
J. Clin. Pathol. 95, 709–714.
75. Remadi, S., Finci, V., Ismail, A., Zacharie, S., and Vassilakos, P. (1995) Herpetic endometri-
tis after pregnancy. Pathol. Res. Pract. 191, 31–34.
76. Sumiyoshi, Y., Akashi, K., and Kikuuchi, M. (1994) Detection of human herpes virus-6
(HHV-6) in the skin of a patient with primary HHV-6 infection and erythroderma. J. Clin.
Pathol. 47, 762–763.
77. Boshoff, C., Schulz, T. F., Kennedy, M. M., et al. (1995) Kaposi’s sarcoma-associated
herpesvirus infects endothelial and spindle cells. Nat. Med. 1, 1274–1278.
78. Reiss, E., Tanaka, K., Bruker, G., et al. (1998) Molecular diagnosis and epidemiology of
fungal infections [Review]. Med. Mycol. 36, (suppl. 1) 249–57.
79. Perez-Jaffe, L. A., Lanza, D. C., Loevner, L. A., Kennedy, D. W., and Montone, K. T.
(1997) In situ hybridization for Aspergillus and Penicillium in allergic fungal sinusitis: a
rapid means of speciating fungal pathogens in tissues. Laryngoscope 107, 233–240.
80. Hanazawa, R., Murayama, S. Y., and Yamaguchi, H. (2000) In-situ detection of Aspergillus
fumigatus. J. Med. Microbiol. 49, 285–290.
81. Park, C. S., Kim, J., and Montone, K. T. (1997) Detection of Aspergillus ribosomal RNA
using biotinylated oligonucleotide probes. Diagn. Mol. Pathol. 6, 255–260.
82. Hayashi, Y., Watanabe, J., Nakata, K., Fukayama, M., and Ikeda, H. (1990) A novel
diagnostic method of Pneumocystis carinii. In situ hybridization of ribosomal ribonucleic
acid with biotinylated oligonucleotide probes. Lab. Invest. 63, 576–580.
83. Kim, J., Yu, J. R., Hong, S. T., and Park, C. S. (1996) Detection of Pneumocystis carinii by
in situ hybridization in the lungs of immunosuppressed rats. Korean J. Parasitol. 34, 177–84.
84. Krimmer, V., Merkert, H., von Eiff, C., et al. (1999) Detection of Staphylococcus aureus
and Staphylococcus epidermidis in clinical samples by 16S rRNA-directed in situ hybridiza-
tion. J. Clin. Microbiol. 37, 2667–2673.
85. Karttunen, T. J., Genta, R. M., Yoffe, B., Hachem, C. Y., Graham, D. Y., and el-Zaatari,
F. A. (1996) Detection of Helicobacter pylori in paraffin-embedded gastric biopsy specimens
by in situ hybridization. Am. J. Clin. Pathol. 106, 305–311.
86. Hayden, R. T., Qian, X., Roberts, G. D., and Lloyd, R. V. (2001) In situ hybridization for
the identification of yeast-like organisms in tissue section. Diagn. Molec. Pathol. 10, 15–23.
87. Assy, N. and Minuk, G. Y. (1997) Liver regeneration: methods for monitoring and their
applications [Review]. J. Hepatol. 26, 945–952.
88. Rautiainen, E., Haapasalo, H., Sallinen, P., Rantala, I., Helen, P., and Helin, H. (1998)
Histone mRNA in-situ hybridization in astrocytomas: a comparison with PCNA, MIB-1
and mitoses in paraffin-embedded material. Histopathology 32, 43–50.
89. DeLellis, R. A. (1994) In situ hybridization techniques for the analysis of gene expression:
applications in tumor pathology. Hum. Pathol. 25, 580–585.
90. Lloyd, R. V., Jin, L., and Chandler, W. F. (1996) In situ hybridization studies in human
pituitaries, Pituitary Adenomas (Landolt, A. M., Vance, M. L., and Reilly, P. L., eds.),
Churchill Livingstone, New York, pp. 47–58.
In Situ Hybridization 45
PROTOCOLS
In Situ Hybridization
Day 2
1. Dip slides in 2× SSC to remove cover slips for 5 min.
2. Wash in PBS for 3 min.
3. Incubate in RNAse A solution (0.1–1 ng/mL in PBS) at 37°C for 30 min.
4. Wash in PBS for 5 min.
5. Wash in 2× SSC at 50°C for 10 min.
6. Wash in 1× SSC at 50°C for 10 min.
7. Wash in 0.5× SSC at 50°C for 10 min.
8. Incubate in buffer A for 1 min.
9. Incubate in buffer A with 1% sheep serum and 0.03% Triton X-100 for 20 min.
10. Add anti-Dig antibody (1/200 in above buffer, from Boehringer Mannheim) and incubate
for 3 h.
11. Wash in buffer A for 5 min, ×2.
12. Wash in buffer B for 5 min, ×2.
13. Develop in nitroblue tetrazolium/bromochloroindolyl phosplate (NBT/BCIP) in the dark
for 30–60 min, monitoring under the microscope.
14. Stop the reaction in buffer B and wash briefly in H2O.
15. Counterstain nucleus with fast red, dehydrate, and coverslip.
46 Protocols
Day 2
1. Dip slides in 2× SSC to remove cover slips.
2. Wash in 2× SSC at 42–44°C for 10 min.
3. Wash in 1× SSC at 42–44°C for 10 min.
4. Wash in 0.5 × SSC at 42–44°C for 10 min.
5. Incubate in buffer A for 1 min.
6. Incubate in buffer A with 1% sheep serum and 0.03% Triton X-100 for 20 min.
7. Add anti-Dig antibody (1/200 in above buffer, from Boehringer Mannheim) and incubate
for 3 h.
8. Wash in buffer A (1 M NaCl, 0.1 M Tris-HCl, 2 mM MgCl2, pH 7.5) for 5 min, ×2.
9. Wash in buffer B (0.1 M NaCl, 0.1 M Tris base, 5 mM MgCl2, pH 9.5) for 5 min, ×2.
10. Develop in NBT/BCIP in the dark for 30–60 min, monitoring under the microscope.
11. Stop the reaction in buffer B and wash briefly in H2O.
12. Counterstain nucleus with fast red, dehydrate, and cover-slip.
INTRODUCTION
In situ hybridization (ISH) has a wide variety of uses in diagnostic pathology (1–4).
ISH for the detection and differentiation of microorganisms is a science that in many
respects is still in its infancy. Early in the development of these technologies, viruses
in tissue sections were the natural targets for identification. These methods have proved
useful for such identification, especially for viruses that are difficult to culture or
when infection was not suspected and tissues were not submitted for culture (5–7).
Unfortunately, many of the early ISH methods were complicated, required a high level
of technical expertise, and may have been viewed as a detection method for the anatomic
pathology laboratories alone (i.e., direct competition for clinical virology). These factors
may have discouraged some from using this technology. Furthermore, microorganisms,
which may be detected by ISH, have been successfully found through alternate techno-
logies. The greatest competing technologies for ISH are immunohistochemistry (IHC),
solution hybridization, and nucleic acid amplification (NAA). In many instances, these
alternate technologies have proved equally effective, possibly more sensitive, and
sometimes more user friendly and/or cost effective than ISH (8). ISH, however, remains
a useful and important tool for the identification of microorganisms in histologic sections,
and its uses are expanding to include the direct identification of microorganisms in
clinical specimens and rapid culture confirmation.
There has been a dramatic evolution of the molecular biology industry over the past
10 years, which has provided a variety of tools to the molecular pathologist and
microbiologist. The relative ease of nucleic acid synthesis to produce high-quality ISH
probes has made this method of microbial detection highly versatile. Nucleic acid
probes directed against ribosomal RNA targets can provide a definitive genus and
species identification for a variety of organisms in a fraction of the time of culture.
These probes may be labeled with fluorescent reporter molecules, or secondary signal
amplification methods may be employed to enhance assay sensitivity. Incredibly, with
ISH and advanced signal amplification methods, it is possible to detect genes present
in only a single copy (9–11). This method holds great potential for detection of the
genes responsible for antimicrobial resistance in bacteria and fungi and could conceiv-
ably replace conventional susceptibility testing.
47
48 Procop and Hayden
In Situ Detection of Infectious Agents 49
With the passage of time, the concept of ISH has become less foreign to the clinical
microbiologist, probably because of the influx of other diagnostic molecular methods
such as polymerase chain reaction (PCR) with Southern blot or solution hybridization.
ISH applications for microorganisms have also expanded from viruses to include bacteria,
fungi, and parasites. These methods have recently been shown to be particularly useful
for the detection of fastidious microbes that are difficult to culture and for rapid culture
confirmation. (5–7,12,13). For example, ISH methods have been used to identify Legio-
nella and Histoplasma definitely in histologic sections and to identify bacteria in blood
culture rapidly, without the delay of subculture and subsequent biochemical testing
(12,13) (Figs. 12, 13, 18 and 19). Furthermore, some have employed ISH directly on the
patient specimen on a rapid basis and substantially improved the time to identification for
Aspergillus species when only hyphae are seen (14). The maintenance of tissue morphol-
ogy afforded by ISH may provide an additional benefit, by helping to discriminate contam-
inants from true infectious pathogens. This could be of particular import in the cases
of organisms such as Aspergillus that may be either saprophytic or pathogenic. These
expanded applications of ISH, as well as the standardization of more user-friendly meth-
ods and the use of automation, are likely to establish ISH as a rapid and effective means
of organism identification in the modern microbiology and pathology laboratory (3,15).
VIRUS DETECTION
Early in the development ISH techniques, viruses were the most common infectious
agents detected. The presence of these viruses was often suspected because of the cyto-
pathic effect seen in histologic sections or cytologic preparations. The most common
human pathogenic viruses that ISH is used to detect are cytomegalovirus (CMV), herpes
simplex virus (HSV), human papillomavirus (HPV), Epstein-Barr virus (EBV), adenovi-
rus, and the JC polyoma virus (Figs. 1–9; see Protocols 8, 16–28). In addition to the
detection of viruses and the confirmation of cause of a cytopathic effect, ISH has also
been used differentiate the high- and low-risk subtypes of HPV, to type herpes simplex
viruses, and to assess the effects of treatment (29–33,17). The presence of a virus in tissue
culture is often confirmed by using specific antibody stains, but ISH may also be used
(34). ISH has also helped to demonstrate associations and possible associations between
particular viruses and specific neoplastic conditions (35–37). HPV, which is associated
with dysplasia/squamous cell carcinoma of the uterine cervix, EBV, which is associated
with posttransplant lymphoproliferative disorder and other malignancies, and human her-
pes virus 8 (HHV8), which is associated with Kaposi’s sarcoma, have all been successfully
detected in the respective neoplastic conditions by ISH (35,36,38–48).
This technology has come into direct competition with immunohistochemistry and
other molecular techniques. In many institutions, immunohistochemistry, which has a
wide variety of applications in anatomic pathology, has replaced ISH or prevented the
institution of ISH for the detection of these agents. This may, in part, be secondary to
efforts to reduce duplication of like technologies and streamline laboratory operations
to lower operating costs. Additionally, the high-quality commercial availability of
antibodies directed against many of these viruses, the use of semiautomated techniques,
and the rapid turn-around time for IHC stains also contributed to the use of IHC rather
than ISH for detection of these agents (26–28,48). More recently, ISH has had to
compete with the accurate, rapid and cost-effective solution hybridization technology
(i.e., Hybrid-Capture, Digene, Beltsville, MD) for the detection and differentiation
of HPV subtypes. Although these issues may have limited the widespread routine use
of ISH techniques, the recent availability of semiautomated platforms and the increasing
number of commercially available probes promise to put these methods within the
reach of many more laboratories.
Interestingly, the use of ISH for the detection of viruses in histologic sections may
come full circle. Many pathologists, particularly those skilled in infectious disease
pathology, may choose not to order additional confirmatory tests such as immunohisto-
chemistry or ISH, particularly if viral cytopathic effects and the histopathologic features
of the disease are characteristic. Furthermore, the rapid direct immunofluorescent testing
now performed in many clinical virology laboratories provides same-day confirmation
of many viral infections, making ISH and IHC unnecessary. For these reasons and
possibly others, some laboratories have experienced a decline in requests for IHC or
ISH testing for viruses. A recent review of IHC antibody use for the detection of viruses
at the Cleveland Clinic Foundation disclosed that orders for HSV, CMV, and HPV
stains were among the most underused. These expensive commercial antibody stains
were expiring prior to use and had to be discarded, resulting in a monetary loss for
that section. In this regard, ISH offers an attractive alternative to IHC. Probes for ISH
generally offer an equal or even greater degree of specificity when compared with IHC,
but they may be ordered in small volumes, frozen, and validated in-house, to help
avoid expiration and monetary loss. Additionally, since any nucleic acid sequence may
be synthesized, ISH probes may be designed for a wide variety of viruses, including
those for which IHC stains are unavailable. Perhaps it is time once again to examine
which method is best for the detection of viruses in histologic sections.
In Situ Detection of Infectious Agents 51
Although ISH has been shown to be useful for the diagnosis and study of viral
diseases in histologic sections, its utility in the routine clinical virology laboratory
remains unclear. ISH has been used to confirm positive viral cultures, but, as with
IHC in histologic sections, in clinical virology ISH has to compete with high-quality
commercially available antibody stains that are used for culture confirmation. The
detection of other viruses, such as CMV and HPV, has been successfully accomplished
using solution hybridization, which is less labor intensive and more cost effective than
ISH. Finally, ISH has to compete with NAA, including quantitative methods, for the
detection of viruses such as HIV, HCV, and EBV. With the variety of molecular
methods currently being used in clinical virology and numerous new applications on
the horizon, the role of ISH in the clinical virology laboratory remains unclear.
BACTERIA DETECTION
Fluorescently labeled oligonucleotide probes have been used for the rapid detection
of bacteria, and the technical parameters of hybridization have been explored (49)
(Braun-Howland). These techniques have been used by microbiologists to study bacterial
physiology, enumerate viable bacteria, and examine biofilms (50–54). Such methods
have been used to study Legionella and Escherichia coli in water (55,56) and have
been employed by the food industry to study the bacterium that causes necrotizing
hepatopancreatitis of farm-raised shrimp (57). Applications also exist in veterinary
medicine. ISH has been used to study Sepulina and Mycoplasma hyopneumoniae
infections in pigs, which cause enterocolitis and respiratory infections, respectively
(58,59). It has been used to investigate the role of Haemobartonella felis in
infectious feline anemia and to study Chlamydia trachomatis infections in swine
and mice (60–62).
These techniques have also been used to examine a variety of agents of human
infections. In formalin-fixed, paraffin-embedded specimens ISH has been used to detect
fastidious bacteria. Several researchers have examined ISH for the detection of Helico-
bacter pylori in gastric biopsies and concur that this method is sensitive, specific and
useful (63–65). Recently, ISH has been shown to be comparable to rapid PCR (i.e.,
Light-Cycler, Roche Diagnostics, Indianapolis, IN) for the detection of Legionella
pneumophila and was found to be superior to the traditional Warthin-Starry stain (Figs.
18 and 19). The use of ISH has been explored for the diagnosis of Chlamydia, given
the difficulty of identifying inclusions during cytologic or histopathologic examination.
The role of Chlamydia trachomatis in chronic prostatitis, coronary artery atherosclerosis,
reactive arthritis, and stillbirth has been examined using ISH (66–69). The presence
of bacteria, undetectable by conventional Gram stain but identifiable by ISH, has been
demonstrated in the sinus-nasal mucosa and mucus of a few patients with chronic
sinusitis (70). The distribution of bacteria in the intestinal mucus layer in patients with
inflammatory bowel disease (IBD) has also been studied by ISH and compared with
patients without IBD, in an effort to describe differences in the protective capacity of
the intestinal mucus in this population (71).
Although it is clear that these techniques are capable of accurately identifying bacteria,
they have not become commonplace in the microbiology laboratory. In the clinical
microbiology laboratory, ISH, like Gram stain, but more highly specific, may be useful
52 Procop and Hayden
for the direct identification of bacteria in clinical specimens (72,73). Such technologies
are clearly important if the agent under investigation is uncultivable. In this regard,
ISH has been used to study the epidemiology of oral treponemes associated with
periodontal disease (74). A careful choice of specific ISH probes may be used to rapid-
ly identify particular organisms that are likely to be associated with particular diseases.
For example, Matsuhisa et al. (72) used ISH for the rapid identification of bacteria
in phagocyte smears from patients with septicemia. This methodology, however, has
not been confirmed in a large cohort of patients with septicemia. Furthermore, consider-
ing that many patients with bacteremia often have few circulating bacteria per milliliter
of blood, the sensitivity of such methods would have to compete with the highly
sensitive automated blood culture instruments used today. Similarly, Krimmer et al.
(73) used ISH to detect Staphylococcus species in specimens associated with prosthetic
orthopedic devices. Although useful for the direct identification of bacteria in clinical
specimens, another, possibly more practical use of ISH in the microbiology laboratory
may be the rapid identification and differentiation of common microorganisms in cul-
tures.
ISH has been used in the clinical microbiology laboratory for the rapid identification of
bacteria in positive blood cultures. Jansen et al. (12) and Kempf et al. (13) simultaneously
published articles examining the ability of ISH to detect and differentiate bacteria in
positive blood culture bottles using probes directed against specific target sequences
present in bacterial rRNA. Both genus-inclusive probes (i.e., all staphylococci or all
Enterobacteriaceae) and species-specific probes (i.e., Staphylococcus aureus or E. coli)
were used. For the identification of S. aureus in blood cultures, Jansen et al. (12)
showed a sensitivity of 66.7%, whereas Kempf et al. (13) demonstrated a sensitivity
of 100%, even though overlapping, nearly identical probe sequences were used. In
both studies, the specificity for the identification of S. aureus was 100%. This ability
to provide rapid speciation of staphylococci in blood cultures would probably have a
significant clinical impact by rapidly identifying S. aureus, the organism more likely
to be associated with serious disease. Furthermore, the rapid identification of S. aureus
by ISH allows for the judicial use of other rapid molecular methods for detecting the
mecA gene, the gene responsible for methicillin resistance. In these blood culture/ISH
studies, the accurate and rapid identification of other Gram-positive cocci (streptococci
and enterococci) and a variety of Gram-negative bacilli (12,13) was also achieved.
Possibly the most attractive feature of this technology is the greatly enhanced time to
identification for ISH compared with conventional methods. Kempf et al. (13) demon-
strated that with a limited number of ISH probes, the cause of 96.5% of positive blood
cultures could be identified by ISH within 2.5 h, whereas conventional culture took
between 1 and 3 days.
This construction of ISH batteries directed against common pathogens of particular
diseases may prove useful in areas other than blood culture. This approach has been
studied for the simultaneous detection and identification of common respiratory patho-
gens in cultures from patients with cystic fibrosis (75). Although it is slightly less
sensitive than culture, Hogardt et al. (75) used ISH to detect and identify Pseudomonas
aeruginosa, Burkholderia cepacia, Stenotrophomonas maltophilia, Haemophilus
influenzae, and S. aureus and demonstrated 90% sensitivity and 100% specificity of
ISH compared with culture.
In Situ Detection of Infectious Agents 53
Mycobacteriology is another area in which ISH may be useful for organism identifica-
tion in direct specimens and in culture. The utility of the direct identification of mycobac-
teria has obvious clinical importance, considering the slow growth of these organisms
and the fastidious nature of some mycobacteria, such as M. leprae, M. genevense, and
M. haemophilum. ISH with peptide nucleic acid probes has been used to distinguish
tuberculous and nontuberculous mycobacteria rapidly in smears from mycobacterial
cultures (76,77). This technology has also been used for rapid identification of M.
tuberculosis and nontuberculous mycobacteria species directly in sputa that contained
acid-fast bacilli (78). ISH applications have also been used in histologic sections to
identify definitely M. leprae, an organism that cannot be cultured in vitro (79).
FUNGUS DETECTION
As with other organism groups, ISH is particularly useful for uncultivable agents.
Several researchers have used ISH to detect the uncultivable fungus, Pneumocystis
carinii (88–90). Unlike most conventional histochemical stains, but like IHC, ISH
detects the trophozoite form of the organism, as well as the cyst forms (88). Immunohis-
tochemical or ISH staining methods for P. carinii may be particularly useful when few
organisms are present or when confusion exists between Pneumocystis and small yeast
forms wherein buds cannot be identified.
ISH has also been applied to the difficult histologic differential diagnosis of morpho-
logically similar yeast and yeast-like forms in human tissues. The nature of disease
and therapeutic options also differ for this group of organisms, which bespeaks the
need for accurate organism identification. ISH methods have been used successfully
to identify accurately Cryptococcus neoformans and the parasitic forms of the dimorphic
fungal pathogens Histoplasma capsulatum, Blastomyces dermatitidis, and Coccidioides
immitis in human tissue sections (84) (Figs. 10–17). The yeast cells of Candida albicans,
Candida tropicalis, and the agent of South American blastomycosis, Paracoccidioides
brasiliensis, have been identified using ISH methods (91–93). The laboratory diagnosis
of systemic fungal disease is often first made by the histopathologist. Therefore, accurate
ISH methods for detection of fungi in histologic sections would represent a significant
advance in diagnosis, especially for organisms like H. capsulatum that produce sugges-
tive, but not diagnostic, forms in tissue and require extended culture incubation for
growth.
As with bacteria, ISH may prove useful for the identification of fungi in the clinical
mycology laboratory. These methods have been used for the direct identification of
common yeast in blood cultures with 100% specificity (13). Additionally, because
saprophytic, nonpathogenic molds exist that demonstrate conidiation in culture similar
to that of the systemic dimorphic pathogens, molecular probes or other techniques are
necessary for confirmation of these fungi. It follows then that the same ISH probes
used to differentiate fungi in tissue sections may be useful for culture confirmation. In
this manner, in situ methods may compete with traditional assimilation and morphologic
methods of yeast and mold identification, especially as automation of ISH methods
becomes more widely available.
PARASITE DETECTION
CONCLUSIONS
REFERENCES
1. Sklar, J. (1985) DNA hybridization in diagnostic pathology. Hum. Pathol. 16, 654–658.
2. Speel, E. J. (1999) Robert Feulgen Prize Lecture. Detection and amplification systems for
sensitive, multiple-target DNA and RNA in situ hybridization: looking inside cells with a
spectrum of colors. Histochem. Cell Biol. 112, 89–113.
3. McNicol, A. M. and Farquharson, M. A. (1997) In situ hybridization and its diagnostic
applications in pathology. J. Pathol. 182, 250–261.
4. Jin, L. and Lloyd, R. V. (1997) In situ hybridization: methods and applications. J. Clin.
Lab. Anal. 11, 2–9.
5. Chandler, F. W. (1997) Approaches to the pathologic diagnosis of infectious diseases, in
Pathology of Infectious Diseases (Connor, D. H. and Chandler, F. W., eds.), Appleton &
Lange, Stamford, CT, pp. 3–7.
6. Chandler, F. W. and Watts, J. C. (1987) Pathologic Diagnosis of Fungal Infections. ASCP
Press. Chicago, IL.
7. Azumi, N. (1997) Immunohistochemistry and In situ hybridization techniques in the detec-
tion of infectious organisms, in Pathology of Infectious Diseases, (Connor, D. H. and
Chandler, F. W., eds.), Appleton & Lange, Stamford, CT, pp. 35–44.
8. Trofatter, K. F. Jr. (1997) Diagnosis of human papillomavirus genital tract infections. Am.
J. Med. 102, 21–27.
9. Zehbe, I., Hacker, G. W., Su, H., Hauser-Kronberger, C., Hainfeld, J. F., and Tubbs,
R. (1997) Sensitive in situ hybridization with catalyzed reporter deposition, streptavidin-
Nanogold, and silver acetate autometallography: detection of single-copy human papilloma-
virus. Am. J. Pathol. 150, 1553–1561.
10. Cheung, A. L., Graf, A. H., Hauser-Kronberger, C., Dietze, O., Tubbs, R. R., and Hacker,
G. W. (1999) Detection of human papillomavirus in cervical carcinoma: comparison of
56 Procop and Hayden
29. Multhaupt, H., Bruder, E., Elit, L., Rothblat, I., and Warhol, M. (1993) Combined analysis
of cervical smears. Cytopathology, image cytometry and in situ hybridization. Acta Cytol.
37, 373–378.
30. Tweddel, G., Heller, P., Cunnane, M., Multhaupt, and Roth, K. (1994) The correlation
between HIV seropositivity, cervical dysplasia and HPV subtypes 6/11, 16/18, 31/33/35.
Gyn. Oncol. 52, 161–164.
31. Tomita, T., Garcia, F., and Mowry, M. (1992) Herpes simplex hepatitis before and after
acyclovir treatment. Immunohistochemical and in situ hybridization study. Arch. Pathol.
Lab. Med. 116, 173–177.
32. Tomita, T., Chiga, M., Lenahan, M., and Balachandran, N. (1991) Identification of herpes
simplex virus infection by immunoperoxidase and in situ hybridization methods. Virchows
Arch. Pathol. Anat. Histopathol. 419, 99–105.
34. McClintock, J. T., Mosher, M., Thaker, S. R., et al. (1991) Culture confirmation of cytomega-
lovirus and herpes simplex virus by direct enzyme-labeled DNA probes and in situ hybridiza-
tion. J. Virol. Methods 35, 81–91.
35. Li, J. J., Huang, Y. Q., Cockerell, C. J., and Friedman-Kien, A. E. (1996) Localization
of human herpes-like virus type 8 in vascular endothelial cells and perivascular spindle-
shaped cells of Kaposi’s sarcoma lesions by in situ hybridization. Am. J. Pathol. 148, 1741–
1748.
36. Ohshima, K., Kikuchi, M., Kobari, S., Masuda, Y., Yoneda, S., and Takeshita, M. (1995)
Demonstration of Epstein-Barr virus genomes, using polymerase chain reaction in situ
hybridization in paraffin-embedded lymphoid tissues. Pathol. Res. Pract. 191, 139–147.
37. Flaegstad, T., Andresen, P. A., Johnsen, J. I., et al. (1999) A possible contributory role of
BK virus infection in neuroblastoma development. Cancer Res. 59, 1160–1163.
38. Papadaki, H. A., Stefanaki, K., Kanavaros, P., et al. (2000) Epstein-Barr virus-associated
high-grade anaplastic plasmacytoma in a renal transplant patient. Leuk. Lymph. 36, 411–415.
39. van de Rijn, M., Bhargava, V., Molina-Kirsch, H., et al. (1997) Extranodal head and
neck lymphomas in Guatemala: high frequency of Epstein-Barr virus-associated sinonasal
lymphomas. Hum. Pathol. 28, 834–839.
40. Pingel, S., Hannig, H., Matz-Rensing, K., Kaup, F. J., Hunsmann, G., and Bodemer, W.
(1997) Detection of Epstein-Barr virus small RNAs EBER-1 and EBER-2 in lymphomas
of SIV-infected rhesus monkeys by in-situ hybridization. Int. J. Cancer 72, 160–165.
41. Braun-Howland, E. B., Danielsen, S. A., and Nierzwicki-Bauer, S. A. (1992) Development
of a rapid method for detecting bacterial cells in situ using 16S rRNA-targeted probes.
Biotechniques 13, 928–934.
42. Vasef, M. A., Ferlito, A., and Weiss, L. M. (1997) Nasopharyngeal carcinoma with emphasis
on its relationship to Epstein-Barr virus. Ann. Otol. Rhinol. Laryngol. 106, 348–356.
43. Chan, J. K., Tsang, W. Y., Ng, C. S., Wong, C. S., and Lo, E. S. (1995) A study of the
association of Epstein-Barr virus with Burkitt’s lymphoma occurring in a Chinese popula-
tion. Histopathology 26, 239–245.
44. Choi, Y. J., Kang, C. S., Shin, W. S., et al. (1994) Epstein-Barr virus associated posttransplant
malignant lymphoma in renal allograft recipients. J. Korean Med. Sci. 9, 162–168.
45. Uner, A. H., Hutchinson, R. E., and Davey, F. R. (1994) Applications of in situ hybridization
in the study of hematologic malignancies. Hematol. Oncol. Clin. North Am. 8, 771–784.
46. Hennig, E. M., Di Lonardo, A., Venuti, A., Holm, R., Marcante, M. L., and Nesland, J.
M. (1999) HPV 16 in multiple neoplastic lesions in women with CIN III. J. Exp. Clin.
Cancer Res. 18, 369–377.
47. Southern, S. A., Graham, D. A., and Herrington, C. S. (1998) Discrimination of human
papillomavirus types in low and high grade cervical squamous neoplasia by in situ hybridiza-
tion. Diagn. Mol. Pathol. 7, 114–121.
48. Cho, N. H., Joo, H. J., Ahn, H. J., Jung, W. H., and Lee, K. G. (1998) Detection of human
58 Procop and Hayden
82. Taga, M. and Murata, M. (1994) Visualization of mitotic chromosomes in filamentous fungi
by fluorescence staining and fluorescence in situ hybrization. Chromosoma 103, 408–413.
83. Tenberge, K. B., Stellamanns, P., Plenz, G., and Robenek, H. (1998) Nonradioactive in
situ hybridization for detection of hydrophobin mRNA in the phytopathogenic fungus
Claviceps purpurea during infection of rye. Eur. J. Cell. Biol. 75, 265–272.
84. Hayden, R. T., Qian, X., Roberts, G. D., and Lloyd, R. V. (2000) In-situ hybridization for
the identification of yeast and yeast-like organisms in tissue sections. Abstract #992, United
States and Canadian Academy of Pathology, 89th Annual Meeting, New Orleans, LA. Mod.
Pathol. 13, 169A.
85. Kobayashi, M., Sonobe, H., Ikezoe, T., Hakoda, E., Ohtsuki, Y., and Taguchi, H. (1999)
In situ detection of Aspergillus 18S ribosomal RNA in invasive pulmonary aspergillosis.
Intern. Med. 38, 563–569.
86. Park, C. S., Kim, J., and Montone, K. T. (1997) Detection of Aspergillus ribosomal RNA
using biotinylated oligonucleotide probes. Diagn. Mol. Pathol. 6, 255–260.
87. Hanazawa, R., Murayama, S. Y., and Yamaguchi, H. (2000) In-situ detection of Aspergillus
fumigatus. J. Med. Microbiol. 49, 285–290.
88. Kim, J., Yu, J. R., Hong, S. T., and Park, C. S. (1996) Detection of Pneumocystis carinii
by in situ hybridization in the lungs of immunosuppressed rats. Korean J. Parasitol. 34, 177–
184.
89. Hayashi, Y., Watanabe, J., Nakata, K., Fukayama, M., and Ikeda, H. (1990) A novel
diagnostic method of Pneumocystis carinii. In situ hybridization of ribosomal ribonucleic
acid with biotinylated oligonucleotide probes. Lab. Invest. 63, 576–580.
90. Kobayashi, M., Urata, T., Ikezoe, T., et al. (1996) Simple detection of the 5S ribosomal
RNA of Pneumocystis carinii using in situ hybridisation. J. Clin. Pathol. 49, 712–716.
91. Lischewski, A., Amann, R. I., Harmsen, D., Merkert, H., Hacker, J., and Morschhauser, J.
(1996) Specific detection of Candida albicans and Candida tropicalis by fluorescent in situ
hybridization with an 18S rRNA-targeted oligonucleotide probe. Microbiology 142, 2731–
2740.
92. Lischewski, A., Kretschmar, M., Hof, H., Amann, R., Hacker, J., and Morschhauser, J.
(1997) Detection and identification of Candida species in experimentally infected tissue
and human blood by rRNA-specific fluorescent in situ hybridization. J. Clin. Microbiol.
35, 2943–2948.
93. De Brito, T., Sandhu, G. S., Kline, B. C., et al. (1999) In situ hybridization in paracoccidioido-
mycosis. Med. Mycol. 37, 207–211.
94. Velasquez, J. N., Carnevale, S., Labbe, J. H., Chertcoff, A., Cabrera, M. G., and Oelemann,
W. (1999) In situ hybridization: a molecular approach for the diagnosis of the microsporidian
parasite Enterocytozoon bieneusi. Hum. Pathol. 30, 54–58.
95. Pereira, M. C., Singer, R. H., and de Meirellas, M. N. (2000) Ultrastructural distribution of
poly (A)+ RNA during Trypanosoma cruzi-cardiomyocyte interaction in vitro: a quantitative
analysis of the total mRNA content by in situ hybridization. J. Eukaryot. Microbiol. 47, 264–
270.
96. Jambou, R., Hatin, I., and Jaureguiberry, G. (1995) Evidence by in situ hybridization for
stage-specific expression of the ATP/ADP translocator mRNA in Plasmodium falciparum.
Exp. Parasitol. 80, 568–571.
97. Unnasch, T. R., Bradley, J., Beauchamp, J., Tuan, R., and Kennedy, M. W. (1999) Character-
ization of a putative nuclear receptor from Onchocerca volvulus. Mol. Biochem. Parasitol.
104, 259–269.
98. Zurita, M., Bieber, D., and Mansour, T. E. (1989) Identification, expression and in situ
hybridization of an eggshell protein gene from Fasciola hepatica. Mol. Biochem. Parasitol.
37, 11–17.
Detecting Adenovirus 61
PROTOCOLS
Detecting Adenovirus
ISH with biotinylated probes is used to detect adenovirus DNA in paraffin sections.
Slides required for the above test include patient slides labeled for positive probe,
negative probe, and DNA probe, (use DNA procedure) and a known positive quality
control slide.
DEPARAFFINIZATION (RNASE-FREE)
1. Dry paraffin slides in 70°C oven 30 min
2. Deparaffinize in xylene, two changes. 7 min each
3. 100% ethanol, two changes. 2 min or 15 dips each
4. 95% ethanol, two changes. 2 min or 15 dips each
5. 2× standard saline citrate (SSC) rinse, two changes. 2 min each
6. 0.2 N HCl (0.2 N hydrochloric acid). 20 min
7. 2× SSC rinse, two changes. 2 min each
Delete steps 8–11 if tissue is not B-5 fixed.
8. Weigert’s iodine (remove mercury deposits). 5 min
9. 0.02% diethyl pyrocarbonate (DEPC)-treated water rinse. 1 min
10. 5% sodium thiosulfate. 1 min
11. 2× SSC rinse, two changes. 2 min each
STRINGENCY WASHING
Stringency washes are performed in shaking waterbath using prewarmed solutions.
Place each probe type into a separate rinse container for stringency washing.
1. Wash in 2× SSC (room temperature), two changes. 3 min each
2. Wash in prewarmed 0.5× SSC (37°C) rinse, and place in 37°C shaking
waterbath. (DNA probe slide requires (37°C) 0.5× SSC rinse and
37°C shaking waterbath.) 20 min
PROBE DETECTION
1. Buffer A (pH 7.5) rinse on orbital shaker. 3 min
2. Apply 300 µL of streptavidin-AP (BRL, 1/100; dilute with buffer
A + 1% normal swine serum and 0.3% Triton X-100) to each slide
and incubate within humid tray in 37°C oven. 20 min
3. Buffer A rinse on orbital shaker, two changes. 3 min each
4. Buffer C (pH 9.5) rinse on orbital shaker. 3 min
5. Prepare NBT/BCIP + levamisole solution by mixing the following
in order (300 µL per slide):
Buffer C (pH 9.5) 1.0 mL 2.0 mL 3.0 mL 4.0 mL 5.0 mL
Add: nitroblue tetrazolium 4.4 µL 8.8 µL 13.2 µL 17.6 µL 22.0 µL
(NBT; BRL)
(invert to mix)
Add: bromochloroindolyl 3.3 µL 6.6 µL 9.9 µL 13.2 µL 16.5 µL
phosphate
(BCIP; BRL)
(invert to mix)
Add: levamisole 8.0 µL 16.0 µL 24.0 µL 32.0 µL 40.0 µL
(Vector)
(invert to mix)
Total 1.0 mL 2.0 mL 3.0 mL 4.0 mL 5.0 mL
Detecting Adenovirus 63
ISH with digoxigenin probes is used to detect Epstein-Barr virus in paraffin sections.
Slides required for the above test are: patient slides labeled for positive probe, negative
probe, and RNA probe (use RNA procedure) and a known positive quality control slide.
DEPARAFFINIZATION (RNASE-FREE)
1. Dry paraffin slides in 70°C oven (minimum 30 min). 30 min
2. Deparaffinize in xylene, two changes. 7 min each
3. 100% ethanol, two changes. 2 min or 15 dips each
4. 95% ethanol, two changes. 2 min or 15 dips each
5. 0.02% DEPC-treated water rinse, two changes. 2 min each
6. Weigert’s iodine (remove mercury deposits). 5 min
7. 0.02% DEPC-treated water rinse. 1 min
8. 5% sodium thiosulfate. 1 min
9. 0.02% DEPC-treated water rinse, two changes. 2 min each
STRINGENCY WASHING
PROBE DETECTION
1. Buffer A (pH 7.5) rinse on orbital shaker. 3 min
2. Apply 300 µL of sheep anti-digoxigenin-AP (BM, 1/100; dilute
with buffer A + 1% normal swine serum and 0.3% Triton X-100)
to each slide and incubate within humid tray in 37°C oven. 20 min
3. Buffer A rinse on orbital shaker, two changes. 3 min each
4. Buffer C (pH 9.5) rinse on orbital shaker. 3 min
5. Prepare NBT/BCIP + levamisole solution by mixing the following
in order (300 µL per slide):
Buffer C (pH 9.5) 1.0 mL 2.0 mL 3.0 mL 4.0 mL 5.0 mL
Add: NBT (BRL) 4.4 µL 8.8 µL 13.2 µL 17.6 µL 22.0 µL
(invert to mix)
Add: BCIP (BRL) 3.3 µL 6.6 µL 9.9 µL 13.2 µL 16.5 µL
(invert to mix)
Add: Levamisole (Vector) 8.0 µL 16.0 µL 24.0 µL 32.0 µL 40.0 µL
(invert to mix)
Total: 1.0 mL 2.0 mL 3.0 mL 4.0 mL 5.0 mL
6. Wipe around section and apply NBT/BCIP + levamisole solution to slides.
Wrap tray with aluminum foil and develop the reaction. Monitor slides
closely the first 5–10 min and then every 5–10 min until a strong signal
free of background (nonspecific staining) is reached. Keep slides wrapped
in aluminum foil during development. Less time may be needed for some
patient cases to prevent background staining. 20–40 min
7. Buffer C rinse, using some agitation. 1 min
Epstein-Barr Virus (Eber I/II) 65
INTRODUCTION
Fluorescence in situ hybridization (FISH) arose from a marriage of classical DNA
hybridization in solution to modern molecular biologic techniques, most notably the
use of restriction endonucleases and, later, the polymerase chain reaction (PCR). The
former has made identification of both genes and relevant interspersed sequences
possible, whereas without the latter, many widely used probes would not be available.
FISH allows one to localize a specific DNA sequence to a specific chromosome, region
of a chromosome, or cell type. Generally, the morphology of the chromosome, cell,
or tissue of interest is preserved to permit unambiguous identification of the target
(1–3). Consequently, the technique has been widely used in a variety of venues. Gene
mapping (4), chromosome identification (5), demonstration of gene amplification in
some solid tumors (6,7), identification of chimeric populations, minimal residual disease,
and tumor cells admixed in normal tissue (8,9) are just a few examples. The advantages
of modern FISH techniques include:
1. Simple, straightforward, standardized techniques
2. Wide assortment of probes available commercially
3. Double labeling and use of multiple probes on a single cell
4. Rapid turnaround times (anywhere from 4 h to a maximum of 48 h).
Comparing modern FISH technology with the original experiments described by
Pardue and Gall (10), which entailed the generation of tritium-labeled DNA or RNA
probes by providing cell cultures with either tritiated thymidine or labeled nucleotide
triphosphates, it is clear that FISH is simpler. The relative ease is emphasized as
additional complex procedures follow labeling of the nucleic acids, i.e., the so-called
satellite sequences (so-called because they floated above the main band DNA in a
sucrose or CsCl gradient) were isolated by density gradient centrifugation and annealed
to chromosome spreads on slides. The radioactive signal was localized using autoradiog-
raphy, which involves coating the slide carrying the specimen with a photographic
emulsion (literally gelatin containing silver halide crystals); the emulsion must be
handled in the dark. Once dry, the coated slides are exposed in the dark, usually at
4°C for days to weeks, before development, drying, and counterstaining. Double-
labeling experiments were possible (11) using isotopes with different path lengths, such
67
68 Sheldon
as tritium and P32 and detecting them by using two coatings of emulsion separated by
a layer of cellophane.
FISH was developed in part due to the technical difficulties and limitations imposed
by autoradiography, in part because of pressure from various regulatory agencies to
reduce the use of radioisotopes, and in part because molecular biology was evolving.
Investigators seeking ways of avoiding labor-intensive classical cytogenetic analysis
by automation propelled the field of molecular cytogenetics. For example, various
combinations of fluorescent dyes with affinity for nucleotides and/or DNA were used
to stain suspensions of isolated chromosomes. The chromosome suspensions were then
analyzed by flow cytometry or fluorescence-activated cell sorting (12). Although flow
cytometry had limited success in completing a karyotype, some of the fluorochromes
and dyes were later used for FISH.
Prior to the advent of modern molecular biology, numerous studies were undertaken
to insert substituents into nucleotides or their precursors. The main obstacles were uptake
of the modified nucleotide through both the cell and nuclear membrane, appropriate
incorporation into DNA or RNA, and visualization of the labeled product.
One of the earliest solutions to the isotope problem was to use mercurated, rather
than tritiated nucleotides (13,14). Pyrimidines were mercurated either chemically or
enzymatically and incorporated into DNA or RNA. Following a hybridization reaction,
the signal was localized by its electron density, or by indirect immunofluorescence
using a hapten-sulfhydral conjugate, in which the sulfur ligand reacted with the mercury.
In other studies, fluorescent derivatives of nucleotides or polynucleotides of interest
were produced (15–17) with varying degrees of success; the adducts could then be
visualized. Later, the thymidine analog bromodeoxyuridine (BrdU), which was widely
used by cytogeneticists during the 1970s to demonstrate sister chromatid exchanges,
was incorporated into some probes; its presence is visualized with an antibody directed
against the BrdU.
In situ hybridization changed dramatically when, in 1981, Langer et al. (18) revolu-
tionized probe technology by synthesizing biotin-labeled polynucleotides that could be
detected with avidin conjugated to either a fluorochrome or peroxidase. DNA technology
became accessible to any reasonably sophisticated immunohistochemistry laboratory.
Early FISH publications demonstrated the localization of unique-sequence human
genes (19–21), but the first probes for human genes to become commercially available
were the α-satellite, or centromere probes. Three types of FISH probes are commonly
used and commercially available: α-satellites, which bind to the pericentromeric region
of a chromosome, unique-sequence probes, which recognize DNA sequences associated
with disease genes, and painting probes, which bind to a series of sequences on a
specific chromosome (summarized in Table 1). The availability of some probes was
markedly curtailed following the patenting of “blocking DNA” and subsequent legal
action by the patent holder, a scenario becoming all too common in modern molecular
biology. The remainder of this discussion describes in more detail the three types of
probes (their use and interpretation of the results), followed by a methods section, using
examples of commonly used probes.
Fluorescent In Situ Hybridization 69
Table 1
Common Types of FISH Probes
Class Size (Kb) Chromosomal location
Repeated sequence
—α-Satellite 150–900 Centromeres of each chromosome
—β-Satellite 150–900 Centromeres of chromosomes 1, 9, and 16
—Classic satellite 150–900 Centromeres of chromosomes 1, 9, 16 and Yq
—Telomere 10 Telomere of each chromosome
Unique-sequence
—Painting 80–500 Covers entire chromosome
—Gene-specific 80–500 Specific locus on chromosomea
—Subtelomere 20–100 Chromosome-specific unique sequenceb
a
Most of these probes are a solution with a marker to identify the specific chromosome.
b
These sequences are located just proximal to the telomere and may be associated with specific syn-
dromes.
Fig. 1. Diagram of the cell cycle. In cycling or dividing cells, the time between mitoses is
about 24 h. The initial phase, G1, is approximately 12 h. DNA synthesis takes 6–8 h, followed
by a second gap, or G2, which lasts about 4 h. Mitosis requires about 20 min.
Fig. 2. (A) Chromosome structure. Chromosomes condense sufficiently for analysis after
completion of DNA synthesis. The chromosomes are composed of four strands of DNA or two
double helices. Two copies of any DNA probe, shown as small boxes, will bind to each
chromosome. (B) Each sister chromatid (arrows) is composed of a DNA double helix. The
sister chromatids are attached at the centromere, which appears as a constriction at metaphase.
The terminal (distal from the centromere) ends of the chromosomes are called telomeres.
into a proliferative cycle, which in normal cells lasts on average 24 h, they move from
G0 to G1; G1 lasts about 12 h. The DNA synthetic or S-phase lasts for 6–8 h. A
second gap, or G2, phase takes about 4 h, although it may last from 2 to 6 h in normal
cells. This is followed by mitosis, which is completed in 20 min. Interphase cytogenetic
analysis is actually an in situ hybridization analysis, as it is performed on cells in
interphase, in other words, on cells that are not dividing.
Once a given chromosome has completed DNA synthesis, that chromosome is
composed of four DNA strands or two DNA double helices, as seen in Fig. 2A. At
Fluorescent In Situ Hybridization 71
metaphase, each double helix forms a sister chromatid; sister chromatids remain associ-
ated at the centromere (Fig. 2B). Whereas one chromosome may have completed DNA
synthesis, the other chromosome of that pair may not complete DNA synthesis at the
same time, e.g., the maternally derived chromosome 2 may go through DNA synthesis
before the paternally derived chromosome 2 does. The former fact explains why in
FISH on chromosome spreads, two closely juxtaposed signals are seen on a single
chromosome; the latter fact explains why in FISH on nuclei, two, three, or four signals
are seen in cells that are diploid. Although chromosome banding techniques are a great
advance in the accurate identification of chromosomes and in gene mapping, it should be
noted that a single chromosome band may encompass 5–10 million bp, or 10–50 genes.
Mitosis itself is composed of four phases. Prophase is when the chromosomes
condense and the nuclear membrane and the nucleoli dissolve. So-called high-resolution
chromosome analysis is done during mid to late prophase; although it may seem that
only a single chromatid is present, the chromosomes are extremely elongated, making
resolution of the chromatids, as well as the centromere, difficult. During metaphase
the chromosomes line up, forming the metaphase plate in preparation for the sister
chromatids to segregate to the two poles during anaphase. Conventional cytogenetic
analysis is done during metaphase, when the condensing chromosomes are shorter and
less likely to overlap than they are at prophase. The sister chromatids and centromere
are easily resolved as the cell moves towards anaphase. During anaphase the sister
chromatids separate and begin to move to opposite poles; the cell membrane begins
to invaginate in order to form two daughter cells. Telophase is signaled by the condensa-
tion of the chromosomes into nuclei, reformation of the nuclear membrane and nuclei,
and pulling apart of the daughter cells.
down into satellites I, II, and III, based on buoyant density. β-Satellites are also located
in the pericentromeric region (24).
The chromosomes that do not have specific α-satellite probes are the so-called D-
and G-group chromosomes, chromosomes 13–15, 21, and 22, which are acrocentric
and carry the ribosomal RNA genes on a “stalk” just above their short arms. To confuse
the issue further, the D and G chromosomes are topped by a structure referred to as
a “satellite,” which is unrelated to satellite sequences in either mouse or human. The
α-satellite sequences are located near the centromere, and not near the structure called
the satellite in these acrocentric chromosomes. There is sufficient sequence homology
between the α-satellite region of chromosomes 13 and 21, and 14 and 22 (25,26) to
preclude the use of conventional α-satellite probes to distinguish these chromosomes.
This becomes relevant when selecting probes for evaluation of cells for trisomy or
confirming that the chromosome of interest is present when using FISH to evaluate
chromosomes for structural abnormalities. For instance, an α-satellite probe is not a
good choice to distinguish trisomy 13 from trisomy 21, an issue that often arises in
prenatal diagnosis. Similarly, if one were evaluating a highly aneuploid cell line for
the presence of translocations involving genes near the retinoblastoma locus on chromo-
some 13, it would be prudent to use a probe for another gene on chromosome 13, but
not an α-satellite probe, to confirm that the cell had two copies of chromosome 13.
α-Satellites are often thought of as chromosome enumeration probes. They are easy
to use, and in general will give a large, bright signal with little background. They are
used in a variety of settings including demonstration of changes of ploidy in tumors,
distinguishing diploid from triploid populations in hydropic pregnancies, identifying
numerical abnormalities in leukemia and lymphoma, sex determination, and for prenatal
diagnosis of trisomy and of sex chromosome anomalies. In Fig. 3, an α-satellite probe
for the X chromosome was used; a single signal from the fluorescein is noted in both
a single chromosome and the interphase nuclei. The result is clear: one X centromere
is present; the interpretation, however, could be a female with Turner’s syndrome, a
normal male, or a person with an intact X centromere, in an otherwise structurally
abnormal X chromosome. α-Satellite probes are not informative for demonstrating
most structural abnormalities, such as a translocations or deletions, unless they involve
the centromere.
Nomenclature
As FISH techniques came into increased use for both investigations and research,
it became obvious that standard cytogenetic nomenclature was not adequate to report
a result. Interphase cytogenetic analyses were particularly problematic. Consequently,
the International Standing Committee on Human Cytogenetic Nomenclature (ISCN)
met and issued a revision of the standard nomenclature in 1995 (27). To summarize
for α-satellite or chromosome enumeration probes, there are three general ways to
report a result. If both a standard karyotype and in situ hybridization have been done,
the results of the karyotype are reported, followed by a period and “ish,” followed by
the in situ hybridization result and the probe used. For example, if there were some
question about the identity of an extra chromosome 22 that was resolved using FISH,
the result would be reported as 47,XX,+22.ish 22(D22Z1×3). If, on the other hand, no
karyotype was performed, and three signals were noted with a probe for chromosome
Fluorescent In Situ Hybridization 73
Fig. 3. FISH performed on cultured human lymphocytes, prepared for conventional cytoge-
netic analysis. The probe used is DXZ1, an alpha satellite directed against the X chromosome’s
centromere. A single bright signal (arrow) is seen on a chromosome and in the interphase nuclei.
Counterstain is ethidium bromide.
18, the result would be written as nuc ish 18cen (D18Z1×3). Although the shorthand
appears nightmarishly complex, clinical cytogenetics reports must still include an inter-
pretation or explanation of the nomenclature.
Telomere Probes
Before leaving the topic of repeated sequences, another type of repeat probe should
be noted, the TTAGGGn repeat at the terminus of the chromosome arm or telomere
(Fig. 2B); this is the signal for DNA polymerase that it has reached the end of the
chromosome. This sequence is repeated over and over to a total size of about 10 kb
(28). Although this is a repeated sequence, the signal generated by a FISH study is
generally quite small. Some of the more interesting work done with these probes is to
utilize them to evaluate whether an apparently terminal translocation is actually recipro-
cal (29), as cytogenetic dogma maintains, or is indeed terminal.
Unique-Sequence Probes
Unique-sequence probes recognize DNA sequences or portions of genes that are
associated with a phenotype of interest. They are associated with either constitutional
disorders or acquired disorders. In the former an abnormal sequence is present or a
normal sequence is absent in all cells in the affected individual. With acquired disorders,
such as malignancies, only those cells involved in the malignant process either will
have an abnormal sequence, generally arising from a translocation, or will have lost a
normal sequence.
74 Sheldon
Table 2
Some Common FISH Probes for Genetic Disorders and Microdeletion Syndromes
Disorder Chromosomal locus Gene or genes involved
Cornelia de Lange syndrome 3q26.3 CDLa
Wolf-Hirschorn syndrome 4p16 WHSCR
Cri du chat syndrome 5p15.2 Two genes: cry and facies
Williams syndrome 7q11.2 Elastin
Retinoblastoma syndrome 13q14 RB1
Angelman’s syndrome 15q11.2 GABRB3
Prader-Willi syndrome 15q11.2 SNRPN
Rubinstein-Taybi syndrome 16p13.3 CREB binding proteina
Smith-Magenis syndrome 17p11.2 FL11
Miller-Diecker syndrome 17p13.3 LIS
DiGeorge syndrome (velocardiofacial) 22q11.2 D22S75 or Tuple1
Kallmann syndrome Xp22
Steroid sulfatase deficiency Xp22 Steroid sulfatase
a
Although FISH is diagnostic, these probes are not commercially available.
Fig. 5. A two-color microdeletion probe. The green signal is a marker for chromosome 15,
in this case the PML gene; the red signal shows the Prader-Willi region on proximal 15q.
Table 3
Common FISH Probes for Malignancies
Disordera Chromosomal locus Gene or genes involved
Neuroblastoma 2p23 N-myc
Prostate 8p22 Lipoprotein lipase
Carcinoma or lymphoma 8q24 C-myc
Breast (prostate, other) 17q12 Her2/neu
Carcinoma 17p13 p53
Breast 20q13 Amplified in breast cancer
Ki-1 lymphoma 2p23;5q35 ALK;NPM
MDS/AML 5q33-q34 5q- critical region
MDS/AML 7q31 7q- critical region
AML-M2 8q22;21q22.1 AML1;ETO
CML/AML/ALL 9q34;22q11.2 BCR;ABL
ALL/AML 11q23 MLL
ALL 12p13;22q22.3 ETV6 (TEL);AML1
CLL 13q14 D13S25
Lymphoma (follicular) 14q32;18q21 IgH;BCL2
AML-M4eo 16p13q22 CBFP
Promyelocytic leukemia 15q2217q21 PML;RARα
MDS or Polycythemia vera 20q12 20q critical region
a
MDS, myelodysplastic syndrome; AML, acute myelogenous leukemia; CML, chronic myelogenous
leukemia; ALL, acute lymphocytic leukemia; CLL, chronic lymphocytic leukemia.
Fig. 6. A two-color probe for BCR and ABL genes on a touch preparation from a spleen.
Note occasional fusion signals (arrows), which are suggestive of a translocation.
Painting Probes
Painting probes, or whole chromosome paints (wcp), are constructed from a collection
of unique DNA sequences that cover or are specific for a single chromosome. A good
painting probe covers the entire chromosome, with little or no crossreactivity with
the other chromosomes, with the following caveat. Because the painting probes are
constructed to bind to unique-sequence DNA, the more highly repetitive DNA sequences
78 Sheldon
Fig. 7. Painting probe for chromosome 9. Note that the centromeres, which contain repeated
sequences, do not paint. Often a faint banding pattern can be seen with paints because the
distribution of unique sequences is not even. (See Color Plate 3 following page 208.)
at the centromeres and telomeres may be spared (Fig. 7). The original painting probes
were made by lysing mitotic cells and sorting out individual chromosomes using a
flow cytometer (31–33). One would wind up with, for example, a tube of intact
chromosome 17. The chromosomes would then be cut with a restriction endonuclease
and the unique sequences on chromosome 17 amplified by so-called Alu PCR or,
more correctly, interspersed repetitive sequence (IRS) PCR (34,35). Alu and other
intermediate repetitive sequences are interspersed through the human genome; conse-
quently, these sequences can be used to create primers for PCR amplification of the
unique sequences between the Alu sites. Because the exact nature of the unique
sequences is not known, this is considered an anonymous DNA library: the genes for
p53, BRCA-1, and the retinoic acid receptor may be included in this mixture, but it
would be both impossible and unnecessary to identify them as individual bands. Then,
either the PCR products are cloned into plasmids or cosmids, or further PCR reactions
are run, this time in the presence of a compound such as biotin or digoxigenin attached
to one or more of the bases in order to label the resulting probe.
Some of the commercial painting probes are available directly conjugated with a
fluorochrome, whereas others use an indirect method and require incubation with
an antibody or binding protein, such as an avidin-fluorochrome or anti-dogoxigenin-
fluorochrome for their detection. Each method has its advantages. The indirect method
allows for amplification of a weak signal; the direct method is faster and easier to use
when more than one chromosome is being probed on the same slide. The specificity
Fluorescent In Situ Hybridization 79
of the binding of a probe to its target sequence is in part obtained by adjusting the
stringency of the hybridization conditions.
Painting probes are useful for identification of chromosomes carrying complex trans-
locations that do not include the centromere, i.e., pieces of chromosomes attached to
the ends of or inserted into other chromosomes. They are also useful for interphase
analysis of translocations for which a specific probe is not available. For instance, if
a patient’s tumor carried an unusual translocation between chromosomes 9 and 22, not
involving the usual breakpoints in a Philadelphia chromosome, using a painting probe
for chromosome 22, the analysis would proceed as follows.
Normal cells will have two copies of chromosome 22; mitotic cells will have two
brightly staining signals. Cells with a translocation, however, will have three signals:
a small chromosome 22, a normal chromosome 22, and a chromosome 9 with a small
piece of chromosome 22 translocated on to it. Only the end of chromosome 9 with the
fluorescein-tagged probe for chromosome 22 would be seen. Chromosomes other than
22 will be stained red with a propidium iodide counterstain, as will the nuclei of cells
that are not in mitosis. A more sophisticated approach might involve a probe for
chromosome 9 labeled with a second fluorochrome, such as rhodamine, as well as the
probe for chromosome 22. This type of analysis has been taken one step further with
the marketing of slides impregnated with combinations of three different painting probes
in a series of eight squares, making it possible to analyze a cell suspension for the
presence of common tumor translocations with a single hybridization reaction.
Spectral Karyotyping and M-FISH
M-FISH and spectral karyotyping (SKY) are specialized methods for using painting
probes so that each pair of chromosomes is visualized as a different color (reviewed
in ref. 36). Each uses a set of five fluorochromes in combination to generate the
colored probes; the former uses combinations of filter sets to visualize the fluorescent
chromosomes (37,38), whereas the latter uses interference spectroscopy or interferome-
try (39). These methods are useful for analysis of complex cytogenetic changes.
Nomenclature for Whole Chromosome Paints
The most common settings in which these probes are used are identification of a
supernumerary unidentified marker chromosome (mar) or when additional chromosomal
material is present on the terminus of a chromosome. In the former case, a result might
be reported as 47,XX,+mar.ish der(15)(wcp15+). In the case of additional material
present on a chromosome 17, which is demonstrated to be part of chromosome 18 by
FISH, the result might be written as 46,XX,add(17)(p13).ish der(17)t(17;18)(p13;p or
q)(wcp18+). The ISCN text (27) should be referred to for more detailed information.
FISH procedures rely on the fact that nucleic acid sequences will bind to one another
if sufficient sequence homology is present. All FISH procedures involve denaturation
of the probe, denaturation of the target sequence, placing the probe and target in contact
and allowing them to reanneal, all of which may be done simultaneously. As outlined
80 Sheldon
above, the probe and the target must be visualized, either by direct or indirect fluores-
cence of the probe. Nonspecific binding of the probe to the cellular DNA is reduced
or eliminated by a wash following the reannealing or hybridization step.
Unlike Southern and PCR-based analyses, much of the specificity of FISH probes
is derived from the posthybridization wash. The concept of stringency has been men-
tioned. This refers to the ability of DNA sequences to bind to one another when the
sequence homology is not exact, for instance, sequences that are rich in adenine and
thymine (A-T) and highly repeated will often hybridize to one another. To eliminate
this nonspecific binding, one would use a high-stringency wash. Factors that increase
the stringency, and make it more difficult for inexactly matched hybrids to remain bound,
include increasing the temperature (from 37°C to 43°C), increasing the formamide
concentration (from 50 to 70%), decreasing the salt concentration (from 0.4X standard
saline citrate [SSC] to 0.25X SSC). α-satellite probes contain the most repeated
sequences and will require a higher stringency posthybridization wash than will either
unique-sequence probes or whole chromosome paints.
Examples
The examples given in the Protocols section will serve to illustrate the specifics of
using FISH probes (see pp. 84–88). In the first example, the specimen is a lymph node
from a patient suspected of having involvement by a tumor that has a trisomy for
chromosome 3. The lymph node was processed for routine histology. An α-satellite
probe can be used on a paraffin section of the lymph node to evaluate it for metastatic
disease. The second example is a touch preparation from the spleen of a patient who
had been treated for chronic myelogenous leukemia, but who is experiencing a new
Fluorescent In Situ Hybridization 81
ACKNOWLEDGMENTS
The author would like to thank Nanci Lefebvre and Beth Cox at the University of
Michigan Cytogenetics Laboratory, Ann Arbor, MI, for assistance with the FISH studies.
The staff of Diagnostic Cytogenetics, Seattle, WA, contributed some of the slides for
two-color FISH.
REFERENCES
1. Haferlach, T., Winkermann, M., Loffler, H., et al. (1996) The abnormal eosinophils are
part of the leukemic cell population in acute myelomonocytic leukemia with abnormal
eosinophils. Blood 87, 2459–2463.
2. Bentz, M., Schroder, M., Herz, M., Stilgenbauer, S., Lichter, P., and Dohner, H. (1993)
Detection of trisomy 8 on blood smears using fluorescence in situ hybridization. Leukemia
7, 752–757.
3. Thompson, C. T., LeBoit, P. E., Nederlof, P. M., and Gray, J. W. (1994) Thick section
fluorescence in situ hybridization on formalin-fixed, paraffin-embedded archival tissue pro-
vides a histogenetic profile. Am. J. Pathol. 144, 237–243.
4. Price, P. M. and Hirschorn, K. (1975) In situ hybridization for gene mapping. Cytogenet.
Cell Genet. 14, 395–401.
5. Popp, S., Jauch, A., Schindler, D., et al. (1993) A strategy for the characterization of minute
chromosome rearrangements using multiple color fluorescence in situ hybridization with
chromosome-specific DNA libraries and YAC clones. Hum. Genet. 96, 527–532.
6. Taylor, C. P., McGuckin, G., Bown, N. P., et al. (1994) Rapid detection of prognostic
genetic factors in neuoblastoma using fluorescence in situ hybridization on tumour imprints
and bone marrow smears. Br. J. Cancer 69, 445–451.
7. Press, M. F., Bernstein, L., Thomas, A., et al. (1997) HER-2/neu gene amplification charac-
terized by fluorescence in situ hybridization. J. Clin. Oncol. 15, 2894–2904.
8. Zhan, L., Chang, K-S., Estey, E. H., Hayes, K., Deisseroth, A. B., and Liang, J. C. (1995)
Detection of residual leukemic cells in patients with acute promyelocytic leukemia by the
fluorescence in situ hybridization method. Blood 85, 495–499.
9. Buno, I., Wyatt, W. A., Zinsmeister, A. R., Dietz-Band, J., Silver, R. T., and Dewald, G.
(1998) A special fluorescent in situ hybridization technique to study peripheral blood
and assess the effectiveness of interferon therapy in chronic myeloid leukemia. Blood
92, 2315–2321.
10. Pardue, M. L. and Gall, J. G. (1970) Chromosomal localization of mouse satellite DNA.
Proc. Natl. Acad. Sci. USA 168, 1356–1358.
82 Sheldon
11. Prescott, D. M. and Bender, M. (1962) RNA and protein synthesis during mitosis in
mammalian tissue culture cells. Exp. Cell Res. 26, 260–268.
12. Cremer, C., Gray, J. W., and Ropers, H-H. (1982) Flow cytometric characterization of a
Chinese hamster × man hybrid cell line retaining the human Y chromosome. Hum. Genet.
60, 262–266.
13. Dale, R. M. and Ward, D. C. (1975) Mercurated polynucleotides: new probes for hybridiza-
tion and selective polymer fractionation. Biochemistry 14, 2458–2469.
14. Hopman, A. H. N., Wiegant, J., and VanDuijn, P. (1987) Mercurated nucleic acid probes,
a new principle for non-radioactive in situ hybridization. Exp. Cell Res. 169, 357–368.
15. Ward, D. C., Reich, E., and Stryer, L. (1969) Fluorescence studies of nucleotides and
polynucleotides. I. Formycin, 2-aminopurine riboside, 2,6-diaminopurine riboside, and their
derivatives. J. Biol. Chem. 244, 1228–1237.
16. Ward, D. C. and Reich, E. (1972) Fluorescence studies of nucleotides and polynucleotides.
II. 7-Deazanebularin: coding ambiguity in transcription with base pairs containing fewer
than two hydrogen bonds. J. Biol. Chem. 247, 705–719.
17. Ward, D. C., Horn, T., and Reich, E. (1972) Fluorescence studies of nucleotides and
polynucleotides. 3. Diphosphopyridine nucleotide analogues which contain fluorescent
purines. J. Biol. Chem. 247, 4014–4020.
18. Langer, P. R., Waldrop, A. A., and Ward, D. C. (1981) Enzymatic synthesis of biotin-
labeled poly nucleotides: novel nucleic acid affinity probes. Proc. Natl. Acad. Sci. USA
78, 6633–6637.
19. Cremer, T., Landegent, J., Bruckner, A., et al. (1986) Detection of chromosome aberrations
in the human interphase nucleus by visualization of specific target DNAs with radioactive
and non-radioactive in situ hybridization techniques; diagnosis of trisomy 18 with probe
LI.84. Hum. Genet. 74, 346–352.
20. Julien, C., Bazin, A., Guyot, B., Forstier, F., and Defos, F. (1986) Rapid prenatal diagnosis of
Down’s syndrome with in situ hybridization of fluorescent DNA probes. Lancet 2, 863–864.
21. Lichter, P., Cremer, T., Tang, C-J., Watkins, P. C., Manuelidis, L., and Ward, D. C. (1988)
Rapid detection of human chromosome 21 aberrations by in situ hybridization. Proc. Natl.
Acad. Sci USA 85, 9664–9668.
22. Manuelidis, L. (1978) Chromosomal location of complex and simple repeated human DNAs.
Chromosoma 66, 23–32.
23. Willard, H. F. (1985) Chromosome-specific organization of human alpha satellite DNA.
Am. J. Hum. Genet. 37, 524–532.
24. Waye, J. S. and Willard, H. F. (1989) Human beta satellite DNA: genomic organization
and sequence definition of a class of highly repetitive tandem DNA. Proc. Natl. Acad. Sci.
USA 86, 6250–6254.
25. Choo, K. H., Vissel, B., Brown, R., Filby, R. G., and Earle, E. (1988) Homologous alpha
satellite sequences on human acrocentric chromosomes with selectivity for chromosomes
13, 14 and 21. Nucleic Acids Res. 16, 1273–1284.
26. Jorgensen, A. L., Kolvraa, S., Jones, C., and Bak, A. L. (1988) A subfamily of alphoid
repetitive DNA shared by the NOR-bearing human chromosomes 14 and 22. Genomics
3, 100–109.
27. ISCN, Mitelman, F., ed. (1995) An International System for Human Cytogenetic Nomencla-
ture. S. Karger, Basel.
28. Moyzis, R. K., Buchingham, J. M., and Cram, L. S. (1988) A highly conserved repetitive
DNA sequence, TTAGGGn, present at the telomeres of human chromosomes. Proc. Natl.
Acad. Sci. USA 85, 6622–6625.
29. Park, V. M., Gutashaw, K. M., and Wathen, T. M. (1992) The presence of interstitial
telomeric sequences in constitutional chromosome abnormalities. Am. J. Hum. Genet. 50,
914–923.
Fluorescent In Situ Hybridization 83
30. Dewald, G., Stallard, R., Alsaadi, A., et al. (2000) A multicenter investigation with D-
FISH BCR/ABL1 probes. Cancer Genet. Cytogenet. 116, 97.
31. Lichter, P., Cremer, T., Border, J., Manuelidis, L., and Ward, D. (1988) Delineation of
individual human chromosomes in metaphase and interphase cells by in situ suppression
hybridization using recombinant DNA libraries. Hum. Genet. 80, 224.
32. Pinkel, D., Landegent, J., Collins, C., et al. (1988) Fluorescence in situ hybridization with
human chromosome specific libraries. Proc. Natl. Acad. Sci. USA 85, 9138–9142.
33. Van Dilla, M., Deaven, L., Albright, K., et al. (1986) Human chromosome-specific DNA
libraries: construction and availability. Biotechnology 4, 537–552.
34. Nelson, D., Ledbetter, S., Corbo, L. et al. (1989) Alu polymerase chain reaction: a method
for rapid isolation of human-specific sequences from complex DNA sources. Proc. Natl.
Acad. Sci. USA 86, 6686–6690.
35. Ledbetter, S., Nelson, D., Warren, S., and Ledbetter, D. (1990) Rapid isolation of DNA
probes within specific chromosome regions by interspersed repeated sequence polymerase
chain reaction. Genomics 6, 475–481.
36. Jalal, S. M. and Law, M. E. (1999) Utility of multicolor fluorescent in situ hybridization
in clinical cytogenetics. Genet. Med. 1, 181–186.
37. Spiecher, M. R., Ballard, S. G., and Ward, D. C. (1996) Karyotyping human chromosomes
by combinatorial multi-fluor FISH. Nat. Genet. 12, 368–375.
38. Eils, R., Uhrig, S., Saracoglu, K., et al. (1998) An optimized, fully automated system for
fast and accurate identification of chromosome rearrangements by multiplex-FISH (M-
FISH). Cytogenet. Cell Genet. 82, 160–171.
39. Schrock, E., duManoir, S., Veldman, T., et al. (1996) Multicolor spectral karyotyping of
human chromosomes. Science 273, 494–497.
40. American College of Medical Genetics. (1999) Standards and Guidelines for Clinical
Genetics Laboratories, 2nd ed., Rockville, MD.
41. Wolfe, K. and Herrington, C. (1997) Interphase cytogenetics and pathology: a tool for
diagnosis and research. J. Pathol. 181, 359–361.
42. van de Kaa, C., Nelson, K., Ramaekers, F., Vooijs, P., and Hopman, A. (1991) Interphase
cytogenetics in paraffin sections of routinely processed hydatidiform moles and hydropic
abortions. J. Pathol. 165, 281–287.
43. Poddighe, P., Ramaekers, F., and Hopman, A. (1992) Interphase cytogenetics of tumors.
J. Pathol. 166, 215–224.
44. Hopman, A., van Hooren, E., van de Kaa, C., Vooijs, P., and Ramaekers, F. (1991) Detection
of numerical chromosome aberrations using in situ hybridization in paraffin sections of
routinely processed bladder cancers. Mod. Pathol. 4, 503–513.
45. Long, A., Mueller, J., Schwartz, J., Barrett, K., Schwartz, R., and Wolfe, H. (1992) High
specificity in situ hybridization. Diagn. Mol. Pathol. 1, 45–57.
46. Arnoldus, E. P. J., Dreef, E., Noodermeer, I. A., et al. (1991) Feasibility of in situ hybridiza-
tion with chromosome specific DNA probes on paraffin wax embedded tissue. J. Clin.
Pathol. 44, 900–904.
47. Seto, E. and Yen, TSB. (1987) Detection of cytomegalovirus infection by means of DNA
isolated from paraffin-embedded tissues and dot hybridization. Am. J. Pathol. 127, 409–413.
84 Protocols
PROTOCOLS
FISH
The sample for the lymph node is a routine 4-µm paraffin section cut on positively
charged slides (PLUS or silane-coated) and then heated in a 60°C oven for no more
than 15 min. The control, both positive and negative, is blood cells present in the
specimen, which should have a diploid number of signals. For the evaluation of BCR/
ABL, a series of touch or imprint preparations were made of a spleen removed from
a patient with possible chronic myelogenous leukemia. One slide was fixed and stained
with Wright-Giemsa to confirm the presence of tumor, and the rest were processed as
described below. Internal controls include the presence of two signals in the majority
of cells (corresponding to the normal cellular genes) and the absence of multiple signals
in stroma and blood cells.
MATERIALS
Fixing Smears and Imprints
1. Methanol.
2. Glacial acetic acid.
A mixture of three parts methanol to one part acetic acid is prepared within 1 h of
use. This fixative is referred to as acid-alcohol, or sometimes as modified Carnoy’s.
Prepare these solutions using sterile reagents and containers; thaw at 37°C and
use immediately.
Denaturation
Prepare all the following solutions using sterile water and containers.
1. 20× SSC: 3 M NaCl (175.32 g/L), 0.3 M NaCitrate (88.23 g/L), 1 L deionized water. Dissolve
the two salts separately and mix. This is a stock solution. The pH will need to be adjusted
to 6.8–7.0 upon dilution.
2. 70% formamide, pH 7.0, in 2× SSC (see Note 1).
FISH 85
Hybridization
The probes used are commercially available and are provided with a suitable hybrid-
ization buffer. The exact ingredients of the hybridization buffers are proprietary, but
they generally contain 50% formamide, 10% dextran sulfate, 0.01% sheared salmon
sperm DNA in 2× SSC. The BCR/ABL (Oncor, Gaithersburg, MD) does not require
dilution. The α-satellite or chromosome enumeration probe (CytoCell, Oxfordshire,
UK) is supplied affixed to a plastic cover slip and comes with hybridization buffer, as
described above.
1. Hybridization chamber: these are available commercially, or use a plastic box with a lid, or
a glass baking dish covered with plastic wrap; place one or two damp paper towels in the
bottom to maintain the humidity.
2. Rubber cement.
3. 37°C incubator, without CO2, or 37°C oven, or slide warmer.
Posthybridization Wash
1. 50% formamide in 2× SSC, pH 7.0, or 0.4× SSC, pH 7.0.
2. Phosphate-buffered detergent (PBD): phosphate-buffered saline, pH 6.8 (1 L), 130 mM
sodium chloride (FW = 58.4) 7.59 g, 7 mM dibasic sodium phosphate (FW = 142) 0.99 g,
3 mM monobasic sodium phosphate (FW = 138) 0.41 g. Dissolve in deionized water in
order given and then add Triton X-100, 0.05% (0.05 mL/L).
Localization of Probe
Anti-digoxigenin antibody labeled with appropriate fluorochrome or avidin labeled
with appropriate fluorochrome; counterstains include DAPI or propidium iodide in
Antifade (Sigma, St. Louis, MO).
METHODS
Preparing the Target
In working with paraffin-embedded tissue, one must remove 1) the paraffin (deparaf-
finization or dewaxing), 2) the protein that has been crosslinked to the DNA, and 3)
DNA crosslinks. Many manufacturers of DNA probes recommend that the sections not
be baked at 60°C; this depends on the probe. It may be helpful to bake for 15 min or
so to improve adhesion of the section to the slide. Slides should be positively charged,
such as silane-treated slides, to improve adhesion.
Fix smears and imprints, such as the spleen in this example, by laying the slides on
a horizontal surface; cover the surface of the specimen with acid-alcohol for 2 min.
The excess fix is tilted off, and the slide is allowed to air-dry for at least 1 h. Skip to
the Denaturation section below. Paraffin sections, such as the section of lymph node
in this example, require dewaxing and several additional steps (steps 1–3) to permit
binding of the probe.
1. Dewaxing. Slides are incubated in a series of Coplin jars at room temperature as follows:
xylene, two changes, 10 min each; xylene/ethanol (50:50) 5 min; ethanol, two changes, 5
min each; air-dry at least 5 min.
2. Thiocyanate pretreatment. 1 M sodium thiocyanate in deionized water; heat to 80°C and
incubate slides for 10 min. This can be done on a slide warmer. Flood the slide with about
86 Protocols
2 mL of the reagent to avoid evaporation. Following incubation, rinse the slide with deionized
water, two changes, 2 min. Briefly drain slides to remove excess water.
3. Enzymatic pretreatment. Several methods are described. This laboratory uses pepsin; some
of the commercial kits use proteinase K. Pepsin 4 mg/mL in 0.2 M HCl (store stock at
−20°C). Incubate at 37°C for 5–15 min (may require up to 45 min; see Note 2). The lymph
node was treated for 20 min. Rinse with deionized water, two changes, 2 min. Dehydrate
through graded alcohols, 70%, 80%, 95%, 100%, and allow to air-dry. This is critical to
avoid diluting the formamide in the following, denaturation step.
Denaturation
Generally, both the probe and the target DNA will require denaturation, but this
depends on the manufacturer’s instructions. Some probes require no denaturation, and
some will tolerate co-denaturation, in which the probe solution is placed on the target
area of the slide, a cover slip is sealed in place with rubber cement, and then the slide
is heated to between 70° and 100°C for 2 : 10 min. To denature the cells or sections
on a slide in 70% formamide.
1. 70% formamide in 2× SSC, pH 7.0 (the pH is critical; see Note 1) is heated in a water bath
in a Coplin jar, with times and temperatures as follows (see Note 3):
a. Paraffin-embedded sections: 8–12 min, 85°C; the section of lymph node was denatured
for 10 min, although this step is not in the probe manufacturer’s directions.
b. Touch preps, smears, chromosome spreads: 2 min, 72°C; the touch prep of spleen was
denatured for 2 min.
2. Immediately place the slide in cold (−20°C) 70% ethanol to stop the denaturation and prevent
renaturation; after 2 min, (see Note 4), transfer to 90, 95, and 100% ethanol, 2 min each,
and allow to air-dry. Serial dehydration is crucial for paraffin sections; 70% followed by
absolute ethanol is adequate for thin smears, chromosome preparations, and so on.
Hybridization
Both the paraffin section of lymph node and the touch prep of spleen were denatured
and dehydrated as outlined above. For the BCR/ABL probe, skip to step 3; for the
chromosome enumeration probe, skip to step 4.
1. Some probes require dilution into hybridization buffer. This buffer contains 70% formamide,
some kind of carrier DNA, buffers, and dextran sulfate (to keep the solution in place on the
slide). The last ingredient makes it difficult to pipet when cold. Bring the stock solution of
the probe to room temperature, briefly vortex, and centrifuge in a microcentrifuge for 2–3
s to collect the contents at the bottom of the tube. If the probe is labeled with a fluorochrome,
it may be useful to work in subdued light at this point (see Note 5). If the probe has passed
its expiration date, altering the dilution is often useful (see Note 6).
2. Some probes require pre-denaturation and/or preannealing; the manufacturer will state this
in the package insert. An aliquot is removed to a microcentrifuge tube and heated to 72°C
for 5–10 min. It is then placed on ice for a few minutes (for α-satellites, to prevent reannealing)
or at 37°C for 30 min (for paints, to allow the repeated sequences to preanneal and prevent
their nonspecific binding to all chromosomes).
3. The BCR/ABL probe does not require dilution but must be denatured. Using a micropipet,
approximately 10 µL of probe per slide is pipetted into a microcentrifuge tube and denatured
in a water bath at 72°C for 5 min. Place the probe on the surface of the section or smear
and cover with a 22-mm2 cover slip. Seal the edges of the cover slip with rubber cement.
Do not try to be neat: you have to peel the cement off again.
FISH 87
4. The α-satellite probe used on the section of lymph node is usually denatured by a co-
denaturation step, per the manufacturer’s instructions. Co-denaturation is a procedure by
which the probe is placed on the slide, and it and the target DNA are denatured simultaneously
and then allowed to hybridize. This technique can be employed for many probes but must
be used with a very nice collection of probes supplied affixed to cover slips (CytoCell, UK).
Notes 5–7 are applicable. For paraffin sections, place 10 µL of hybridization buffer on the
section, place the cover slip with the attached probe on the slide, and seal with rubber cement.
Allow the rubber cement to dry for 5 min, and then place the slide on a 75°C hotplate for
5 min to denature the section.
5. Place the slides in the hybridization chamber, and incubate at 37°C for 30 min to overnight.
For paraffin sections and new probes, it is useful to start with the longer incubation and
decrease the time on subsequent trials. In these examples, the probe is incubated for 16 h.
Posthybridization Wash
1. Peel the rubber cement from the cover slip; often the cover slip comes off, too. If it does
not, “swish” the slide in a beaker of PBD to float the cover slip off.
2. Immediately place slide in one of the following solutions:
a. 50% formamide, 37°C, 15 min (paraffin sections, irrespective of probe, paints, unique-
sequence probes).
b. 65% formamide, 43°C, 15 minutes (alpha and beta satellite probes) or
c. 0.4× SSC, 72°C, 2 minutes, for directly labeled probes; this is the rapid wash technique,
see Note 8.
For both BCR/ABL and the α-satellite used in this example, the rapid wash was used.
3. Following a formamide wash, place slides in 2× SSC, 37°C, 5 min; for the rapid SSC wash,
follow with 30 s in PBD. Hybridized slides can be held in PBD at 4°C overnight prior to
the next step.
4. Score a minimum of 250–500 nuclei for so-called interphase cytogenetics. This is particularly
necessary for tissue sections, as the entire nucleus may not be in the section.
5. Cells labeled with a two-color probe should have at least one signal present in the nucleus
in each color to demonstrate that the probes worked and contained the appropriate labels.
6. Slides may be stored either before or after viewing (see Note 11).
7. If results are unsatisfactory, many specimens can successfully be destained (see Notes 12
and 13) and/or rehybridized (see Note 14).
NOTES
1. Formamide becomes basic as it degrades; adjust the pH of the solution with HCl. Deionizing
the formamide and freezing it in polypropylene tubes will increase the stability. Ultra-pure
formamide (BRL, Bethesda, MD), stored frozen, will maintain its pH for years.
2. To determine whether the digestion is adequate, cover-slip the wet slide and view with a
fluorescent microscope with a fluorescein isothiocyanate (FITC) filter. If there is green
fluorescence, more digestion time is needed. If distinct nuclei are not seen, the tissue
is overdigested.
3. For specimens to be denatured in formamide, it is useful to prewarm the slides on a slide
warmer prior to placing them in the formamide. Denature no more than four slides at a
time to maintain the temperature of the formamide, and recheck the temperature between
batches of slides.
4. For most specimens, there is a tradeoff between denaturation sufficient to allow the probe
to bind and over-denaturation so that morphology is lost. In general, do not exceed 2 min
unless the specimen is paraffin embedded. The exception is older specimens, which can be
denatured for 21⁄2–3 min.
5. For most commercial probes, subdued light is more than adequate. When you are working
with only a few slides, and the fluorochrome is not directly conjugated, the relatively brief
exposure to room light does not appear to be a problem.
6. Probes that are 6 months to a year or more past their stated expiration date have been used
successfully by using a higher ratio of probe to diluent (twice the recommended amount
of probe works well). If the probe does not require dilution, a larger volume (20 µL rather
than 10 for a 22-mm2 area) is often successful.
7. At least one vendor (CytoCell, UK) supplies their probes bound to cover slips. These can
be used quite successfully. After placing the cover slip over the hybridization solution on
the target area of the slide and sealing with rubber cement, it is important to wait at least
5 min before doing the denaturation step, as the probe requires this time to dissolve off
the cover slip.
8. A variety of posthybridization washes are described. For probes that are conjugated to a
fluorochrome, a “quick wash” is preferable.
9. Some manufacturers of fluorochrome conjugates sell kits that include a variety of “blocking
reagents,” designed to reduce nonspecific binding of avidin to the tissue, buffers for washing,
and polypropylene cover slips for use during incubation of the conjugate. “Plastic” cover
slips can be cut from Parafilm, which will reduce the scratching of the specimen during
this incubation. A 22 × 50-mm2 surface area is preferable for this step.
10. Two-color indirect immunofluorescence can be completed by adding both fluorochrome-
labeled antibodies simultaneously. In some cases, adding separate aliquots of the fluoro-
chromes effectively dilutes each in half; in other cases it may be necessary to do separate
incubations, add marked excess of antibody solution, and increase the incubation time or
increase the concentration of each one.
11. Most commercial probes are designed for viewing with a 100-W rather than a 50-W mercury
vapor lamp or equivalent. Signals may not be visible at 50 W. A 100-W bulb cannot be
put in a 50-W lamp socket.
FISH 89
12. Slides can be stored in the dark in the refrigerator for 3–7 days. Some slides may be usable
for several months if stored in a freezer. In either case, wipe the oil off the cover slip
before storing, as it may be necessary to remount the cover slip prior to further viewing.
13. If the signal from a probe is faint, sometimes it can be improved by removing some of the
counterstain. Propidium iodide is readily removed by a brief rinse in PBD. DAPI is much
more difficult to remove; the slides will have to be processed through two or three 5-min
washes in PBD, preferably in the dark.
14. Most specimens can be rehybridized or reprobed at least once. Wipe any oil from the cover
slip, and then remove the cover slip, either by gently prying it off with a razor blade or
by soaking in PBD. Rinse in fresh PBD and dehydrate. Any probe bound to the slide will
be removed during denaturation.
6
Practical Applications of the FISH Technique
INTRODUCTION
Fluorescence in situ hybridization (FISH), often referred to as molecular cytogenetics,
is currently recognized as a reliable technique and an alternative to classic cytogenetics
(karyotyping), with many applications for detecting chromosomal abnormalities. The
technique can analyze interphase nuclei of resting cells or metaphase spreads of mitotic
cells. In addition, specific protocols for karyotyping analysis combine FISH with clas-
sic cytogenetics.
FISH was introduced in 1988 by Pinkel et al. (1), who utilized a nonisotopic,
fluorescence technique for cytogenetics. However, the credit for initial application of
interphase cytogenetics is given to Lichter et al. (2), who in 1988 applied a radioactive
DNA probe to demonstrate chromosomal aberrations in nonmitotic human nuclei.
FISH TECHNIQUE
The main principle of FISH is the formation of the hybridization product of a DNA
probe with the target chromosomal DNA (Fig. 1). Briefly, a denatured fluorescent-
labeled DNA probe is added to the denatured specimen DNA and used to perform
hybridization. Posthybridization washings, under specific stringency conditions, remove
the unbound probe. After nuclear counterstaining, the specimen is ready for study with
a fluorescent microscope equipped with appropriate excitation filters and a photographic
camera. Digital image analysis systems equipped with appropriate software can enhance
the fluorescent signals.
Metaphase FISH is ideal for detecting locus-specific gene rearrangements and for
generating high-resolution physical genetic maps (3). However, it is often difficult to
obtain metaphase spreads from solid tumors. Selective cell growth of nontumorous
cells (lymphocytes, fibroblasts), which may result in loss of overall genetic information
and lead to false results, represents the main disadvantage of this technique (4,5). In
addition, metaphase FISH requires specific equipment and expertise in tissue cultures.
Interphase FISH is useful in detecting normal and defective genes such as deletions
or gains. The technique is ideal for analyzing solid tumors, particularly those with a
low proliferation rate. Nuclei from touch imprints, specimens of fine needle aspiration
91
92 Kontogeorgos, Kapranos, and Thodou
biopsies, preparations of centrifuged biologic fluids, and nuclei isolated from frozen
or paraffin-embedded tissues are appropriate for interphase FISH analysis (6–11). The
technique can also be applied directly to frozen or paraffin sections; however, cautious
interpretation of the results is required due to partial loss of nuclear DNA mass during
tissue sectioning (7–14).
Table 1
Selection of Filters for Optimal Fluorescent Microscopy, According to Detection
Systems, Fluorochromes, and Counterstains
Filter needed to Filter needed to
visualize counter- visualize probe
stain excitation/ signal excitation/
Probe label Detection reagent Counterstain emission (nm) emission
Digoxigenin FITC/anti-digoxigenin Propidium iodide (PI) 520/610 490/525 (long-pass)
Rhodamine/ DAPI 365/480 540/550
anti-digoxigenin
Biotin FITC-avidin PI 520/610 490/525 (long pass)
Texas Red-avidin DAPI 365/480 540/525
Digoxigenin Dual-color detection DAPI 365/480 Triple-bandpass
and biotin
From: Workshop Manual. Introduction to Molecular Cytogenetics: Tissue in situ hybridization. Oncor,
Gaithersburg, MD, pp. 18, 1995.
of light to pass through, are used for single or simultaneous dual-signal fluorescence
detection (Table 1). Optimal use of filters can subject the fluorescent signal to the
lowest amount of excitation light to retain fluorescence as much as possible. The
microscope should be equipped with a photographic camera. A 400-ASA film is appro-
priate, although 100-ASA films provide higher resolution. Alternatively, a digital,
preferably cool, camera allows direct transfer of the digitized signals to a computer
for storage and further processing using proper software (16).
PRACTICAL APPLICATIONS
Detection of Numerical Aberrations
α-Satellite centromeric, β-satellite pericentromeric, or classical satellite probes, spe-
cific for tandem repeats of DNA sequences of individual chromosomes, are widely
used, mostly with interphase FISH, to detect copy number aberrations of chromosomes
(3,22) (Fig. 3). Alternatively, paint (coatsome) probes, complementary to hundreds of
䉴
(See Color Plate 4 following page 208.)
Fig. 3. (A) Examples of numerical chromosome aberrations using an α-satellite, centromere-
specific DNA probe. The nucleus in the left panel shows two fluorescent signals corresponding
to a pair of normal chromosome 11, whereas more than two signals, as shown in the nucleus
of the right panel are characteristic for polysomies (FITC/PI, 400×). (B) Monosomy of chromo-
some 11 as the dominant abnormality in a mixed somatotroph-lactotroph, mostly prolactin-
producing pituitary adenoma (FITC/PI). (C) Chromosome 1 in ductal carcinoma of the breast.
Most cell exhibit three or more fluorescent signals (FITC/PI). (D) Chromosome 1 breast carci-
noma of the ductal type. All cells display aberrant fluorescent signals (FITC/PI). (E) Localization
of the DiGeorge gene on a chromosome in the 22q11 region in normal metaphase spreads. A
pair of fluorescent signals corresponds to the alleles of the DiGeorge gene. The third signal
marks the centromeric region of chromosome 22 (FITC/PI). (F) Dual-color FISH in metaphases
of normal lymphocytes demonstrates the bcr gene on chromosome 22 (green) and the abl gene
on chromosome 9 (red) (FITC/TR/DAPI). (G) Mixture of whole paint (coatsome) DNA probe
for chromosome 11 (red) and α-satellite centromere-specific probe for chromosome 2 (green)
in normal metaphase spreads (FITC/TR/DAPI). (H) All human telomeres demonstrate all chro-
mosomes in metaphase spreads (FITC/PI, original magnification ×400 for A and B; ×160 for
C and D; ×600 for E–G; ×720 for H.) E–H: from personal experiments (G.K., H.T.) in the
Appligene Oncor Labs, Illkirch, France.
Practical Applications of FISH 95
96 Kontogeorgos, Kapranos, and Thodou
Fig. 4. (A) Uses of centromeric and locus-specific probes. (B) Use of locus-specific probes
to detect structural aberrations.
unique DNA sequences specific to part or the entire length of the chromosome, can
be used for metaphase FISH analysis (3,5) (Fig. 4). Probes to the total human genome
can hybridize the entire chromosome and distinguish it from somatic cell hybrids
(18). Lastly, probes complementary to the highly conserved repetitive telomeric DNA
sequence (TTAGGG)n can also demonstrate all human chromosomes in metaphase
spreads and differentiate them from other species (23) (Fig. 4).
Aneuploidy in Solid Tumors
Interphase FISH utilizing centromere-specific DNA probes discloses numerical aber-
rations and thus provides information on tumor DNA ploidy. Normal diploid cells
display two centromeres for each pair of chromosome. The presence of one centromere,
corresponding to loss of one chromosome copy, is characteristic for monosomy, whereas
aneuploid cells with polysomies display more than two centromeres, corresponding to
extra chromosome copies (Fig. 3).
Numerical aberrations in some solid tumors (such as of chromosomes 1 and 17 in
breast cancer, 7 and 12 in follicular thyroid tumors, and of chromosome 8 in prostate
carcinomas and 11 in pituitary adenomas) are frequent and in most cases are correlated
with the DNA index. Thus, FISH is useful to estimate aneuploidy in these tumors (14,24–
27).
Aneuploidy in Prenatal Diagnosis
FISH has many applications for prenatal, and postnatal diagnosis and for the detection
of preimplantation genetic abnormalities. Specific DNA probes can rapidly detect
Practical Applications of FISH 97
c-erbB-2, was introduced for clinical use, particularly for metastatic breast cancer
refractory to chemotherapy (37). Therefore, c-erbB-2 analysis became of great impor-
tance for prognosis and treatment. Interphase FISH is ideal for detection and enumeration
of gene amplification (35). A commercially provided kit for FISH analysis approved
by the Food and Drug Administration is now available. A centromeric probe specific
for chromosome 17 and a sequence-specific probe for c-erbB-2 are necessary for FISH
analysis. A scoring system utilized with FISH is based on the ratio of gene copy to
centromeric copy number. Controls from cell lines showing a normal and amplified
c-erbB-2 gene should be used for testing the technique and for counting the fluorescent
signals. The expected normal ratio is 1.0; when the ratio grater than 2.0, the gene is
considered amplified. Although overexpression of c-erbB-2 protein can be detected by
immunohistochemistry, FISH analysis is preferable, particularly for tumors showing a
moderate degree of typical membranous immunostaining (38).
Androgen Receptor Gene Amplification
Amplification of the androgen receptor (AR) gene located on chromosome Xq11-
q13 may occur during androgen deprivation therapy of recurrent carcinomas from
patients who failed androgen endocrine treatment (39). Untreated prostate cancers show
no AR gene amplification (40). High levels of AR amplification are demonstrated in
specimens from recurrent tumors, but in none of those taken from the same patients
before therapy (41). Clinicopathologic studies indicate that AR amplification occurs
mostly in tumors that initially showed good response for more than 12 months following
endocrine therapy (39).
DiGeorge’s Syndrome
DiGeorge’s syndrome represents a developmental defect of the third and fourth
pharyngeal pouches associated with and characteristic of facial dysmorphism, congenital
heart defect, and hypoplasia or aplasia of the thymus and parathyroids. The DiGeorge
gene is located on chromosome 22q11. FISH and DNA dosage analysis showed deletion
of the DiGeorge gene region in all 16 cases studied (42). In another investigation of
23 patients with DiGeorge’s syndrome by prometaphase chromosome analysis and/or
by FISH, four patients displayed an interstitial deletion in band 22q11.2, whereas in
18 patients, the deletion was disclosed only by the FISH technique (43). Detection of the
DiGeorge defect gene is very important in diagnosis, prognosis, and genetic counseling.
Kallmann’s Syndrome
Kallmann’s syndrome is an inherent disorder characterized by hypogonadotropic
eunuchoidism, anosmia, and hyposomia. The affected individuals may also have border-
line normal intelligence, cleft lip and palate, deafness, and renal and cardiac anomalies.
Males are predominately affected. The patients show microdeletion of the KAL gene
located in the Xp22.3 region. FISH is a reliable technique for detecting the KAL-1
deletion in people with normal karyotypes but features consistent with Kallmann’s
syndrome (44).
Ewing’s Sarcoma/Primitive Neuroectodermal Tumors
The balanced translocation t(11;22)(q24;q12) is specific for Ewing’s sarcoma/primi-
tive neuroectodermal tumors. This translocation results in the fusion of the EWS gene
Practical Applications of FISH 99
ACKNOWLEDGMENTS
This work was supported in part by the E-52/96 Award of the Medical Council of
Greece (to Dr. Kontogeorgos).
REFERENCES
1. Pinkel, D., Landegent, J., Collins, C., et al. (1988) Fluorescence in situ hybridization
with human chromosome-specific libraries: detection of trisomy 21 and translocations of
chromosome 4. Proc. Natl. Acad. Sci. USA 85, 9138–9142.
2. Lichter, P., Cremer, T., Borden, J., Manuelidis, L., and Ward, D. C. (1988) Delineation of
individual human chromosomes in metaphase and interphase cells in situ suppression
hybridization using recombinant DNA libraries. Hum. Genet. 8, 235–246.
3. Meyne, J. and Moyzis, R. K. (1994) In situ hybridization protocols, in Methods in Molecular
Biology, vol. 33 (Choo, K. H. A., ed.), Humana, Totowa, NJ, pp. 63–74.
4. Wolman, S. R. (1997) Applications of fluorescence situ hybridization techniques in cytopa-
thology. Cancer 81, 193–197.
5. Cowan, J. M. (1994) Fishing for chromosomes. The art and its applications. Diagn. Mol.
Pathol. 3, 224–226.
6. Cajulis, R. S. and Frias-Hidvegi, D. (1993) Detection of numerical chromosomal abnormali-
ties in malignant cells in fine needle aspirates by fluorescence in situ hybridization of
interphase cell nuclei with chromosome-specific probes. Acta Cytol. 37, 391–396.
7. Kim, S. Y., Lee, J. S., Ro, J. Y., Gay, M. L., Hong, W. K., and Hittelman, W. N.
(1993) Interphase cytogenetics in paraffin sections of lung tumors by non-isotopic in situ
hybridization. Mapping genotype/phenotype heterogeneity. Am. J. Pathol. 142, 307–317.
8. Zitzelsberger, H., Szucs, S., Weier, H. U., et al. (1994) Numerical abnormalities of chromo-
some 7 in human prostate cancer detected by fluorescence in situ hybridization (FISH)
on paraffin-embedded tissue sections with centromere-specific DNA probes. J. Pathol.
172, 325–335.
9. Xiao, S., Renshaw, A., Cibas, E. S., Hudson, T. J., and Fletcher, J. A. (1995) Novel
fluorescence in situ hybridization approaches in solid tumors. Characterization of frozen
specimens, touch preparations, and cytological preparations. Am. J. Pathol. 147, 896–904.
10. McManus, D. T., Patterson, A. H., Maxwell, P., et al. (1999) Interphase cytogenetics of
chromosomes 11 and 17 in fine needle aspirates of breast cancer. Hum. Pathol. 30, 137–144.
11. Kontogeorgos, G. (2000) The art and applications of FISHing in endocrine pathology.
Endocr. Pathol. 11, 123–136.
12. Hopman, A. H. N., van Hooren, E., van de Kaa, C. A., Vooijs, P. G. P., and Ramaekers,
F. C. S. (1991) Detection of numerical chromosome aberrations using in situ hybridization
in paraffin sections of routinely processed bladder cancers. Mod. Pathol. 4, 503–513.
13. Blanco, R., Lyda, M., Davis, B., Kraus, M., and Fenoglio-Preiser, C. (1999) Trisomy 3 in
gastric lymphomas of extranodal marginal zone B-cell (mucosa-associated lymphoid tissue)
origin demonstrated by FISH in intact paraffin tissue sections. Hum. Pathol. 30, 706–7112.
14. Kontogeorgos, G., Kapranos, N., Kokka, E., Orphanidis, G., and Rologis, D. (1999) Molecu-
lar cytogenetics of chromosome 11 in pituitary adenomas: a comparison of fluorescence in
situ hybridization and DNA ploidy study. Hum. Pathol. 30, 1377–1382.
15. Kapranos, N., Kontogeorgos, G., Frangia, K., and Kokka, E. (1997) Effect of fixation on
100 Kontogeorgos, Kapranos, and Thodou
PROTOCOLS
Caution: Colcemid is a toxic chemical and should be handled with care. Avoid skin or
eye contact and inspiration.
5. PHA solution
a. Phytohemagglutinin is supplied lyophilized; it should be dissolved in the appropriate
amount of sterile distilled water according to the supplier.
b. The solution can be kept in frozen aliquots for several weeks. Defrost an aliquot just
before use.
c. Do not leave it at room temperature for long.
*It is important to get the cell pellet completely broken up. However, any harsh treatment
of the cells at this stage may rupture them, resulting in incomplete metaphases on the
slides. It is better to avoid passing cell suspension through narrow spaces such as pipets.
104 Protocols
*Alternatively, slide cleaning can be done by passing through a series of 100, 95, 70,
and 50% ethanol and then in running tap water for 1 h. Air-dry in the hood before use.
of mitoses present. If cell density is too high, add a few drops of the fixative. If cell density
is too low, centrifuge and resuspend in less fixative.
10. If pleased with the result, make more slides in the same way, and let them air-dry.
11. Dehydrate slides in a series of 70, 95, and 100% ethanol, 5 min in each, at room temperature.
Allow them to air-dry.
12. Store slides in a slide box for 2 weeks at room temperature before hybridization.***
*The final volume of fixative will depend on the size of the cell pellet and may need
to be adjusted if metaphases are too sparse or many. It is advisable to try a volume
of no more than 1 mL.
**Alternatively, using the same glass Pasteur pipet, drop 1–3 drops of the chromo-
some preparation on different points across the slide from above. Good results are
obtained when breathing on the slide until condensation forms and then dropping the
cell suspension from a distance of at least 50 cm.
***If slides are to be used for hybridization at a time shorter than 2 weeks they
have to be pretreated in 2× SSC before dehydration in ethanols. This pretreatment
artificially ages the chromosomes, making them less sensitive to over-denaturation.
Prewarm 40 mL of 2× SSC, pH 7.0, in a Coplin jar to 37°C in a water bath. Place
prepared metaphase slides in the jar and incubate for 30 min.
the first method, adjust the temperature of water in the staining dish with the slides,
to warmer if drying time is too long or colder if drying time is very short.
*Insulin and glutathione use is optional. They are believed to propagate cell growth.
2.3. SPECIMEN
3.2.4. Hybridization
1. Apply probe-hybridization mixture on each specimen (20 µL is adequate for an area under
a 22 × 22-mm cover slip).
2. Cover with a glass cover slip. Seal the edges with cover slip sealant if necessary.
3. Place slides within a prewarmed humidified chamber and incubate specimens at 37°C. Thirty
minutes of hybridization time is adequate for satellite probes, but overnight hybridization is
preferred for all other probes.
Hybridization temperature should be kept strictly at 37°C to avoid cross-hybridization to
other satellite sequences (<37°C) or decline of signal intensity (>37°C).
Table 2
FISH Nuclear Counterstain According to Probe Labeling (Oncor)
Probe label Counterstain (g/mL)
Fluorescein Propidium iodide (0.3)
Texas red DAPI (0.1)
Fluorescein and Texas red DAPI (0.1)
3.2.9. Counterstain
Apply to the specimen 10–15 µL of the appropriate counterstain (Table 2), and
cover with a glass cover slip.
4.3.2. Hybridization
Place 10–20 µL of probe on each slide (depending on tissue size) and cover with
a glass cover slip.
1. Denature probe and tissue DNA simultaneously by placing the slides on a hot plate or oven
at 67–69°C for 5 min.
2. Incubate slides in a humidified chamber at 37°C for 2–16 h (satellite probes) or 16 h (unique-
sequence probes).
INTRODUCTION
The relative insensitivity of in situ hybridization (ISH) has been cited as an important
limiting factor in the potential applications of this technique (1). Therefore, a major
initiative on the part of researchers who use molecular biologic techniques is the
improvement of sensitivity of ISH. There are two means of achieving this: the amplifica-
tion of target sequences and the amplification of signal (2). The former encompasses
such techniques as in situ polymerase chain reaction (PCR) and in situ reverse transcript-
ase (RT)-PCR, and is not discussed in this review. The latter includes catalyzed reporter
deposition, a major component of which is the use of labeled tyramide as a means of
signal amplification. This technique, which has subsequently been adapted for use in
ISH from its initial immunohistochemical applications, is a powerful tool for increasing
sensitivity in molecular biologic testing. Through the use of tyramide signal amplifica-
tionTM (TSATM), the theoretical limits of ISH can be approached.
As is the case with other highly sensitive assay techniques, the increased sensitivity
of TSA has inherent problems and limitations. With increased signal comes increased
noise, which creates inherent difficulties. Furthermore, the technique is in a continuing
state of evolution, so any review of the topic finds itself almost immediately behind
the times as new protocols are continuously developed. Because of the rapid and
ongoing development in this exciting area, the dissemination of new protocols via
electronic means, e.g., the Internet, plays an important role. It is highly recommended
that the interested reader consult the website maintained by Dr. Gerhard Hacker,
http://www.sbg.ac.at/kgg/protocols, for information on new protocols as they arise.
The purpose of this chapter is to discuss TSA as applied to ISH. After a brief
discussion of the historical development of TSA, a summary of the theoretical basis
of TSA and a partial description of protocols are given. The limitations of TSA, as
well as a discussion of practical troubleshooting techniques, follows.
HISTORICAL BACKGROUND
TSA (NEN Life Science, Boston, MA) is a proprietary term protected by U.S. patents
5,731,158, 5,583,001, and 5,196,306 and foreign equivalents for catalyzed reporter
113
114 Frater and Tubbs
deposition (CARD) used to amplify an ISH signal. CARD was originally described in
1989 by Bobrow et al. (3) as a novel amplification method for solid phase immunoassays.
Amplification is achieved by the deposition of biotin-labeled phenols at the site of
interest in a reaction catalyzed by the analyte-dependent reporter enzyme horseradish
peroxidase. Bobrow et al. (4–6) subsequently modified this technique for use in other
bioassays. Although the term CARD strictly refers to the technique using any source
of biotin-labeled phenols, this particular CARD technique uses tyramide as the source
of the phenol moiety. Therefore, for practical purposes, all CARD techniques currently
use TSA.
The subsequent modification of TSA for use in fluorescence in situ hybridization
(FISH) was reported by Raap et al. in 1995 (7). TSA was described by Raap as a two-
layered antibody technique, the first layer being a peroxidase-conjugated anti-hapten
immunoglobulin/streptavidin. The peroxidase substrate was tyramide. The second layer
consisted of (strept)avidin conjugates and was used to detect the biotin molecules
deposited in the first layer. The authors used several different tyramide conjugates, which
reacted with the peroxidase, resulting in the creation of a highly reactive intermediate
substance that aided and enhanced the localization of fluorochromes. Modifications of
this original protocol provide the basis for all CARD methods using TSA. Other authors
have described ELISA-based (4–6) and immunohistochemical methods (8) using TSA.
THEORY
All protocols currently use the CARD method initially proposed by Bobrow and the
TSATM modification described by Raap et al. in 1995 (3,7). The exact mechanism of
TSA is not entirely known, but it appears to involve the following general principles,
described in significant detail in a number of cogent reviews (2,9,10). The object of
TSA is to increase the sensitivity of ISH by an increase in detectable signal.
As mentioned in the previous section, TSA is a modification of Bobrow’s CARD
technique. Briefly, the TSA process consists of the addition of a fluorochrome- or
hapten-conjugated tyramide to a region of RNA or DNA altered by a horseradish
peroxidase-catalyzed chemical reaction. The tyramide intermediate compound is highly
reactive and binds to certain amino acids lying in close proximity to the horseradish
peroxidase. A basic understanding of the chemistry of the amino acid tyramine and its
derivatives is helpful to understand this pivotal step in the TSA process. From the
standpoint of TSA, the most important part of the tyramine ([4-(2-aminoethyl) phenol])
molecule is its phenol group (Fig. 1). It is thought that horseradish peroxidase acts on
this site to create an unstable, highly reactive intermediate that binds the amino acid
binding sites. The tyramides deposited in this fashion are then detected using fluores-
cence or a chromogenic method if a hapten-conjugated tyramide has been used (Fig. 2).
The end result of this process is an improvement in signal detection and hence
increased sensitivity using TSA. Macechko et al. (11) demonstrated this, comparing
immunologic amplification and TSA using a 40-kb cosmid probe and a 15-kb phage
probe. A single fluorochrome-conjugated antibody layer was only weakly detected,
whereas subsequent use of a second unlabeled secondary antibody and a third layer
consisting of an antibody conjugated to the same fluorochrome resulted in a significantly
enhanced signal. However, after a 1:10 dilution of the probes, the immunologically
Tyramide Amplification Methods 115
Fig. 1. The tyramine ([4-(2-aminoethyl) phenol]) molecule. It is postulated that the highly
reactive phenol moiety (shaded region) of tyramide intermediates generated by the tyramide
signal amplification process binds to amino acids in close proximity to the horseradish peroxidase
molecules (see text).
amplified signal was lost, whereas it was still detectable using TSA. Thus it is possible to
detect smaller targets and use less probe reagent using TSA compared with immunologic
amplification methods (11).
In addition to comparing the sensitivities of ISH with and without TSA, several
authors have shown that different tyramide conjugates confer different sensitivities
using ISH. The original TSA report of Raap et al. (7) described the use of four different
tyramine conjugates: biotin, fluorescein, tetramethylrhodamine, and aminomethylcou-
marin acetic acid. In 1997 van Gijlswijk et al. (12) compared the performance of several
different tyramide conjugates using a middle repetitive specific probe (28S rRNA) and
a single-copy gene-specific cosmid probe with a 30–40 kb target size. Nearly all
conjugates examined using these probes were easily visualized. The best performing
conjugates used BODIPY FL and rhodamine, although nearly all conjugates could be
visualized. Only two conjugates had suboptimal results (defined as decreased signal
compared with nonconjugated ISH), X-rhodamine and Texas Red. They noted that
these two compounds differed from the other conjugates in having an additional sulfonic
acid moiety, which conveyed an additional negative charge to these conjugates, possibly
impacting performance. Additionally, the authors demonstrated that TSA could be
adapted for multicolor FISH. The best results using multiple rounds of hybridization
alternating with tyramide detection were obtained using 7-hydroxycoumarin-, fluores-
cein-, and rhodamine-tyramide conjugates sequentially. Using this approach, intermit-
tent hydrogen peroxidase inactivation was unnecessary (12).
Current estimates of the limits of detection sensitivity with ISH using tissue sections
are approximately 10–20 mRNA/viral DNA copies per cell (13). TSA markedly
increases sensitivity, allowing detection of single RNA or DNA copies. However, the
technical limitations, including the need for relatively large probes and background
tissues, may impose practical constraints on the sensitivity of the method. However,
TSA does appear to improve ISH sensitivity substantially, and when performed by
individuals familiar with molecular techniques, it can approximate the theoretical claims
regarding improved sensitivity, including single-copy gene detection.
116 Frater and Tubbs
Tyramide Amplification Methods 117
the remaining peroxidase completely between steps. To achieve this, mild acid treatment
with 0.01 N HCl may be used (18,21,22). If this step is not performed, repeated
deposition of the tyramide conjugate at the prior target site could result.
Despite the limitations of TSA, it appears to be superior to techniques such as in
situ PCR, which has considerably greater practical limitations compared with TSA. In
situ PCR is an example of a target amplification procedure in which pretreated cells
are permeabilized to allow entrance of primers, nucleotides, and polymerase enzymes.
The amplified products are detected using ISH techniques. This technique has been
plagued with difficulties, including low amplification efficiency and poor reproducibility.
Because of these serious limitations, this technique has been of only limited use in
diagnostic pathology (9,17). Additionally, the TSA product can be detected using
techniques such as FISH and brightfield ISH (BRISH) (17).
POTENTIAL USES
A recent review (17) has summarized the potential uses of TSA. Because of the
markedly increased sensitivity that can be achieved using this technique, there are
numerous possibilities for TSA combined with ISH, including those listed below.
Other Uses
Although this chapter is limited to a discussion of ISH methods using TSA, it
should be noted that there are several other uses for TSA in diagnostic and
experimental pathology. Adapting biotinylated tyramides for use in immunohistochem-
istry, Adams (8) reported a 1000-fold increase in sensitivity in tissue compared
with specimens prepared by the conventional avidin-biotinylated complex technique.
Others have reported much more modest increases in sensitivity using TSA and
immunohistochemistry, on the order of 5–50% compared with conventionally treated
specimens (26,28–30). The interested reader is directed to a recent review of these
subjects (17).
120 Frater and Tubbs
Tyramide Amplification Methods 121
Fig. 3. Signal amplification achieved with CARD. (A) SiHa cells, known to contain 1–2
copies of HPV-16 per cell, conventional chromogenic ISH. HPV is undetected (DAPI/H2O2
detection). (B) SiHa cells, CARD ISH procedure; 1–2 copies of HPV identified (DAPI/H2O2
detection). (C) CaSki cells, conventional chromogenic ISH (DAPI/H2O2 detection). (D) CaSki
cells, CARD ISH. Numerous and in part confluent signals are identified (DAPI/H2O2 detection).
An interesting new development that may further increase TSA sensitivity involves
its use in conjunction with branched DNA (bDNA) ISH. bDNA, (also called oligodeoxy-
ribonucleotide) is another method whereby signal amplification may be improved.
bDNA has been tested most extensively in quantitative measurement of viral load in
chronic hepatitis in the solution phase. After nucleic acid denaturation and hybridization
of probes to target and solid phase performed in the usual fashion, the bDNA multimer
is then hybridized to the probe. After the incorporation of alkaline phosphatase molecules
into the bDNA, the system acts as a signal amplifier. This method has been successful
in detection of viral nuclei acid with a detection limit of 1–3 plaque-forming units
(pfu). This method has been reported to be highly reproducible and easy for experienced
lab personnel to perform (31–33). In combination with TSA, further increases in
sensitivity are theoretically possible. Because of the reported reproducibility and ease
of handling of bDNA, this may be feasible with no appreciable loss of specificity.
Flow cytometry is another modality that has been reported to benefit from TSA
technology. Kaplan and Smith (34) compared fluorescence intensity of conventionally
and TSA-treated samples from the Jurkat malignant T-cell line. They reported a 10–100-
fold increase in amplification using TSA for a number of antibodies (MHC class I, CD5,
CD3, CD4, CD6, CD7, CD34, CD45, MHC class II, Fas ligand, and phosphatidylserine).
TSA has been adapted for a number of sophisticated clinical and research uses,
including high-resolution mapping of DNA sequences, multiple color techniques for
simultaneous detection of multiple probes, and comparative genomic hybridization
(34a). Combined FISH/immunohistochemistry with TSA has been described, which
allows simultaneous detection of glial cell line-derived neurotropic factor 2 mRNA
and localization of its co-receptor Ret (35). Comparative genomic hybridization may
be particularly useful in the assessment of tumors. Another study, looking at the
122 Frater and Tubbs
REFERENCES
1. Tubbs, R. R., Pettay, J., Grogan, T., et al. (2000) Super sensitive in situ hybridization
by tyramide signal amplification (TSA) and nanogold-silver staining: the contribution
of autometallography and catalyzed reporter deposition (CARD) to the rejuvenation of
in situ hybridization, in Gold and Silver in Molecular Morphology, Eaton, Natick,
M. A., in press.
2. Komminoth, P. and Werner, M. (1997) Target and signal amplification: approaches to
increase the sensitivity of in situ hybridization. Histochem. Cell Biol. 108, 325–333.
3. Bobrow, M. N., Harris, T. D., Shaughnessy, K. J., and Litt, G. J. (1989) Catalyzed reporter
deposition, a novel method of signal amplification. Application to immunoassays. J. Immu-
nol. Methods 125, 279–285.
4. Bobrow, M. N., Shaughnessy, K. J., and Litt, G. J. (1991) Catalyzed reporter deposition,
a novel method of signal amplification. II. Application to membrane immunoassays. J.
Immunol. Methods VOL?, 103–112.
5. Bobrow, M. N., Litt, G. J., Shaughnessy, K. J., Mayer, P. C., and Conlon, J. (1992) The
use of catalyzed reporter deposition as a means of signal amplification in a variety of
formats. J. Immunol. Methods 150, 145–149.
6. Bobrow, M. N., Litt, G. J. (1993) Method for the detection or quantitation of an analyte using
an analyte dependent enzyme activation system. United States Patent Number 5196306.
7. Raap, A. K., van de Corput, M. P. C., Vervenne, R. A. W., van Gijlswijk, R. P. M., Tanke,
H. J., and Wiegant, J. (1995) Ultra-sensitive FISH using peroxidase-mediated deposition
of biotin- or fluorochrome tyramides. Hum. Mol. Genet. 4, 529–534.
8. Adams, J. C. (1992) Biotin amplification of biotin and horseradish peroxidase signals in
histochemical stains. J. Histochem. Cytochem. 40, 1457–1463.
9. Totos, G., Tbakhi, A., Hauser-Kronberger, C., Tubbs, R. R. (1997) Catalyzed reporter
deposition: a new era in molecular and immunomorphology—nanogold-silver staining and
colorometric detection and protocols. Cell Vision 4, 433–442.
10. Hopman, A. H. N., Ramaekers, F. C. S., Speel, E. J. M. (1998) Rapid synthesis of biotin-,
digoxigenin-, trinitrophenyl-, and fluorochrome-labeled tyramides and their application for
in situ hybridization using CARD amplification. J. Histochem. Cytochem. 46, 771–777.
11. Macechko, P. T., Krueger, L., Hirsch, B., Erlandsen, S. R. (1997) Comparison of immuno-
logic amplification vs. enzymatic deposition of fluorochrome-conjugated tyramide as detec-
tion systems for FISH. J. Histochem. Cytochem. 45, 359–363.
12. Van Gijlswijk, R. P. M., Zijlmans, H. J. M. A. A., et al. (1997) Fluorescence-labeled
tyramides: use in immunocytochemistry and fluorescence in situ hybridization. J. Histochem.
Cytochem. 45, 375–382.
13. Speel, E. J. M., Seremaslani, P., Roth, J., Hopman, A. H. N., Komminoth, P. (1998) Improved
mRNA in situ hybridization on formaldehyde-fixed and paraffin-embedded tissue using signal
amplification with different haptenized tyramides. Histochem. Cell Biol. 110, 571–577.
14. Kerstens, H. M. J., Poddighe, P. J., Hanselaar, G. J. M. (1995) A novel in situ hybridization
Tyramide Amplification Methods 123
30. Toda, Y., Kono, K., Abiru, H., et al. (1999) Applications of tyramide signal amplification
system to immunohistochemistry: a potent method to localize antigens that are not detectable
by ordinary method. Pathol. Int. 49, 479–483.
31. Hendricks, D. A., Stowe, B. J., Hoo, B. S., et al. (1995) Quantitation of HBV DNA in
human serum using a branched DNA (bDNA) signal amplification assay. Am. J. Clin.
Pathol. 104, 537–546.
32. Horn, T., Chang, C-A., and Urdea, M. S. (1997) Chemical synthesis and characterization
of branched oligodeoxyribonucleases (bDNA) for use as signal amplifiers in nucleic acid
quantification assays. Nucleic Acids Res. 25, 4842–4849.
33. Urdea, M. S., Wuestehube, L. J., Laurenson, P. M., and Wilber, J. C. (1997) Hepatitis C—
diagnosis and monitoring. Clin. Chem. 43, 1507–1511.
34. Kaplan, D. and Smith, D. (2000) Enzymatic amplification staining for flow cytometric
analysis of cell surface molecules. Cytometry 40, 81–85.
34a.Joos, S., Fink, T. M., Ratsch, A., Lichter, P. (1994) Mapping and chromosome analysis;
the potential of fluorescence in situ hybridization. J. Biotech. 35, 135–153.
35. Zaidi, A. U., Enomoto, H., Milbrandt, J., and Roth, K. A. (2000) Dual fluorescent in
situ hybridization and immunohistochemical detection with tyramide signal amplification.
J. Histochem. Cytochem. 48, 1369–1375.
35a.Buonamici, L., Serra, M., Losi, L., Eusebi, V. (2000) Application of CARD-ISH for
assessment of numerical chromosome aberrations in interphase nuclei of human tumor cells.
Int. J. Surg. Pathol. 8, 201–206.
36. Schöfer, C., Weipoltshammer, K., Almeder, M., and Wachtler, F. (1997) Signal amplification
at the ultrastructural level using biotinylated tyramides and immunogold detection. Histo-
chem. Cell Biol. 108, 313–319.
37. Hainfeld, J. F. (1987) A small gold-conjugated antibody label: improved resolution for
electron microscopy. Science 230, 450–453.
38. Van de Corput, M. P. C., Dirks, R. W., van Gijlswijk, R. P. M., and van de Rijke, F. M.
(1998) Fluorescence in situ hybridization using horseradish peroxidase-labeled oligodeoxy-
nucleotides and tyramide signal amplification for sensitive DNA and mRNA detection.
Histochem. Cell Biol. 110, 431–437.
39. Van de Corput, M. P. C., Dirks, R. W., van Gijlswijk, R. P. M., et al. (1998) Sensitive
mRNA detection by fluorescence in situ hybridization using horseradish peroxidase-labeled
oligodeoxynucleotides and tyramide signal amplification. J. Histochem. Cytochem. 46,
1249–1259.
40. King, G., Chambers, G., Murray, G. I. (1999) Detection of immunoglobulin light chain
mRNA by in situ hybridisation using biotinylated tyramine signal amplification. J. Clin.
Pathol. Mol. Pathol. 52, 47–51.
41. Yang, H., Wanner, I. B., Roper, S. D., and Chaudhari, N. (1999) An optimized method for
in situ hybridization with signal amplification that allows the detection of rare mRNAs.
J. Histochem. Cytochem. 47, 431–445.
42. Schmidt, B. F., Chao, J., Zhu, Z., DeBiasio, R. L., and Fisher, G. (1997) Signal amplification
in the detection of single-copy DNA and RNA by enzyme-catalyzed deposition (CARD)
of the novel fluorescent reporter substrate Cy3.29-tyramide. J. Histochem. Cytochem. 45,
365–373.
TSA 125
PROTOCOLS
INTRODUCTION
This is the first TSA protocol reported in the literature, described by Raap et al. in
1995 (7). It was subsequently refined and reported in greater detail by other authors
but is provided here to illustrate further the basic principles common to all TSA protocols.
ISH
ISH was performed by the usual method. Experiments were performed using DNA
with biotin- or digoxigenin-dUTP DNA ISH, using human chromosome spreads, and
RNA ISH with HeLa rat-9G cells and Plasmodium berghei.
INCUBATION
COMPLETION
The slides were then washed, incubated in streptavidin-FITC, washed, and mounted
in Vectashield (Vector), a commercially available antifading agent.
126 Protocols
INTRODUCTION
The following protocol is adapted from a recent review of the subject (1); it uses
CARD/TSA in combination with Nanogold, a technique successfully employed to detect
such minute targets as single human papillomavirus (HPV) DNA copies. Nanogold was
first reported by Hainfeld for use in electron microscopy and subsequently adapted for
this current use (9,23,37). An advantage of this technique is that it can be successfully
employed in formalin-fixed, paraffin-embedded tissue sections as well as formalin-
fixed cytologic specimens. As is true with the other protocols, the extreme sensitivity
of this method can cause problems with excessive background staining when issues
with fixation and probes arise.
SPECIMEN PREPARATION
Specimens are deparaffinized with fresh xylene washes (2 × 15 min each) followed
by rinsing in absolute EtOH (2 × 5 min). The specimens are then treated with a solution
of 3% hydrogen peroxide in methanol at room temperature for 30 min, followed by
rinsing in double-distilled (ultrapure) water for 10 s and phosphate-buffered saline
(PBS) for 3 min. After this, the specimens are incubated with 0.1 mg/mL proteinase
K (Boehringer Mannheim) in PBS at 37°C for approx 8 min. This last time may be
variable, and optimization is necessary. Microwave treatment at this point may be
helpful. The specimen is then washed twice with PBS, 3 min each, and then with
ultrapure water for 10 s. The specimens are then hydrated with graded EtOH (50, 70,
100%) for 5 min each, followed by air-drying.
HYBRIDIZATION
Specimens are first prehybridized with a 1:1 solution of deionized formamide and
20% dextran sulfate in 2× standard saline citrate (SSC) at 50°C for 5 min. After careful
shaking of the prehybridization block, 1 drop of biotinylated DNA probe is added on
the section, which is then covered with a cover slip. Entrapment of air bubbles under
the cover slip is to be carefully avoided. The DNA is then denatured by heating the
sections on heating blocks at 92–94°C for 8–10 min. The sections should then be
incubated in a moist chamber for at least 2 h. The sections are then washed with two
changes of 2× SSC for 5 min each, followed by washes with 0.5× SSC, 0.2× SSC,
and then distilled water, which should remove the cover slips and prepare the sections
for the next step.
water for 2 min. Drain the sections, wipe the parts of the slide around the section, and
apply PAP-pen (Dako) to the area of the slide adjacent to the sections. Then incubate
the sections with blocking solution (4× SSC containing 5% casein sodium salt or 0.5%
blocking powder [NEN Life Science]). Wash the sections in 4× SSC containing Tween-
20 (Boehringer Mannheim) for 2 min.
TSA
Incubate sections with streptavidin-biotin-peroxidase complex (Dako) at room tem-
perature for 30 min. Wash with three changer of 4× SSC containing 0.05% Tween-20
for 2 min each, followed by two changes of PBS for 2 min each. Then incubate the
sections with biotinylated tyramide at room temperature for 10 min (NEN kit) or 15 min
(Dako). Follow manufacturers’ recommendations for appropriate use of these products.
NANOGOLD
Immerse the sections in PBS containing 0.1% fish-gelatin for 5 min. Then incubate
the sections with streptavidin-Nanogold (Nanoprobes) diluted 1:750 in PBS containing
1% bovine serum albumin at room temperature for 60 min. Then wash in three changes
of PBS-gelatin for 5 min each. After this, wash the specimens in ultrapure (EM
grade) water.
AUTOMETALLOGRAPHY
Perform silver acetate autometallography or GoldEnhance development (Nano-
probes) according to the manufacturer’s instructions.
COMPLETION
After performing autometallography, counterstain sections with hematoxylin and
eosin or nuclear fast red, dehydrate, and mount using Permount or DPX. Avoid the
use of Eukitt for this purpose.
INTRODUCTION
Detection of cellular mRNA can be difficult for a number of reasons. First, compared
with DNA, RNA degrades much more rapidly, requiring prompt specimen treatment
for detection by FISH. Also, mRNA is present in relatively small quantities within the
cell compared with DNA. For this reason, low sensitivity has plagued many attempts
at RNA detection. Because of its greater sensitivity, FISH with radioactive probes has
been successful. However, the current attempts to minimize the use of radioisotopes
in the clinical laboratory setting has made this method problematic. An additional
128 Protocols
SPECIMEN PREPARATION
Two experimental protocols were described. In the first, cells from human urothelial
carcinoma and osteosarcoma cell lines were cultured on uncovered glass slides.
RNA-FISH
The target RNA was denatured prior to hybridization with 70% deionized formamide/
2× SCC. Following this, 10 µL of hybridization mixture was added, and a cover slip
was placed on each slide. Hybridization was performed in a moist chamber and lasted
30 min. Next, the cover slips were removed, and the cells were washed with 2× SSC
and 40% formamide (5 min) and 2× SSC. Then the cells were again washed three
times at 2 min each with 2× SSC and rinsed with TBS (100 mM Tris-HCl, 150 mM
NaCl) with 0.05% Tween-20. The hybridization product was detected using TSA with
reagent (NEN Life Science Products) with a 30-min incubation.
SIGNAL DETECTION
The slides were examined using a standard microscope with a single-bandpass filter
for fluorescein and Texas Red.
PROTOCOLS
8
Ultrastructural In Situ Hybridization
INTRODUCTION
In 1969, the in situ hybridization (ISH) technique was introduced for the detection
of ribosomal RNA gene products (1,2). Many investigators have since improved this
technique to identify specific genes or gene products in cells. The development of
synthetic oligonucleotide probes, which could be easily designed and produced, contrib-
uted greatly to the refinement of ISH (3). Nonradioactive synthesized oligonucleotide
probes labeled with biotin or digoxigenin were introduced for the detection of ISH
signals (4,10). ISH under a brightfield light microscope (LM-ISH) has since become
a widely used method for examining the tissue distribution and expression of mRNA.
The LM-ISH method is lacking in the resolution required for studies on the spatial
relationship between mRNA and the protein product. This type of information can only
be provided by ultrastructural studies.
Ultrastructural ISH with an electron microscope (EM-ISH) is a recently developed
technique and is essential for the intracellular identification of mRNA and to study the
role of mRNA in protein synthesis. Ultrastructural ISH for RNA was first described
by Jacob et al. (11). This method has been further developed by several investigators
(12–26). Three different approaches have been applied by investigators to EM-ISH
studies: the preembedding method (13,17,21–26), the nonembedding method using
ultrathin frozen sections (14,15), and the postembedding method (18–20,24,26).
Recently, we developed a nonradioisotopic EM-ISH method using biotinylated synthe-
sized oligonucleotide probes and applied it to the ultrastructural visualization of growth
hormone (GH) and prolactin (PRL) mRNAs and pathophysiological studies in rat
pituitary cells (24–26). In addition, we developed a combined EM-ISH and immunohis-
tochemistry method for the purpose of simultaneous identification of pituitary hormone
and its message in the same cell (27–33). In this chapter, the previously reported EM-
ISH techniques including our nonradioisotopic EM-ISH method are reviewed, along
with our combined immunohistochemistry and nonradioisotopic preembedding ISH
methods. We also present an overview on the methods reported in the literature.
129
130 Matsuno et al.
Fig. 1. Photograph of preembedding EM-ISH for rat GH mRNA. Rat GH mRNA is localized
diffusely on the polysomes of the entire rough endoplasmic reticula (RER). Bar = 200 nm.
(Reproduced with permission from ref. 30.)
(ABC-HRP), using Vectastain’s ABC kit (Vector, Burlingame, CA), and thereafter hybrid-
ization signals are developed with 0.017% H2O2 in DAB solution (complete Graham-
Karnovsky’s solution) for 7 min. At this stage, hybridization signals are confirmed light
microscopically.
11. 2% osmium tetroxide is applied to the sections for 1 h. After dehydration with graded
ethanol (50, 70, 80, 90, 95, 100%), tissue sections are embedded in Epon resin by the
inverted beam capsule method.
12. Polymerization of Epon resin at 60°C for 2 days. Ultrathin sections are inspected under
electron microscopy.
13. The negative control experiments should include hybridization with probes of sense or
scrambled sequences and without probes, as well as pretreatment with ribonuclease A (100
µg/mL) at 37°C for 45 min before hybridization.
Results
This preembedding EM-ISH study with an antisense biotinylated oligonucleotide
probe localized rat GH mRNA diffusely on the polysomes of the entire rough endoplas-
mic reticulum (RER) (Fig. 1). The preembedding EM-ISH study with a sense probe
for rat GH mRNA showed negative signals (Fig. 2).
132 Matsuno et al.
Fig. 2. Control study with a sense probe for rat GH mRNA. Rat GH mRNA is not localized
on the RER. Bar = 200 nm. (Reproduced with permission from ref. 30.)
Comments
In the preembedding method, both ultrastructure and mRNA are sufficiently pre-
served. The preembedding method is valuable in that hybridization signals can be
confirmed light microscopically. The only drawback of this preembedding method is
failure of quantification of hybridization signals that are detected enzymatically.
Pitfalls:
1. Prolonged fixation of the tissues will decrease the messages.
2. Excessive treatment with proteinase K will destroy the morphology.
Fig. 3. Hybridization signals for rat GH mRNA are localized on the polysomes of the RER
using 20 nm streptavidin gold. Bar = 100 nm. The hybridization signal intensity is lower than
that for the preembedding method. (Reproduced with permission from ref. 30.)
4. After hybridization, the grids are dipped in 2× SSC, 1× SSC, and then 0.5× SSC for 5
min, respectively.
5. The hybridization signals are developed for 30 min with 20 nm streptavidin gold (British
Biocell, Cardiff, UK) diluted 1:50 in 1% BSA-PBS. After being dipped in PBS and distilled
water and dried at room temperature, the grids are inspected under electron microscopy.
6. The control experiments are hybridizations with probes of sense or scrambled sequences
and without probes.
Results
The postembedding EM-ISH study localized hybridization signals for rat GH mRNA
on the polysomes of the RER using 20 nm streptavidin gold (Fig. 3). The EM-ISH
study with a sense probe for rat GH mRNA showed no hybridization signals (Fig. 4).
The hybridization signal intensity is lower than that for the preembedding method.
Comments
In the postembedding method using tissues embedded in LR White resin, the ultra-
structure is also sufficiently preserved. Compared with the preembedding method, the
postembedding method has several drawbacks: 1) message preservation is difficult
during polymerization of LR White resin at relatively high temperatures for extended
periods, which leads to mRNA degradation; and 2) nonspecific signals may be seen
134 Matsuno et al.
Fig. 4. EM-ISH with a sense probe for rat GH mRNA shows no hybridization signals. Bar =
100 nm. (Reproduced with permission from ref. 30.)
Fig. 5. Photograph of pituitary cells from female rats given estrogen for 7 weeks subjected
to preembedding electron microscopic ISH. Rat PRL mRNA is localized frequently on the
polysomes of the whirling RER. Bar = 200 nm. (Reproduced with permission from ref. 31.)
vesiculated, and partly dilated RER (Fig. 6). There were also increased numbers of
secretory granules, which resulted in increased PRL immunoreactivity induced by
bromocriptine.
Comments
For quantitative analyses of PRL mRNA expression, other experiments including
Northern blot hybridization and LM-ISH studies are required. Using the 32P-labeled
oligonucleotide probe for rat PRL mRNA, a 1.0-kb transcript was detected on the
nitrocellulose membrane, and the PRL mRNA and β-actin mRNA hybridization signal
density ratios for the pituitary glands of untreated, control female rats, those treated
with estrogen alone for 3 and 7 weeks, and those treated with estrogen plus bromocriptine
were evaluated densitometrically (26). As shown in our previous report, the hybridization
signal density of PRL mRNA was enhanced as the duration of estrogen treatment
increased, and it decreased markedly after bromocriptine adminsitration (26). Our
previous LM-ISH studies revealed that the hybridization signals of PRL mRNA from
estrogen-treated rats were more intense than those from normal rats without estrogen
administration; the hybridization signal frequency increased as the duration of estrogen
PROTOCOL: Ultrastructural In Situ Hybridization 137
COMBINED IMMUNOHISTOCHEMISTRY
AND PREEMBEDDING EM-ISH
As for mRNA preservation, the preembedding EM-ISH method using frozen sections
fixed in 4% paraformaldehyde has more advantages over the postembedding EM-
ISH method using tissues embedded in LR White resin. Frozen sections fixed in
4% paraformaldehyde have better morphologic preservation than immediately frozen
sections. Based on this assessment à propos of maintaining tissue morphology and
retaining the messages, we utilize the preembedding EM-ISH method using 6-µm-thick
frozen sections fixed in 4% paraformaldehyde for the simultaneous detection of mRNA
and encoded protein.
Method
Steps 1–11 are the same as described in the section on the preembedding EM-
ISH method.
Results
As shown in the previous section, EM-ISH with an antisense probe for rat GH
mRNA revealed diffuse localization on the polysomes of the entire RER. Subsequent
immunohistochemical staining using anti-rat GH antibody (monkey, diluted 1:100 with
PBS), supplied by the National Institute of Diabetes and Digestive and Kidney Diseases
(NIDDK, Bethesda, MD), and 20 nm protein A colloidal gold identified rat GH on the
secretory granules (Fig. 7). Colloidal gold signals for GH were distributed mainly on
the secretory granules and were also identified in the cisternae of the RER (Fig. 7).
The immunoreactivity retrieved after the etching process with 10% H2O2 was similar
and comparable to that with 4% sodium periodate. Immunohistochemical control experi-
ments substituting normal monkey serum for anti-GH antibody combined with the
preembedding EM-ISH method revealed no reactions of protein A colloidal gold parti-
cles on the secretory granules (Fig. 8).
Comments
The only flaw with this combined method is the deosmification and degradation
of the signals of mRNA caused by the etching process using H2O2 or sodium
PROTOCOL: Ultrastructural In Situ Hybridization 139
Fig. 8. Immunohistochemical control experiments at the ultrastructural level with the substitu-
tion of normal monkey serum for anti-rat GH antibody. No positive reactions of protein A
colloidal gold particles are observed on the secretory granules, whereas positive signals for rat
GH mRNA are identified as osmium black on the polysomes of the RER. Bar = 200 nm.
(Reproduced with permission from ref. 30.)
periodate. To solve this problem, we have recently used LR White resin for tissue
embedment (29).
Pitfalls:
1. Prolonged fixation of the tissues will decrease the messages.
2. Excessive treatment with proteinase K will destroy the morphology.
3. Excessive etching will shade off the reaction products (osmium black) of mRNA.
Fig. 9. Using LR White resin for tissue embedment, rat GH mRNA is also localized on the
polysomes of the entire RER. Subsequent immunohistochemical staining identifies rat GH
protein both on the secretory granules (arrow) and in the cisternae of the RER (arrowhead).
Bar = 200 nm. (Reproduced with permission from ref. 31.)
15. The grids are washed with 0.1 M PB, and the immunoreaction is visualized at room
temperature for 1 h with 20 nm protein A colloidal gold diluted 1:40 in 0.1 M PB.
16. After being washed with 0.1 M PB and distilled water, the grids are dried at room temperature
and observed under an electron microscopy.
17. The immunohistochemical negative control experiment is substitution of normal serum for
the antibody.
Results
EM-ISH using LR White resin for tissue embedding also localized rat GH mRNA
on the polysomes of the entire RER (Fig. 9). Subsequent immunohistochemical staining
identified rat GH protein both on the secretory granules and in the cisternae of the
RER (Fig. 9).
Comments
As shown in our previous reports (24,25,27,28), GH mRNA is distributed diffusely on
the RER. In these reports using tissues embedded in Epon resin, somewhat heterogeneous
electron density was observed in GH mRNA expression. In LR White resin-embedded
tissues, GH mRNA was also distributed diffusely on the RER, its electron density
PROTOCOL: Ultrastructural In Situ Hybridization 141
Significance
There are two major problems to be resolved in the ultrastructural double-staining
method to visualize mRNA and encoded protein simultaneously in the same cell. One
is to retain the messages and the other is to maintain the immunoreactivity of the
encoded protein in the same cell. Based on the aforementioned assessment à propos
of maintaining tissue morphology and retaining the messages, we employ the preembed-
ding EM-ISH method using 6-µm-thick frozen sections fixed in 4% paraformaldehyde
for the simultaneous detection of mRNA and encoded protein. The immunoreactivity
can be retrieved by the etching process with H2O2 or sodium periodate even after
modification such as osmification and embedding in Epon resin, and tissues embedded
in Epon resin can serve for simultaneous ultrastructural detection of messages and
encoded proteins. The only problem is the deosmification and degradation of the mRNA
signals, which are caused by etching with H2O2 or sodium periodate. To solve this
problem, we have recently used LR White resin for tissue embedding (29). In LR
White resin-embedded tissues, retrieval of immunoreactivity with H2O2 or sodium
periodate is not required, and therefore degradation of mRNA signals can be avoided.
To our knowledge, only five reports, except for our previous ones (27–33), describing
the simultaneous ultrastructural detection of mRNA and encoded protein have been
published, in each of which the postembedding EM-ISH method using colloidal gold
particle was utilized (16,35–38). However, in these postembedding EM-ISH studies,
the relatively frequent nonspecific reactions of colloidal gold particles used for the
detection of mRNA were observed in the cisternae of the RER. The EM-ISH method
for Lowicryl K4M-embedded tissues is generally supposed to be faulty in morphologic
preservation. Decreased message preservation was observed in postembedding EM-
ISH studies using LR White-embedded tissues (24,38). From the viewpoint of specific
hybridization signal detection and morphologic preservation, the combined immunohis-
tochemistry and preembedding EM-ISH method is considered superior to the previously
reported postembedding EM-ISH method and is successful and satisfactory for simulta-
neous ultrastructural identification of mRNA and encoded protein.
Translation of mRNA for secreted proteins is initiated on free ribosomes and then
translocated to the polysomes on the RER with the aid of signal recognition particles
once the signal peptide is produced. Synthesized proteins are secreted into the luminal
space of the RER and subsequently transported to secretory granules via the Golgi
142 Matsuno et al.
apparatus. This ultrastructural double-staining method for mRNA and encoded protein
can provide an important clue for elucidating the intracellular correlation of mRNA
translation and secretion of translated protein.
REFERENCES
1. Gall, J. G. and Pardue, M. L. (1969) Formation and detection of RNA-DNA hybrids in
cytological preparations. Proc. Natl. Acad. Sci. USA 63, 378–383.
2. John, H. A., Birnstiel, M. L., and Jones, K. W. (1969) RNA-DNA hybrids at the cytological
level. Nature 223, 582–587.
3. Lewis, M. E., Sherman, T. G., and Watson, S. J. (1985) In situ hybridization histochemistry
with synthetic oligonucleotides: strategies and methods. Peptides 6, 75–87.
4. Guitteny, A. F., Fouque, B., Mougin, C., Teoule, R., and Bloch, B. (1988) Histological
detection of messenger RNAs with biotinylated synthetic oligonucleotide probes. J. Histo-
chem. Cytochem. 36, 563–571.
5. Hankin, R. C. and Lloyd, R. V. (1989) Detection of messenger RNA in routinely processed
tissue sections with biotinylated oligonucleotide probes. Am. J. Clin. Pathol. 92, 166–171.
6. Larsson, L. I. (1989) In situ hybridization using biotin-labeled oligonucleotides: probe
labeling and procedures for mRNA detection. Arch. Histol. Cytol. 52, 55–62.
7. Farquharson, M., Harvie, R., and McNicol, A. M. (1990) Detection of messenger RNA
using a digoxigenin end labeled oligodeoxynucleotide probe. J. Clin. Pathol. 43, 424–428.
8. Pringle, J. H., Ruprai, A. K., Primrose, L., et al. (1990) In situ hybridization of immunoglobu-
lin light chain mRNA in paraffin sections using biotinylated or hapten-labeled oligonucleo-
tide probes. J. Pathol. 162, 197–207.
9. Schmitz, G. G., Walter, T., Seibl, R., and Kessler, C. (1991) Nonradioactive labeling of
oligonucleotides in vitro with the hapten digoxigenin by tailing with terminal transferase.
Anal. Biochem. 192, 222–231.
10. Shorrock, K., Roberts, P., Pringle, J. H., and Lauder, I. (1991) Demonstration of insulin and
glucagon mRNA in routinely fixed and processed pancreatic tissue by in-situ hybridization.
J. Pathol. 165, 105–110.
11. Jacob, J., Todd, K., Birnstiel, M. L., and Bird, A. (1971) Molecular hybridization of 3H-
labelled ribosomal RNA with DNA in ultrathin sections prepared for electron microscopy.
Biochim. Biophys. Acta 228, 761–766.
12. Webster, H. F., Lamperth, L., Favilla, J. T., Lemke, G., Tesin, D., and Manuelidis, L.
(1987) Use of a biotinylated probe and in situ hybridization for light and electron microscopic
localization of Po mRNA in myelin-forming Schwann cells. Histochemistry 86, 441–444.
13. Guitteny, A. F. and Bloch, B. (1989) Ultrastructural detection of the vasopressin messenger
RNA in the normal and Brattleboro Rat. Histochemistry 92, 277–281.
14. Morel, G., Chabot, J. G., Gossard, F., and Heisler, S. (1989) Is atrial natriuretic peptide
synthesized and internalized by gonadotrophs? Endocrinology 124, 1703–1710.
15. Morel, G., Dihl, F., and Gossard, F. (1989) Ultrastructural distribution of growth hormone
(GH) mRNA and GH intron 1 sequences in rat pituitary gland: effects of GH releasing
factor and somatostatin. Mol. Cell. Endocrinol. 65, 81–90.
16. Singer, R. H., Langevin, G. L., and Lawrence, J. B. (1989) Ultrastructural visualization of
cytoskeletal mRNAs and their associated proteins using double-label in situ hybridization.
J. Cell Biol. 108, 2343–2353.
17. Wolber, R. A., Beals, T. F., and Maassab, H. F. (1989) Ultrastructural localization of herpes
simplex virus RNA by in situ hybridization. J. Histochem. Cytochem. 37, 97–104.
18. Jirikowski, G. F., Sanna, P. P., and Bloom, F. E. (1990) mRNA coding for oxytocin is
present in axons of the hypothalamo-neurohypophyseal tract. Proc. Natl. Acad. Sci. USA
87, 7400–7404.
PROTOCOL: Ultrastructural In Situ Hybridization 143
19. Le Guellec, D., Frappart, L., and Willems, R. (1990) Ultrastructural localization of fibro-
nectin mRNA in chick embryo by in situ hybridization using 35S or biotin labeled cDNA
probes. Biol. Cell 70, 159–165.
20. Le Guellec, D., Frappart, L., and Desprez, P. Y. (1991) Ultrastructural localization of
mRNA encoding for the EGF receptor in human breast cell cancer line BT20 by in situ
hybridization. J. Histochem. Cytochem. 39, 1–6.
21. Le Guellec, D., Trembleau, A., Pechoux, C., Gossard, F., and Morel, G. (1992) Ultrastruc-
tural nonradioactive in situ hybridization of GH mRNA in rat pituitary gland: preembedding
vs ultrathin frozen sections vs postembedding. J. Histochem. Cytochem. 40, 979–986.
22. Trembleau, A., Calas, A., and Fevre-Montange, M. (1990) Ultrastructural localization
of oxytocin mRNA in the rat hypothalamus by in situ hybridization using a synthetic
oligonucleotide. Brain Res. Mol. Brain Res. 8, 37–45.
23. Pomeroy, M. E., Lawrence, J. B., Singer, R. H., and Billings-Gagliardi, S. (1991) Distribution
of myosin heavy chain mRNA in embryonic muscle tissue visualized by ultrastructural in
situ hybridization. Dev. Biol. 143, 58–67.
24. Matsuno, A., Ohsugi, Y., Utsunomiya, H., et al. (1994) Ultrastructural distribution of growth
hormone, prolactin mRNA in normal rat pituitary cells: a comparison between preembedding
and postembedding methods. Histochemistry 102, 265–270.
25. Matsuno, A., Teramoto, A., Takekoshi, S., et al. (1994) Application of biotinylated oligonu-
cleotide probes to the detection of pituitary hormone mRNA using Northern blot analysis,
in situ hybridization at light and electron microscopic levels. Histochem. J. 26, 771–777.
26. Matsuno, A., Ohsugi, Y., Utsunomiya, H., et al. (1995) Changes in the ultrastructural
distribution of prolactin and growth hormone mRNAs in pituitary cells of female rats after
estrogen and bromocriptine treatment, studied using in situ hybridization with biotinylated
oligonucleotide probes. Histochem. Cell Biol. 104, 37–45.
27. Matsuno, A., Utsunomiya, H., Ohsugi, Y., et al. (1996) Simultaneous ultrastructural identifi-
cation of growth hormone and its messenger ribonucleic acid using combined immunohisto-
chemistry and non-radioisotopic in situ hybridization: a technical note. Histochem. J. 28,
703–707.
28. Matsuno, A., Nagashima, T., Takekoshi, S., et al. (1998) Ultrastructural simultaneous
identification of growth hormone and its messenger ribonucleic acid. Endocr. J. 45
(suppl.), S101–S104.
29. Matsuno, A., Ohsugi, Y., Utsunomiya, H., et al. (1998) An improved ultrastructural double-
staining method of rat growth hormone and its mRNA using LR White resin: a technical
note. Histochem. J. 30, 105–109.
30. Matsuno, A., Nagashima, T., Osamura, R. Y., and Watanabe, K. (1998) Application of
ultrastructural in situ hybridization combined with immunohistochemistry to pathophysio-
logical studies of pituitary cell: technical review. Acta Histochem. Cytochem. 31, 259–265.
31. Matsuno, A., Itoh, J., Osamura, R. Y., Watanabe, K., and Nagashima, T. (1999) Electron
microscopic and confocal laser scanning microscopic observation of subcellular organelles
and pituitary hormone mRNA: application of ultrastructural in situ hybridization and immu-
nohistochemistry to the pathophysiological studies of pituitary cells. Endocr. Pathol. 10,
199–211.
32. Matsuno, A., Nagashima, T., Ohsugi, Y., et al. (2000) Electron microscopic observation
of intracellular expression of mRNA and its protein product: technical review on ultrastruc-
tural in situ hybridization and its combination with immunohistochemistry. Histol. Histopa-
thol. 15, 261–268.
33. Matsuno, A., Nagashima, T., Osamura, R. Y., and Watanabe, K. (2000) Electron microscopic
in situ hybridization and its combination with immunohistochemistry, in Molecular Histo-
chemical Techniques (Springer Lab Manual) (Koji, T., ed.), Springer, New York, pp. 204–
221.
144 Matsuno et al.
34. Lloyd, R. V., Jin, L., and Chandler, W. F. (1991) In situ hybridization in the study of
pituitary tissues. Pathol. Res. Pract. 187, 552–555.
35. Escaig-Haye, F., Grigogiev, V., Sharova, I., Rudneva, V., Buckrinskaya, A., and Fournier,
J. G. (1992) Ultrastructural localization of HIV-1 RNA and core proteins. Simultaneous
visualization using double immunogold labelling after in situ hybridization and immunocyto-
chemistry. J. Submicrosc. Cytol. Pathol. 24, 437–443.
36. Egger, D., Troxler, M., and Bienz, K. (1994) Light and electron microscopic in situ hybridiza-
tion: nonradioactive labeling and detection, double hybridization, and combined hybridiza-
tion-immunocytochemistry. J. Histochem. Cytochem. 42, 815–822.
37. Gingras, D. and Bendayan, M. (1995) Colloidal gold electron microscopic in situ hybridiza-
tion: combination with immunocytochemistry for the study of insulin and amylase secretion.
Cell Vision 2, 218–225.
38. Morey, A. L., Ferguson, D. J. P., and Fleming, K. A. (1995) Combined immunocytochemistry
and nonisotopic in situ hybridization for the ultrastructural investigation of human parvovirus
B19 infection. Histochem. J. 27, 46–53.
9
Quantitation of In Situ Hybridization Analysis
INTRODUCTION
In situ hybridization is a powerful method for detecting specific DNA and RNA
sequences in cells and tissue sections. Major areas of applications include cytogenetics,
microbiology, and studies of gene expression and regulation. The methods are based
on the fact that the purine and pyrimidine bases that make up DNA (adenine, A;
guanine, G; thymine, T; and cytosine, C) or RNA (A, G, C, and uracil, U) can form
specific base pairs (A-T, A-U, C-G) that link two complementary strands of DNA or
RNA to each other. Thus, provided the sequence of interest (target) is available or
known, it is possible to synthesize and label the complementary sequence and use it
as a probe for in situ hybridization detection of the target. Since the reaction will
contain only four (DNA-DNA, RNA-RNA) or, at most, five (DNA-RNA) bases as
reactants, it is possible to define conditions mathematically under which only totally
complementary sequences will hybridize. The strength of the hybridization is tempera-
ture dependent, and the melting point Tm defines the temperature at which half the hybrids
formed will dissociate or “melt.” The Tm depends on the degree of complementarity (lack
of mismatches), the number of bases (length of the probe), the frequency of GC base
pairs (which bind more strongly than A-T or A-U), and the ionic strength. Addition
of formamide is often used to lower the Tm.
Formulas incorporating all these factors make it possible to define the experimental
conditions needed for a highly specific (stringent) detection of the target with little or
no risk of cross-hybridization to related sequences (1). However, in practice, factors
other than the mathematically definable base-pairing will contribute to nonspecific
staining, and one should be aware that this method is also not without its caveats
and pitfalls. These sources of error can, however, be reasonably well identified and
circumvented by the use of appropriate controls and supportive biochemical procedures.
Thus, in situ techniques have a great appeal and impact on many basic scientific as
well as diagnostic disciplines.
Every examination of an object involves a comparison with other objects and, hence,
a quantitation. Descriptions of cytochemical reactions usually involve expressions like
strong, medium, and weak reactivity, implying a visual quantitation that is unsubstanti-
ated by real measurements. This can be problematic because humans are ill-equipped
145
146 Larsson
Fig. 1. A demonstration of the imperfection in our perception of intensities. The four larger
squares are of different intensities; the centrally placed smaller squares are of exactly the same
intensity. However, depending on whether the small squares are placed against a darker or
lighter background, we perceive them as being either lighter or darker, respectively.
TISSUE PRETREATMENT
In situ hybridizations require that cells and tissues be fixed as soon as possible.
Thus, RNA in particular is very vulnerable to degradation. For quantitative studies,
standardized tissue handling and rapid fixation is of utmost importance.
The purpose of the fixation for in situ hybridization is threefold: 1) to inhibit
nucleases that may degrade the target, 2) to immobilize the target, and 3) to preserve
the morphology. To obtain a rapid and uniform fixation of the tissue, we prefer to use
perfusion fixation. If this is not possible, thin slices (1 mm thick) of tissue are cut and
immersed in fixative. Handling of clinical material is problematic as it is often difficult
to standardize. To correct for variations in tissue quality, reference hybridizations to
ubiquitous RNAs like ribosomal RNAs have been suggested (2). A problem inherent
Quantitation of ISH Analysis 147
in the use of reference RNAs is that housekeeping and ribosomal RNAs may also
change under experimental and pathologic conditions (3).
Although in the past many different fixatives have been used for in situ hybridization,
there is now general consensus that formaldehyde-based fixatives are preferable. At
the conditions of temperature, concentration, and pH commonly used, formaldehyde
does not react chemically with nucleotides but acts by trapping target sequences in a
crosslinked protein network.
The fixed tissue may be embedded either in paraffin or in plastic, cut on a Vibratome,
or be frozen for cryostat sectioning. Formaldehyde-fixed cultured or isolated cells may
be stored dry or in 70% ethanol. In addition, methods for freezing unfixed tissues
directly for cryostat sectioning or for freeze-drying have been used (4–10). Whatever
method is chosen, the ensuing sectioning must meticulously avoid contaminating the
tissue with RNases. Use of contaminated microtome knives is one obvious risk. A
second risk is connected with the use of proteins like albumin, gelatin, or egg white
for attaching sections to slides. Many of these adhesives are contaminated with RNases
and should be avoided. The use of silane-coated slides, which are expensive to buy
but may be inexpensively manufactured in the laboratory (9), circumvents this problem.
Needless to say, close attention should be paid to the potential contamination of
laboratory equipment and water with RNases. In our experience, water quality may
represent a considerable problem and many workers prefer to add diethylpyrocarbonate
(DEPC) to their distilled water, to inhibit RNases. One should be aware that handling
of this chemical poses a certain health hazard and that appropriate routines for destroying
it by autoclaving are necessary (11). If tissue sections are used for hybridization, due
attention should be paid to the reproducibility of the section thickness, as this parameter
influences both background and signal (12).
The next step of the procedure usually involves permeabilization. Thus, depending
on the tissue and the conditions of fixation and embedding, target sequences may
become more or less inaccessible to the probes. Fixation in conventional 10% buffered
formalin combined with paraffin embedding may totally mask target sequences, whereas
the same fixation combined with cryostat sectioning causes much less masking (7).
Moreover, in certain cell types like keratinocytes, masking will be more severe than
in other cells, due to the content of keratins and other cell-specific proteins.
Accordingly, it is not possible to establish one single permeabilization protocol that
will work for all tissues and pretreatment procedures. Popular methods for permeabiliza-
tion include the use of proteolytic enzymes like proteinase K or pepsin. Digestion with
these enzymes must be carefully optimized since underdigestion produces a suboptimal
signal, and overdigestion may lead to extraction of target. Thus, use of these methods
often results in a compromise between unmasking and loss of target sequences. This
is one of the reasons why accurate enumeration of endogenous target copies is difficult
or impossible. A further problem rests with the use of proteinase K. Being a natural
product, this enzyme is contaminated with RNases, and it is therefore recommended
to let it autodigest (and thus proteolytically to destroy nucleases) before use. This is
not necessary with pepsin, which is used in 0.2 M hydrochloric acid, where nucleases
are inactive. However, pepsin performs best in certain tissues, and proteinase K is best
in others (11). This may reflect differences between cellular proteins and restricts our
options for permeabilization. Despite their shortcomings, proteases are needed for
Quantitation of ISH Analysis 149
obtaining optimal results in most tissues (Fig. 3). After protease digestion, many workers
postfix sections briefly in formaldehyde to prevent losses of target molecules. In cases
of weakly fixed cells and tissues, no or weaker permeabilizing agents like hydrochloric
acid or sodium dodecyl sulfate may be used (8).
Different methods are used after permeabilization to minimize nonspecific binding
of the probe. These include acetylation of positively charged tissue amino groups,
which otherwise may bind the probe ionically. In addition, prehybridization with inert
DNA and/or RNA are often used. In many cases, these procedures are needed, but it
may be worthwhile to check whether they really are helpful. Sometimes prehybridization
may in fact increase background.
reasons, single-stranded probes are preferred for mRNA hybridizations. Three types
of probes are used for this purpose: 1) synthetic oligodeoxynucleotide (oligo) probes,
2) RNA probes, and 3) single-stranded polymerase chain reaction (PCR)-generated
DNA probes.
Oligo-probes are synthesized on DNA synthesis machines and may be labeled either
during or after synthesis. Radioactive labeling is performed after synthesis either by
enzyme-catalyzed incorporation of radioactive phosphor in the 5′-end or of radioactive
nucleotides in the 3′-end (13). The latter reaction is catalyzed by terminal deoxynucleoti-
dyl transferase (terminal transferase) and is the most common approach for labeling
oligoprobes. The nucleotides may be labeled with several different isotopes, including
3
H, 35S, 32P, or 33P. Of these, 35S and 33P are preferred. 3H is seldom used since the
energy of its radiation is so low that much of it is absorbed within the tissue before
reaching the autoradiographic emulsion. Consequently, 3H produces a very precise, but
insensitive, localization. The opposite applies to 32P, which produces radiation of such
high energy that the precision of localization suffers. Thus, 32P may be used for low-
resolution work for which cellular precision is not needed but high sensitivity is required.
A further concern relates to the shelf life of these probes. Thus, whereas 3H has a very
long shelf life, 32P-labeled probes can at most be stored for a week. 35S and 33P have
better precision and longer shelf life than 32P and possess a higher sensitivity than 3H.
Accordingly, these two isotopes are the ones usually employed for radioactive in situ
hybridization and quantitation (14).
These isotopes have been extensively used for quantitative in situ hybridizations in
the central nervous system. In the gastrointestinal tract, 35S-labeled probes have caused
problems by binding nonspecifically to a population of connective tissue cells (15,16).
Although this source of error can be detected by appropriate controls, it is nevertheless
disturbing. Interestingly, the nonspecific binding is 35S-specific and is not seen with
nonradioactively labeled probes (15). It may reflect reactions with disulphide-rich pro-
teins.
Nonradioactive, chemical labeling of oligo-probes may be carried out either during
or after synthesis. Our studies have demonstrated that such nonenzymatic labeling can
be highly advantageous (7,17). It results in sizeable quantities of probes that contain
only the specified number of reporter molecules and, accordingly, produce highly
reproducible results. In contrast, incorporation of nonradioactively labeled nucleotides
in the 3′-end by terminal transferase produces a highly heterogeneous mixture of probes
that may vary severalfold in tail length (7). This reduces the sensitivity of detection
(7). Moreover, the quality of different batches of terminal transferase may vary, leading
to problems with reproducibility that are not encountered with chemically labeled probes.
In our routine, oligo-probes averaging 30 nucleotides in length (30-mers) are labeled
with three or four biotin or digoxigenin molecules in the 3′-end (5). Addition of more
markers per oligoprobe does not seem to improve sensitivity. Although biotin is often
an excellent nonradioactive marker for in situ hybridization, it produces problems in
tissues that are rich in biotin, like the pancreas. Problems may also occur with cells
cultured in biotin-rich media. Although methods and commercial kits for blocking
endogenous biotin exists, they have not been entirely satisfactory. Therefore, use of
digoxigenin as a marker is preferred when biotin background is a problem. Alternatively,
Quantitation of ISH Analysis 151
other markers like fluorescein isothiocyanate (FITC) may be used alone or in combina-
tions with biotin or digoxigenin for multiple hybridizations. Digoxigenin has stood
the test of time as a reporter yielding very low background even after prolonged
developments.
People buying oligo-probes from commercial vendors should take care to specify
that the oligos are intended for use as probes and not as PCR primers. The reason for
this is that PCR primers can be fabricated cheaply with dilute chemistry in the DNA
synthesis machines, whereas oligo-probes need to be of better quality. Methods for
selecting oligonucleotide probes have been discussed elsewhere (5,18,19).
RNA probes are usually produced by in vitro transcription. They can be generated
from sequences cloned into suitable vectors or may be generated by PCR using primers
that incorporate short synthetic RNA polymerase promoter sequences (20,21). Needless
to say, they are more vulnerable to degradation than DNA probes. However, this is not
a major issue since RNase-free conditions are needed anyway for mRNA hybridization.
During in vitro transcription, the probes may be labeled with either radioactively (usually
35
S) or nonradioactively (usually digoxigenin) tagged nucleotides (22,23). Subsequently,
they are usually hydrolyzed to fragments of 100–200 bp, as this may improve penetration
into tissues (23). The hybrids that form in the tissue are resistant to RNase (cf. RNase
protection assays), and nonhybridized probes that are trapped nonspecifically in the
tissue can be removed by RNase A digestion after hybridization (21,22). This may
dramatically reduce background.
PCR-generated single-stranded DNA probes may be produced by assymetrical PCR
but are more efficiently produced by the λ exonuclease method (24). They may be
labeled with radioactively or nonradioactively tagged nucleotides during PCR. Apart
from better stability, they have no specific advantages over RNA probes and are less
commonly used.
The selection between labels and probes is dictated by personal experience and by
specific requirements. For instance, a molecular biologist who wants to have a quick
look at the expression pattern of a newly cloned gene would probably reach that goal
most easily by producing a 35S-labeled RNA probe. At the other extreme, clinical
analyses of chromosome abberations will need the better resolution and faster results
provided by nonradioactive hybridization. For quantitative work, the resolution needed
may also vary. Thus, studies of the central nervous system frequently employ low-
resolution analyses of variations of gene expression in macroscopically defined nuclei.
These studies do not require the high resolution offered by the nonradioactive approach.
In contrast, studies at the cellular level generally benefit from nonradioactive methods
that also permit the use of determinations of cell size and number.
Quantitative in situ hybridization results may be misleading if analyses of cell
numbers and sizes are not undertaken (25). In the past, many authors have claimed
that radioactive methods possess a higher sensitivity than nonradioactive techniques.
With the recent demonstration by the Singer laboratory (26) of single-copy detection
by nonradioactive hybridization, this objection no longer seems to hold. Although,
admittedly, single-copy detection is not routinely attainable by either radioactive or
nonradioactive techniques, advances in nonradioactive methods including the use of
catalyzed reporter deposition (CARD) techniques (27) in conjunction with sensitive
152 Larsson
DETECTION SYSTEMS
In situ hybridization signals revealed by either autoradiographic or nonautoradio-
graphic techniques can be quantitated. In the neurosciences, a low-resolution approach
for autoradiographic detection is often employed. This method makes use of film sheets
that are placed in contact with the hybridized sections. Following a suitable exposure
time (days to weeks), the sheets are processed for development under standardized
conditions and used for densitometry. The density obtained will be proportional to the
degree of blackening of the film. Measurements should be referred to as density since
the autoradiographic silver grains do not absorb light but block its transmission. In this
type of approach, cellular resolution is not obtained. However, different groups of nerve
cells can be identified and average changes in them recorded. In addition, automatic
imagers can be used for this approach (41,42). For obtaining a higher resolution,
conventional autoradiography using liquid emulsion is employed.
Quantitation of ISH Analysis 153
Fig. 5. Determination of gastrin and somatostatin (SRIF) hybridization intensities (left) and
cell numbers in antrophyloric mucosa of freely fed (FF) rats and in rats that have starved (S)
for 20 and 72 h. The data are expressed as percent changes ± SEM relative to the levels in
freely fed animals. Note that both the hybridization intensity and numbers of gastrin mRNA-
positive cells decrease with starvation, whereas the reverse applies to somatostatin mRNA-
positive cells. The magnitude of these changes would be grossly underestimated were only the
hybridization intensities measured.
method for calculating how many probe molecules have hybridized to the section is
to hybridize with different mixtures of labeled and unlabeled probes (46).
Many studies have documented that it is also possible to quantitate nonradioactive
hybridization using enzymes or fluorescent reporter molecules as markers (3,4,12,34,
35,37,39,47–51). As in conventional hybridization cytochemistry, the nonautoradio-
graphic approach is more complicated. However, in addition to advantages already
mentioned, it also yields superior morphologic definition that allows determinations of
cell sizes and cell numbers. The importance of this aspect is often overlooked in
autoradiographic applications (Fig. 5) (25).
Many studies have employed quantitations using alkaline phosphatase as a marker.
The reason for this is that the pioneering work by Butcher (52,53) on quantitation of
this enzyme made it possible to apply his principles directly to the hybridization
analysis (12). Moreover, several popular in situ hybridization methods employ alkaline
phosphatase as a reporter enzyme for nonradioactive hybridization detection. The devel-
oper mostly used for alkaline phosphatase detection is bromochloroindolyl phosphate-
nitroblue tetrazolium (BCIP-NBT). This developer yields formazan reaction products
that exist in a red-absorbing and a blue-absorbing form. The red product forms first
during development and is responsible for the reddish coloration of weakly reactive
sites, whereas the blue product forms later and imparts a deep blue color to strongly
reactive sites. The mixture of the two produces a purple color of sites of intermediate
Quantitation of ISH Analysis 155
from light by aluminum foil. Visual inspection of these slides during development
should be avoided. Development is stopped in a buffer containing EDTA for chelating
the divalent cations needed for alkaline phosphatase activity, and the sections are
mounted in a water-miscible medium.
IMAGE ANALYSIS
of pixels or µm2) and integrated and mean density of defined objects. Objects (often
referred to as regions of interest [ROIs] can be defined manually by tracing with the
mouse or automatically with a “magic wand” that selects contiguous pixels within a
user-definable tolerance zone of gray values. Most common, however, is the use of
thresholding operations. These usually work through sliders that will define a certain
zone of the grayscale (e.g., 27–81 on a scale from 0 to 255). Pixels with gray values
falling within this user-defined zone will become highlighted (e.g., in red) during the
thresholding. As a consequence, it may be difficult to decide whether the highlighted
pixels correspond exactly to the desired ROI. This is made much easier by simultane-
ously displaying the original grayscale image on the monitor. Regrettably, several image
analysis programs do not offer this valuable option.
In addition to manual or interactive thresholding, it is, under certain circumstances,
possible to define automatic, and thus, objective conditions for thresholding. Under most
circumstances, including in situ hybridization measurements, automatic thresholding is,
however, not possible. To avoid the observer bias that may occur during manual
thresholding (59), coded specimens should be analyzed by one single trained operator.
The conditions for thresholding can be made easier by filtering the grayscale image to
enhance its contrast and definition. Several such filters are standard components of
image analysis programs.
Once thresholding is completed, the positional information (x-y coordinates) of the
ROI pixels can be stored in a bitmap image. This is a binary image in which the pixels
of interest are given the value of 1 and the rest are given the value of 0. Several image
analysis programs provide a set of operators that work on binary images through dilation
and erosion of the pixels. In this way, holes in the original binary image can be closed,
edge-bound objects removed, and shapes defined through masked propagation (dilation
of a seed [eroded] image using another image as a mask, behind the contours of which
the seed image cannot be propagated).
Elegant use of these simple operators can achieve a very precise definition of the
objects to be measured. In their simplest form, measurements can be returned by
multiplying the pixels of the binary image with the pixels of the original (unperturbed)
gray level image. Thus, pixels with the same x-y coordinates will be multiplied with
each other. Since the ROI pixels have the value of 1 (and the rest the value of 0) in
the binary image, only the corresponding gray values will remain in the gray value
image. The histogram of this image can be used for obtaining the density of the ROI
in pixel values (Fig. 2). This represents a simplified account of image measurement
routines, and most commercial image analysis software programs include far more
sophisticated procedures that also return positional information and other parameters.
In addition, background readings should be done and subtracted from the measurements
of the hybridized cells. In many cases, background measurements can be made on
nearby negative cells or section areas. This may not be possible with ubiquitous RNAs,
studies of which may necessitate background readings from control sections (e.g.,
pretreated with RNases before hybridization or hybridized with labeled non-sense or
scrambled probes). For studies of the linearity of the hybridization system, cytochemical
models can be used (12,60) (Fig. 6). If the system is calibrated, the pixel values can
be recalculated to numbers of hybridizable copies of target, OD, or radioactivity units.
Quantitation of ISH Analysis 159
Fig. 6. Cytochemical models using sense sequences of MSH or gastrin immobilized on glass
slides as described (60) and subsequently hybridized with complementary biotinylated probes
followed by alkaline phosphatase-based detection. Note that the signal increases in parallel with
increasing concentrations of target.
The number of ROI pixels will also be known and can be used to calculate mean
density, area, perimeter, form factor, and so on.
Different image analysis packages are available for different needs. One excellent,
introductory possibility is the shareware program NIH Image (available from http://
zippy.nimh.nih.gov or http://rsb.info.nih.gov/NIH-Image/download.html). Unfortu-
nately, this program is only available for use on Macintosh computers. It can be used
alone or in combination with more heavyweight programs like TCL (now SCIL) Image
(TNO Institute of Applied Physics, Delft, The Netherlands). We have used these
programs together on the Macintosh platform for well over 10 years. The NIH Image
program has been used for controlling the frame grabber with respect to shade correction
and frame averaging, whereas most measurements have been performed using the TCL
Image package, which permits some very sophisticated filtering routines on the gray
level as well as on binary images. The TCL/SCIL Image program (61) is available on
many platforms including DOS/Windows.
Another program that is available on both Macintosh and PCs is IP Lab, which we
have had excellent experiences with on the Macintosh platform (IPLab Spectrum, Signal
Analytics, Vienna, Virginia). This program can be used for analysis of both color and
grayscale images. Many of the larger image analysis programs take some time to learn.
This is not the case with the free NIH Image program, which constitutes the main
image analysis package in many laboratories. As an alternative, some workers use the
Adobe Photoshop program for image analysis (62,63). This program is available for
both PCs and Macintoshes. A very worthwhile alternative is the Image-Pro Plus program
(Media Cybernetics, Silver Spring, MD). Compared with other heavyweight image
analysis packages, this program seems to be extremely easy to use. Needless to say,
the above concerns relate only to image analysis programs that we have tested personally,
and many more are available on the market. Fortunately, many demo versions are now
available so one can test the program for free before buying it.
160 Larsson
ACKNOWLEDGMENTS
I thank the Danish Medical Research Council and the Danish Cancer Society for
support.
REFERENCES
1. Fitzpatrick-McElligott, S., Lewis, M. E., Tyler, M., Baldino, F. Jr., and Davis, L. G. (1988)
In situ hybridization with radiolabeled synthetic oligodeoxynucleotide probes. Dupont Bio-
technol. Update 3, 2–3.
2. Yoshii, A., Koji, T., Ohsawa, N., and Nakane, P. K. (1995) In situ localization of ribosomal
RNAs is a reliable reference for hybridizable RNA in tissue sections. J. Histochem. Cyto-
chem. 43, 321–327.
3. Guiot, Y. and Rahier, J. (1997) Validation of non-radioactive in situ hybridization as a
quantitative approach of messenger ribonucleic acid variations. A comparison with Northern
blot. Diagn. Mol. Pathol. 6, 261–266.
4. Guiot, Y. and Rahier, J. (1995) The effects of varying key steps in the non-radioactive
hybridization protocol: a quantitative study. Histochem. J. 27, 60–68.
5. Hougaard, D. M., Hansen, H., and Larsson, L-I. (1997) Non-radioactive in situ hybridization
for mRNA with emphasis on the use of oligodeoxynucleotide probes. Histochem. Cell Biol.
108, 335–344.
6. Larsson, L-I., Christensen, T., and Dalbøge, H. (1988) Detection of proopiomelanocortin
mRNA by in situ hybridization using a biotinylated ligodeoxynucleotide probe and avidin-
alkaline phosphatase histochemistry. Histochemistry 89, 109–116.
7. Larsson, L-I. and Hougaard, D. M. (1990) Optimization of non-radioactive in situ hybridiza-
tion: image analysis of varying pretreatment, hybridization and probe labelling conditions.
Histochemistry 93, 347–354.
8. Lawrence, J. B. and Singer, R. H. (1985) Quantitative analysis of in situ hybridization
methods for the detection of actin gene expression. Nucleic Acids Res. 13, 1777–1799.
9. Rentrop, M., Knapp, B., Winter, H., and Schweizer, J. (1986) Aminoalkylsilane-treated
glass slides as support for in situ hybridization of keratin cDNAs to frozen tissue sections
under varying fixation and pretreatment conditions. Histochem. J. 18, 271–276.
10. West, M. J., Ostergaard, K., Andreassen, O. A., and Finsen, B. (1996) Estimation of the
number of somatostatin neurons in the striatum: an in situ hybridization study using the
optical fractionator method. J. Comp. Neurol. 370, 11–22.
11. Larsson, L-I. and Hougaard, D. M. (1993) Non-radioactive in situ mRNA hybridization
using synthetic oligonucleotides: principles, combination with immunocytochemistry and
quantitation. Neurosci. Protocols 20, 1–18.
12. Larsson, L-I., Traasdahl, B., and Hougaard, D. M. (1991) Quantitative non-radioactive in
situ hybridization. Model studies and studies on pituitary proopiomelanocortin cells after
adrenalectomy. Histochemistry 95, 209–215.
13. Collins, M. L. and Hunsaker, W. R. (1985) Improved hybridization assays employing tailed
oligonucleotide probes: a direct comparison with 5′-end-labeled oligonucleotide probes and
nick-translated plasmid probes. Anal. Biochem. 151, 211–224.
14. Baskin, D. G. and Stahl, W. L. (1993) Fundamentals of quantitative autoradiography by
computer densitometry for in situ hybridization with emphasis on 33P. J. Histochem. Cyto-
chem. 41, 1767–1776.
15. Larsson, L-I., Tingstedt, J-E., Madsen, O. D., Serup, P., and Hougaard, D. M. (1995) The
LIM-homeodomain protein Isl-1 segregates with somatostatin but not with gastrin expression
during differentiation of somatostatin/gastrin precursor cells. Endocrine 3, 519–524.
Quantitation of ISH Analysis 161
16. Panula, P. and Wasowicz, K. (1993) Action of antiulcer drugs. Science 262, 1454–
1455.
17. Larsson, L-I. and Hougaard, D. M. (1993) Sensitive detection of rat gastrin mRNA by in
situ hybridization with chemically biotinylated oligodeoxynucleotides: validation, quantita-
tion and double-staining studies. J. Histochem. Cytochem. 41, 157–163.
18. Lewis, M. L., Sherman, T. G., and Watson, S. J. (1985) In situ hybridization histochemistry
with synthetic oligonucleotides: strategies and methods. Peptides 6, 75–87.
19. Stahl, W. L., Eakin, T. J., and Baskin, D. G. (1993) Selection of oligonucleotide probes
for detection of mRNA isoforms. J. Histochem. Cytochem. 41, 1735–1740.
20. Brink, P. E. and Grimm, P. C. (1994) Rapid nonradioactive in situ hybridization for
interleukin-2 mRNA with riboprobes generated using the polymerase chain reaction.
J. Immunol. Methods 167, 83–89.
21. Sitzmann, J. H. and LeMotte, P. K. (1993) Rapid and efficient generation of PCR-derived
riboprobe templates for in situ hybridization histochemistry. J. Histochem. Cytochem. 41,
773–776.
22. Komminoth, P., Merk, F. B., Leav, I., Wolfe, H. J., and Roth, J. (1992) Comparison of
35
S- and digoxigenin-labeled RNA and oligonucleotide probes for in situ hybridization.
Histochemistry 98, 217–228.
23. Schaeren-Wiemers, N. and Gerifn-Moser, A. (1993) A single protocol to detect transcripts
of various types and expression levels in neural tissue and cultured cells: in situ hybridization
using digoxigenin-labelled cRNA probes. Histochemistry 100, 431–440.
24. Hannon, K., Johnstone, E., Craft, L. S., et al. (1993) Synthesis of PCR-derived, single-
stranded DNA probes suitable for in situ hybridization. Anal. Biochem. 212, 421–427.
25. McCabe, J. T. and Bolender, R. P. (1993) Estimation of tissue mRNAs by in situ hybridiza-
tion. J. Histochem. Cytochem. 41, 1777–1783.
26. Femino, A. M., Fay, F. S., Fogarty, K., and Singer, R. H. (1998) Visualization of single
RNA transcripts in situ. Science 280, 585–590.
27. Adams, J. C. (1992) Biotin amplification of biotin and horseradish peroxidase signals in
histochemical stains. J. Histochem. Cytochem. 40, 1457–1463.
28. Kerstens, H. M. J., Poddighe, P. J., and Hanselaar, G. J. M. (1995) A novel in situ
hybridization signal amplification method based on deposition of biotinylated tyramide. J.
Histochem. Cytochem. 43, 347–352.
29. Raap, A. K., van de Corput, M. P. C., Vernenne, R. A. W., van Gijlswijk, R. P. M., Tanke,
H. J., and Wiegant, J. (1995) Ultrasensitive FISH using peroxidase-mediated deposition of
biotin- or fluorochrome tyramides. Hum. Mol. Genet. 4, 529–534.
30. Van Giljswijk, R. P. M., Wiegant, J., Raap, A. K., and Tanke, H. J. (1996) Improved
localization of fluorescent tyramides for fluorescence in situ hybridization using dextran
sulphate and polyvinyl alcohol. J. Histochem. Cytochem. 44, 389–392.
31. Long, A. A., Mueller, J., Andre-Schwartz, J., Barrett, K. J., Schwartz, R., & Wolfe, H.
(1992) High-specificity in-situ hybridization. Diagn. Mol. Pathol. 1, 45–57.
32. Lloyd, R. V. and Jin, L. (1995) In situ hybridization analysis of chromogranin A and B
mRNAs in neuroendocrine tumors with digoxigenin-labeled oligonucleotide probe cocktails.
Diagn. Mol. Pathol. 4, 143–151.
33. Trembleau, A. & Bloom, F. E. (1995) Enhanced sensitivity for light and electron microscopic
in situ hybridization with multiple simultaneous non-radioactive oligodeoxynucleotide
probes. J. Histochem. Cytochem. 43, 829–841.
34. de Pauw, E. S., Verwoerd, N. P., Duinkerken, N., et al. (1998) Assessment of telomere
length in hematopoietic interphase cells using in situ hybridization and digital fluorescence
microscopy. Cytometry 32, 163–169.
35. Hultdin, M., Gronlund, E., Norrback, K., Erikson-Lindstrom, E., Just, T., and Roos, G.
162 Larsson
(1998) Telomere analysis by fluorescence in situ hybridization and flow cytometry. Nucleic
Acids Res. 26, 3651–3656.
36. Poon, S. S., Martens, U. M., Ward, R. K., and Lansdorp, P. M. (1999) Telomere length
measurements using digital fluorescence microscopy. Cytometry 36, 267–278.
37. Rufer, N., Dragowska, W., Thornbury, G., Roosnek, E., and Landsdorp, P. M. (1998)
Telomere length dynamics in human lymphocyte subpopulations measured by flow cytome-
try. Nat. Biotechnol. 16, 743–747.
38. Larsson, L-I. (1997) Quantitative in situ hybridization. Endocr. Pathol. 8, 3–9.
39. Leeuw, T. and Pette, D. (1994) Kinetic microphotometric evaluation of in situ hybridization
for mRNA of slow myosin heavy chain in type I and C fibres of rabbit muscle. Histochemistry
102, 105–112.
40. Cash, E. and Brahic, M. (1986) Quantitative in situ hybridization using initial velocity
measurements. Anal. Biochem. 157, 236–240.
41. Vizi, S. and Gulya, K. (2000) Calculation of maximal hybridization capacity (Hmax) for
quantitative in situ hybridization. A case study for multiple calmodulin mRNAs. J. Histo-
chem. Cytochem. 48, 893–904.
42. Laniece, P., Charon, Y., Cardona, A., et al. (1998) A new high resolution radioimager for
the quantitative analysis of radiolabelled molecules in tissue section. J. Neurosci. Methods
86, 1–5.
43. Jonker, A., de Boer, P. A. J., van den Hoff, M. J. B., Lamers, W. H., and Moorman, A. F.
M. (1997) Towards quantitative in situ hybridization. J. Histochem. Cytochem. 45, 413–423.
44. Palfi, A., Hatvani, L., and Gulya, K. (1998) A new quantitative film autoradiographic
method of quantifying mRNA transcripts for in situ hybridization. J. Histochem. Cytochem.
46, 1141–1149.
45. Singer, R. H., Lawrence, J. B., & Villnave, C. (1986) Optimization of in situ hybridization
using isotopic and non-isotopic detection methods. Biotechniques 4, 230–250.
46. Gerfen, C. R., McGinty, J., and Young, W. S. III. (1991) Dopamine differentially regulates
dynorphin, substance P, and enkephalin expression in striatal neurons: in situ hybridization
histochemical analysis. J. Neurosci. 11, 1016–1031.
47. Albalwi, M., Hammond, D. W., Goepel, J. R., Hough, R. E., and Goyns, M. H. (1999)
Semi-quantitative fluorescence in situ hybridization analysis indicates that the myc protein
is consistently stabilized both before and after transformation of low-grade follicular center
to high-grade diffuse large cell lymphoma. Lab. Invest. 79, 707–715.
48. Cheng, L., Bucana, C. D., and Wei, Q. (1996) Fluorescence in situ hybridization method
for measuring transfection efficiency. Biotechniques 21, 486–491.
49. Higo, N., Oishi, T., Yamashita, A., Matsuda, K., and Hayashi, M. (1999) Quantitative non-
radioactive in situ hybridization study of GAP-43 and SCG10 mRNAs in the cerebral cortex
of adult and infant macaque monkeys. Cereb. Cortex 9, 317–331.
50. Rauch, J., Wolf, D., Craig, J. M., Hausmann, M., and Cremer, C. (2000) Quantitative
microscopy after fluorescence in situ hybridization—a comparison between repeat-depleted
and non-depleted DNA probes. J. Biochem. Biophys. Methods 44, 59–72.
51. Salchi, M., Barron, M., Merry, B. J., and Goyns, M. H. (1999) Fluorescence in situ
hybridization analysis of the fos/jun ratio in ageing brain. Mech. Ageing Dev. 107, 61–71.
52. Butcher, R. G. (1972) Precise cytochemical measurement of neotetrazolium formazan by
scanning and integrating microdensitometry. Histochemie 32, 171–190.
53. Butcher, R. G. (1978) The measurement in tissue sections of the two formazans derived
from nitroblue tetrazolium in dehydrogenase reactions. Histochem. J. 10, 739–744.
54. Mason, D. Y. (1985) Immunocytochemical labelling of monoclonal antibodies by the
APAAP immunoalkaline phosphatase technique: In Techniques in Immunocytochemistry,
vol. 3 (Bullock, G. B. and Petrusz, P., eds.), Academic, New York, pp. 25–42.
55. Guerin-Reverchon, I., Chardonnet, Y., Chignol, M. C., & Thivolet, J. (1989) A comparison
Quantitation of ISH Analysis 163
10
Confocal Laser Scanning Microscopy
INTRODUCTION
Immunoelectron microscopy and electron microscopic in situ hybridization are
undoubtedly the best methods for following the dynamic changes of subcellular organ-
elles; however, these techniques require specific tissue preparation and equipment.
More recent developments include a more refined and sophisticated technique, confocal
laser scanning microscopy (CLSM), which was originally described by Minsky in 1957
(1) and has since been applied to the field of medical biology. In early experiments,
only fluorescent signals were detectable by CLSM (2–9); however, recent innovations
have made possible the visualization of nonfluorescent signals such as horseradish
peroxidase (HRP) and diaminobenzidine (DAB) signals by CLSM (10,11). Moreover,
the combination of CLSM and the image analysis system (IAS) (12) has allowed us
to visualize subcellular organelles three-dimensionally in routinely processed light
microscopic specimens. We applied CLSM to specimens prepared for light microscopy
(12) and demonstrated the intracellular identification of subcellular organelles and
pituitary hormone mRNA, comparable to that of electron microscopy (13,14). We also
applied CLSM to the study of tumor angiogenesis (15) and the microvessel environment
of hormone-secreting cells (16). The visualization of subcellular organelles, mRNA
and protein products, as well as three-dimensional images of microvessel environment
of hormone-secreting cells is discussed in this chapter.
GENERAL PRINCIPLES
The fundamentals of CLSM date back to 1957, when Minsky (1) described this new
type of microscopy. As commonly applied, CLSM has a viable pinhole, and its modify-
ing scan angles allow the adjustment of magnification. A schematic example of the
pinhole of CLSM is given in Fig. 1 The laser light is focused via the scanner through
the tube lens and the objective, illuminating a small spot in the specimen. Emitted light
emanating from the focal plane and the planes above and below is directed via the
scanner to the dichroic beam splitter (dichroic mirror), where it is decoupled and
165
166 Matsuno et al.
Fig. 2. Schematic drawing of transmittance and confocal images. The reflectance confocal
image is an individual dimensional image obtained at any optional cross section, whereas the
transmittance image is a projected or integrated two-dimensional image from three-dimen-
sional images.
Probes
Nonfluorescent as well as fluorescent signals are detectable under CLSM (2–9).
The non-fluorescent signals, including HRP, DAB, osmium black, azo dye, nitroblue
tetrazolium (NBT), and heavy metals, have valuable advantages over fluorescent signals
168 Matsuno et al.
since these nonfluorescent signals are permanently preserved and allow repeated or
retrospective examination.
Results
CLSM observation of a paraffin section of a human somatotroph adenoma immuno-
stained with anti-GH antibody revealed secretory granules positively immunostained
for GH (Fig. 3).
6. Tissue sections were treated with 0.1 µg/mL proteinase K at 37°C for 30 min, followed
by treatment for 10 min with 0.25% acetic anhydride in 0.1 M triethanolamine.
7. The slides were washed in 2× sodium chloride sodium citrate (SSC) at room temperature
for 3 min and then prehybridized at 37°C for 30 min. The prehybridization solution consisted
of 10% dextran sulfate, 3× SSC, 1× Denhardt’s solution (0.02% Ficoll/0.02% BSA/0.02%
polyvinylpyrrolidone), 100 µg/mL salmon sperm DNA, 125 µg/mL yeast tRNA, 10 µg/mL
polyadenylic-cytidylic acid, 1 mg/mL sodium pyrophosphate, pH 7.4, and 50% formamide.
8. The biotinylated oligonucleotide probe at a concentration of 0.1 ng/µL was diluted with
this prehybridization solution, and hybridization was carried out at 37°C overnight. The
sequence of antisense synthesized oligonucleotide probe for rat PRL mRNA and the details
of biotinylation were described in our previous reports (14,17–22).
9. The slides were washed at room temperature with 2× SSC, 1× SSC, and then 0.5× SSC
for 15 min.
10. After washing in PBS for 5 min, the slides were dipped in DAB solution (incomplete
Graham-Karnovsky’s solution) for 1 h.
11. The hybridization signals were detected with streptavidin-biotin-horseradish peroxidase
(ABC-HRP), using Vectastain’s ABC kit (Vector, Burlingame, CA), and thereafter hybrid-
ization signals were developed with 0.017% H2O2 in DAB solution (complete Graham-
Karnovsky’s solution) for 7 min. At this stage, hybridization signals were confirmed light
microscopically.
12. CLSM observation of the immunostained tissue specimens was as described above.
Results
With the CLSM transmittance mode, diffuse positive signals of PRL mRNA were
observed, evidently in the cytoplasm, and the whirling structure of the rough endoplasmic
170 Matsuno et al.
Fig. 4. With CLSM transmittance mode, diffuse positive signals of PRL mRNA were
observed, evidently in the cytoplasm, and the whirling structure of RER (arrow) was also noted.
Nu, nucleus. Bar = 1 µm. (Reproduced with permission from ref. 14.).
reticulum (RER) was also noted (Fig. 4). With the CLSM reflectance confocal mode
combined with IAS, positive signals of PRL mRNA were observed (Fig. 5). With
combined transmittance and reflectance confocal images, positive signals of PRL mRNA
and granular polysome-like structures were noted on the whirling RER in the cytoplasm
(Fig. 6). Under high-power magnification, these signals were interpreted to be similar
and comparable to ISH signals of PRL mRNA on the whirling RER observed under
EM (Fig. 7) (14).
Comments
The ultrastructural resolution of CLSM may be insufficient compared with that
of electron microscopy. Nevertheless, a reflectance confocal image combined with a
transmittance image has merit in that it can provide an individual dimensional image
at any optional cross-sections of a routinely processed light microscopic specimen.
Another merit of CLSM is the large number of cells that can be observed, whereas
only a limited number of cells can be observed under electron microscopy. CLSM can
identify, at the ultrastructural level, ISH signals of mRNA, or immunohistochemical
signals of protein together with subcellular organelles; these are similar and comparable
to those observed under electron microscopy (13,14). CLSM is available for the study
of the spatial relationship of mRNA and encoded protein and may be another useful
tool for the ultrastructural ISH study of mRNA.
PROTOCOL: CLSM Observations 171
Fig. 5. With CLSM reflectance confocal mode combined with IAS (LSM software version
3.98), positive signals of PRL mRNA (arrow) were observed. Nu, nucleus. Bar = 1 µm (Repro-
duced with permission from ref. 14.)
Fig. 6. With a combination of transmittance images and reflectance confocal images, positive
signals of PRL mRNA and granular polysome-like structures were noted on the whirling RER
in the cytoplasm (arrow). Nu, nucleus. Bar = 1 µm. (Reproduced with permission from ref. 14.)
172 Matsuno et al.
Fig. 7. Under high-power magnification, these signals were interpreted as being similar and
comparable to ISH signals of PRL mRNA on the whirling RER observed under EM (arrow).
Bar = 1 µm. (Reproduced with permission from ref. 14.)
4. Then FITC-conjugated gelatin was dialyzed in 0.01 M PBS, pH 7.4, at 37°C in the dark for
1 week.
5. Aliquots of the dialyzed FITC-conjugated gelatin in a total volume of 200 mL were injected
into rats.
Tissue Preparation
1. Female adult Wistar rats were used after bilateral adrenalectomy or adrenocorticotropic
hormone (ACTH) intramuscular administration (0.5 mg/day) for 2 weeks. Nontreated rats
served as control.
2. The abdominal and chest walls of the animals were opened under diethyl ether inhalation
anesthesia. To demonstrate microvessels, 200 mL of dialyzed FITC (30 mg/mL) conjugated
gelatin solution was injected into the left ventricle of the heart, cutting off the right atrium.
3. After complete perfusion over approximately 3–5 min, the animals were fixed in cold 4%
paraformaldehyde (PFA; Wako Pure Chemical, Osaka, Japan) added to 15% picric acid
(Wako) solution (0°C) over approximately 25–30 min.
4. Then wet pituitary glands were resected and immediately fixed overnight in graded cold
PFA (0.5–8%) containing 15% picric acid overnight at 4°C.
5. For simultaneous visualization of microvessel networks and immunohistochemical localiza-
tion of ACTH and GH, the pituitary tissues were cut into 2–3-mm-thick blocks, placed on
50 × 70-mm coverslips (Matsunami Glass, Osaka, Japan), and mounted in 0.05 M Tris-HCl
buffer, 1,4-diazabicyclo[2,2,2]octane (DABCO; Wako).
6. The pituitary tissue specimens were used for frozen sections. These frozen sections were
obtained from tissues that had been fixed overnight in 4% PFA dissolved in 0.01 M PBS,
pH 7.4, immersed in graded concentrations of sucrose in PBS (10% for 1 h, 15% for 2 h,
and 20% for 4 h), embedded in OCT compound, and stored at −80°C until use.
7. Frozen sections (30 µm thick) were mounted on 3-aminopropylmethoxysilane-coated slides.
Immunohistochemistry
An immunohistochemical analysis of hormone-secreting endothelial cells was per-
formed by the use of indirect immunofluorescent and an immunoperoxidase labeled
antibody, as described by Nakane in 1975 (45).
1. The first antibody employed was a rabbit polyclonal anti-ACTH antibody (1:200; DAKO,
Santa Barbara, CA) or a rabbit polyclonal anti-GH antibody (1:200; DAKO).
2. The second antibody employed was tetramethylrhodamine isothiocyanate (TRITC)-conju-
gated anti-rabbit IgG (swine, 1:10; DAKO) or HRP-conjugated anti-mouse IgG (sheep, 1:50;
Amersham, Poole, UK).
3. The peroxidase coloring reaction was performed with DAB. The nuclei were then counter-
stained with 5% methyl green (MG) buffered by 0.1 mol/L Veronal acetate at pH 4.0
(Chroma, Kögen, Norway).
Confocal Laser Scanning Microscopy System
CLSM was employed on thick sections (1–2 mm) to elucidate the relationship
between hormone-secreting endothelial cells and the microvessel network by stereo
pair, colored depth focus, projection, and three-dimensional cutting at reconstruction
models. The system included a 488-nm argon laser (for FITC) and/or a 543-nm helium-
neon laser (for TRITC and DAB). Optical fluorescence signals of microvessel network
patterns labeled by FITC were observed using a dichroic beam splitter (NT 80/20/543;
Carl Zeiss) and emission filter (BP510-525; Carl Zeiss). Optical fluorescence signals
of ACTH-secreting cells labeled with TRITC were observed using a dichroic beam
splitter (NT 80/20/543; Carl Zeiss) and emission filter (LP570; Carl Zeiss). Optical
174 Matsuno et al.
reflectance signals of GH cells labeled by HRP were observed using a dichroic beam
splitter (NT 80/20/543; Carl Zeiss) and no emission filter. Double- or triple-labeled
volumetric data sets of these images taken with CLSM can be digitized and made
susceptible to image analysis manipulation and computer-assisted three-dimensional
reconstructions using LSM software.
Optical tomographic imaging (Z-images) for three-dimensional reconstruction of
the hormone-secreting endocrine cells and their microvessels was done with a plan-
apochromat (×63, oil immersion, NA 1.4; Carl Zeiss) and a plan-neofluar (×10, NA
0.30; ×20, NA 0.50; Carl Zeiss) objective lens. Z-directional movement for optical
sectioning of entire specimens was controlled by a stepping motor unit for axial scanning
at 0.5–20-µm focus steps. The image resolution was 512 × 512 and/or 1024 × 1024
pixels (8 bits, 256 gray levels).
Results
Distribution of Three-Dimensional Microvessels as Determined by Fluorescent CLSM
By CLSM, FITC fluorescence-conjugated gelatin injected into the left ventricle of
the heart was observed to enter the capillaries of the control rat pituitary glands (Fig.
8). The CLSM images in the fluorescence confocal mode identified straight vessels,
arcs, branches, loops, and well-developed networks of capillaries in the anterior pituitary
gland three dimensionally.
Immunohistochemical Localization of ACTH and GH in Nontreated Control Rats
by CLSM (See Color Plate 5 following page 208.)
Double three-dimensional imaging with the visualization of hormone-producing cells
by TRITC- and/or HRP-labeled antibodies demonstrated the shape of ACTH and GH
cells and their relationship to the surrounding capillaries (Fig. 9). The lumina of the
capillaries were labeled yellowish green by FITC. ACTH (Fig. 9a,b) and GH cells
(Fig. 9c,d) fluoresced red. Many stellate cells were positive for ACTH, and microvessels
were clearly visible in the pituitary glands. In ACTH cells, cells were frequently attached
to the capillaries, and long cytoplasmic processes terminated at the capillaries (Fig.
9a). A high-power view of the image in Fig. 9a reveals more clearly the relationship
between microvessel networks and ACTH-positive cells (Fig. 9b). The terminals of
the ACTH processes and the capillary lumina are separated by approximately 600 nm.
This suggests endothelial cells and basement membranes. The reaction products of
ACTH appeared granular (Fig. 9b). In contrast, GH cells were generally rounded and
frequently lay in free spaces among the FITC fluorescence-labeled capillaries (Fig. 9c).
GH also appeared granular in the cytoplasm (Fig. 9d).
DAB (osmium black) also gave reflectance signals at a wavelength of 543 nm. DAB
labeling for GH was applied to the double three-dimensional imaging with FITC
fluorescence. CLSM detected fluorescence of FITC at a depth of focus of 2000 µm (Fig.
8) and fluorescence of TRITC and reflectance signals for DAB at 300 µm (Fig. 9a,c).
Immunohistochemical Localization of ACTH in Adrenalectomy and ACTH
Administration Rat Anterior Pituitary Gland by CLSM
In the adrenalectomized rats, the anterior pituitary gland showed markedly increased
numbers of ACTH-immunoreactive cells (Fig. 10a). The positive cells revealed that
PROTOCOL: CLSM Observations 175
Comments
Three-dimensional imaging by CLSM has clearly demonstrated well-developed capil-
lary networks and a relationship with the hormone-secreting endocrine cells. In addition
to normal pituitary vessels that form morphologic patterns, straight vessels that parallel
each other, parallel vessels that crosslink, arcs (incompletely closed loops), arcs with
branching, closed loops, and networks of closed loops (Fig. 8) have been identified.
The presence of networks of closed loops has been found to have the strongest association
with hormone-secreting endocrine cells of any parameter, including cell shape, cell
176 Matsuno et al.
Fig. 9. Three-dimensional projection images of ACTH and GH cells and combined signals
of fluorescent (FITC) and nonfluorescent (DAB) probe in normal rat pituitary glands. Red:
ACTH cells (anti-ACTH-TRITC), fluorescence mode (A, B); GH cells (anti-GH-HRP-DAB)
reflectance mode (C, D). Green: microvessels (FITC-conjugated gelatin), fluorescence mode.
(A) ACTH cells are closely related to the microvessel networks. Low power. (B) ACTH signals
show a granular pattern in the cytoplasm. High power. (C) GH cells are clearly visible three-
dimensionally by CLSM reflectance mode. Low power. (D) GH signals appear granular in the
cytoplasm three-dimensionally by CLSM reflectance mode. High power. Nu, nucleus. Bars =
25 µm in A and C; 5 µm in B and D. (See Color Plate 5 following page 208.)
size, and cell type. The presence of parallel vessels with crosslinking is also strongly
associated with the development or processing of hormone-secreting endocrine cells
(Fig. 9).
Our technique, which combined FITC-gelatin injection with immunohistochemistry
using HRP- and/or fluorescence-labeled antibody as described here, was designed to
visualize the microcirculatory patterns in three dimensions and to provide a simultaneous
three-dimensional visualization of the microcirculation and hormone-secreting endo-
crine cells processing ACTH or GH by an FITC-conjugated gelatin injection method.
The critical points for the demonstration of the capillaries seemed to lie in 1) warming
of the FITC fluorescence-conjugated gelatin and 2) peristaltic infusion of the solution
to prevent thrombotic occlusion of the capillaries. For demonstration of ACTH and
GH, effusion fixation by picric acid paraformaldehyde solution (Zamboni) preserved
antigenicity as well as cell shape and cytoplasmic structures.
The technique of combined double immunolabeling and/or histochemical labeling,
CLSM, and computer-assisted three-dimensionally reconstructed analysis does have
advantages. First, as mentioned above, each of the labels used was detectable in deeper
PROTOCOL: CLSM Observations 177
levels and was able to reveal the relationship between vessels and tissues in the pituitary
gland. Second, unlike scanning electron microscopy by cast corrosion methods, which
make it impossible to observe the relationship between casts and tissues, this technique
permits the three-dimensional study of samples from low- to high-power magnification.
The technique described here is most suitable for the study of small lateral areas. To
create a three-dimensionally reconstructed image, one must overlap data not only along
the x- and y-axes but also along the z-axis, a procedure that is tedious and requires
extensive computer resources.
By this three-dimensional method, ACTH cells were clearly depicted as polygonal
cells with long cytoplasmic processes, which terminated on the capillaries. This shape
of ACTH cells has frequently been described as unique and characteristic. GH cells
178 Matsuno et al.
were round and were located among the capillaries. It should be emphasized that this
method not only elucidates the shape and relationship to the capillaries but also depicts
a granular localization on CLSM, suggestive of secretory granules.
Our experimental study to visualize simultaneously the immunohistochemical local-
ization of ACTH and GH and the microvessel network using CLSM with the fluorescence
probes FITC, TRITC, and DAB revealed a clear characteristic intracellular localization
of ACTH and GH in the rat anterior pituitary glands on thick specimens. These images
showed both the immunohistochemical localization of ACTH and GH and the pattern
of the microvessel network. With this technique, FITC injection combined with immuno-
histochemistry simultaneously revealed differences in the localization or distribution.
Permanently preserved sections allow repeated or retrospective examinations as well
as observation of organs in vivo.
This method clearly showed the relationship between the microvasculature and the
hormone-secreting endocrine cells on thick light microscopic specimens. It also permit-
ted recognition of subcellular sites of the microvasculature and hormone-secreting
endocrine cells, obviating most of the cumbersome procedures associated with patho-
logic diagnosis in tissues containing tumor microvessels.
REFERENCES
1. Minsky, M. (1957) Microscopy apparatus. U.S. Patent No. 3013467.
2. Arndt-Jovin, D. J., Robert-Nicoud, M., and Jovin, T. M. (1990) Probing DNA structure
and function with a multi-wavelength fluorescence confocal laser microscope. J. Microsc.
157, 61–72.
3. Arndt-Jovin, D. J., Robert-Nicoud, M., Kaufman, S. J., and Jovin, T. M. (1985) Fluorescence
digital imaging microscopy in cell biology. Science 230, 247–256.
4. Bauman, J. G., Bayer, J. A., and van Dekken, H. (1990) Fluorescent in-situ hybridization
to detect cellular RNA by flow cytometry and confocal microscopy. J. Microsc. 157, 73–81.
5. Hozak, P., Novak, J. T., and Smetana, K. (1989) Three-dimensional reconstructions of nucleo-
lus-organizing regions in PHA-stimulated human lymphocytes. Biol. Cell 66, 225–233.
6. Michel, E. and Parsons, J. A. (1990) Histochemical and immunocytochemical localization
of prolactin receptors on Nb2 lymphoma cells: applications of confocal microscopy.
J. Histochem. Cytochem. 38, 965–973.
7. Takamatsu, T. and Fujita, S. (1988) Microscopic tomography by laser scanning microscopy
and its three-dimensional reconstruction. J. Microsc. 149, 167–174.
8. Tao, W., Walter, R. J., and Berns, M. W. (1988) Laser-transected microtubules exhibit
individuality of regrowth; however, most free new ends of the microtubules are stable.
J. Cell Biol. 107, 1025–1035.
9. White, J. G., Amos, W. B., and Fordham, M. (1987) An evaluation of confocal versus
conventional imaging of biological structures by fluorescence light microscopy. J. Cell
Biol. 105, 41–48.
10. Itoh, J., Utsunomiya, H., Komatsu, N., Takekoshi, S., Osamura, R. Y., and Watanabe, K.
(1992) A new application of confocal laser scanning microscopy (C-LSM) to observe
subcellular organelles utilizing non fluorescent probe (osmium black). Histochem. J. 24, 550.
11. Robinson, J. M. and Batten, B. E. (1989) Detection of diaminobenzidine reactions using
scanning laser confocal reflectance microscopy. J. Histochem. Cytochem. 37, 1761–1765.
12. Itoh, J., Osamura, R. Y., and Watanabe, K. (1992) Subcellular visualization of light micro-
scopic specimens by laser scanning microscopy and computer analysis: a new application
of image analysis. J. Histochem. Cytochem. 40, 955–967.
PROTOCOL: CLSM Observations 179
13. Itoh, J., Sanno, N., Matsuno, A., Itoh, Y., Watanabe, K., and Osamura, R. Y. (1997)
Application of confocal laser scanning microscopy (CLSM) to visualize prolactin (PRL)
and PRL mRNA in the normal and estrogen-treated rat pituitary glands using non-fluorescent
probes. Microsc. Res. Tech. 39, 157–167.
14. Matsuno, A., Itoh, J., Osamura, R. Y., Watanabe, K., and Nagashima, T. (1999) Electron
microscopic and confocal laser scanning microscopic observation of subcellular organelles
and pituitary hormone mRNA: application of ultrastructural in situ hybridization and immu-
nohistochemistry to the pathophysiological studies of pituitary cells. Endocr. Pathol. 10,
199–211.
15. Itoh, J., Yasumura, K., Takeshita, T., et al. (2000) Three-dimensional imaging of tumor
angiogenesis. Anal. Quant. Cytol. Histol. 22, 85–90.
16. Itoh, J., Kawai, K., Serizawa, A., Yasumura, K., Ogawa, K., and Osamura, R. Y. (2000)
A new approach to three-dimensional reconstructed imaging of hormone-secreting cells
and their microvessel environments in rat pituitary glands by confocal laser scanning
microscopy. J. Histochem. Cytochem. 48, 569–578.
17. Matsuno, A., Ohsugi, Y., Utsunomiya, H., et al. (1994) Ultrastructural distribution of growth
hormone, prolactin mRNA in normal rat pituitary cells: a comparison between preembedding
and postembedding methods. Histochemistry 102, 265–270.
18. Matsuno, A., Teramoto, A., Takekoshi, S., et al. (1994) Application of biotinylated oligonu-
cleotide probes to the detection of pituitary hormone mRNA using Northern blot analysis,
in situ hybridization at light and electron microscopic levels. Histochem. J. 26, 771–777.
19. Matsuno, A., Ohsugi, Y., Utsunomiya, H., et al. (1995) Changes in the ultrastructural
distribution of prolactin and growth hormone mRNAs in pituitary cells of female rats after
estrogen and bromocriptine treatment, studied using in situ hybridization with biotinylated
oligonucleotide probes. Histochem. Cell Biol. 104, 37–45.
20. Matsuno, A., Nagashima, T., Osamura, R. Y., and Watanabe, K. (1998) Application of
ultrastructural in situ hybridization combined with immunohistochemistry to pathophysio-
logical studies of pituitary cell: technical review. Acta Histochem. Cytochem. 31, 259–265.
21. Matsuno, A., Nagashima, T., Ohsugi, Y., et al. (2000) Electron microscopic observation
of intracellular expression of mRNA and its protein product: technical review on ultrastruc-
tural in situ hybridization and its combination with immunohistochemistry. Histol. Histopa-
thol. 15, 261–268.
22. Matsuno, A., Nagashima, T., Osamura, R. Y., and Watanabe, K. (2000) Electron microscopic
in situ hybridization and its combination with immunohistochemistry, in Molecular Histo-
chemical Techniques (Springer Lab Manual) (Koji, T., ed.), Springer, New York, pp. 204–
221.
23. Lakkakorpi, T. J., Yang, M., and Rajaniemi, H. J. (1994) Processing of the LH/CG receptor
and bound hormone in rat luteal cells after hCG-induced down-regulation as studied by a
double immunofluorescence technique in conjunction with confocal laser scanning micros-
copy. J. Histochem. Cytochem. 42, 727–732.
24. Rummelt, V., Gardner, L. M., Folberg, R., et al. (1994) Three-dimensional relationships
between tumor cells and microcirculation with double cyanine immunolabeling, laser scan-
ning confocal microscopy, and computer-assisted reconstruction: an alternative to cast
corrosion preparations. J. Histochem. Cytochem. 42, 681–686.
25. Strong, L. H. (1964) The early embryonic pattern internal vascularization of the mammalian
cerebral cortex. J. Comp. Neurol. 123, 121–138.
26. Conradi, N. G., Engvall, J., and Wolff, J. R. (1980) Angioarchitectonics of rat cerebral
cortex during pre- and postnatal development. Acta Neuropathol. (Berl.) 50, 131–138.
27. Elias, K. A. and Weiner, R. I. (1984) Direct arterial vascularization of estrogen-induced
prolactin-secreting anterior pituitary tumors. Proc. Natl. Acad. Sci. USA 81, 4549–4553.
28. Kimura, K., Tojo, A., Nanba, S., Matsuoka, H., and Sugimoto, T. (1990) Morphometric
180 Matsuno et al.
11
Polymerase Chain Reaction
INTRODUCTION
The polymerase chain reaction (PCR) was introduced in 1985 by Saiki et al. (1),
who recognized that sequence-specific oligonucleotides could be used to prime synthesis
of complementary DNA strands by any of several DNA polymerases and that these
primers could be chosen so as to replicate both strands of the DNA positioned between
their cognate sequences in the template (1,2). Initial studies with the technique were
laborious and thus limited in scope. Thermostable DNA polymerases were not yet
commercially available, so the experimenter had to add fresh polymerase after the
DNA denaturing step of each cycle (1,2). The process was not automated, so a student
or technician was needed to transfer the reactions between water baths or heating blocks
set for each temperature of the cycle. The original report was largely a “proof of
principle” study in which a portion of the human β-globin gene was amplified and
then analyzed for the presence of normal versus sickle cell coding sequences (1,3). In
1988, the use of a thermostable DNA polymerase, isolated from Thermus aquaticus,
was used in this procedure instead of the Escherichia coli Klenow fragment (4,5). This
modification constituted the major advance needed to make the technique practical and
commercially viable. Various preparations of heat-stable polymerases from T. aquaticus,
as well as other thermophilic bacteria, were soon commercially available, as were
several models of programmable instruments for rapid and reliable heating and cooling
of samples. The utility of PCR was immediately recognized by molecular biologists,
geneticists, and scientists developing diagnostic procedures for human genetic diseases
and forensic investigations. The ability to amplify directly and analyze specific portions
of individual genes from only a few cells or from archival tissue has revolutionized
the field of pathology as well as other clinical disciplines that involve genetic analysis.
The importance of PCR was acknowledged by the scientific community at large when
Kary Mullis was awarded the 1993 Nobel Prize in Chemistry for his contribution to
its development.
This chapter describes the methodology of designing and executing PCR assays,
including typical protocols of nucleic acid extractions from fresh and paraffin-embedded
tissue, as well as various controls that are necessary and problems that are typically
encountered. Examples of typical protocols of specific applications of PCR to diagnose
181
182 Nikiforov and Howles
GENERAL PRINCIPLES
The main principal of PCR is an exponential amplification of a target DNA sequence
that results in the generation of millions of identical copies of a DNA fragment starting
from just a few copies. With the addition of a reverse transcription step, PCR can also
be applied to amplify RNA fragments. The assay is simple in general design and is
summarized in Fig. 1. Short, unique segments of DNA that are positioned on either
side of the sequence of interest are chosen. These pieces of DNA, the primers, are
typically 20–25 nucleotides long. The length of DNA to be amplified, the template,
varies from as little as 100 to a few thousand nucleotides. The primers are synthesized
such that when the pair forms hybrids with their cognate sequences in the template,
their 3′ ends are oriented toward each other. As the reaction proceeds, the DNA
polymerase elongates the primers, with the result that an exact copy is made of the
template DNA between the two primers. The DNA fragments are denatured (made
single-stranded) by heating the reactions to 94°C. Upon cooling to the appropriate
temperature (see below), primers anneal to the appropriate sequence, either the original
template or one of the copies. The temperature is raised to 72°C (the optimum tempera-
ture for the common thermostable polymerases) to allow DNA synthesis, and the process
is repeated. In an ideal reaction, a sequence of 30 cycles of polymerization results in
approx 1 × 109 copies of the original template.
DNA in high salt buffers with isopropanol. A number of commercial kits offer conve-
nient alternatives, minimizing handling and exposure to chemicals, offset by a
higher cost.
Method
1. Harvest fresh tissue using sterile scissors or disposable blade and immediately freeze in
liquid nitrogen. Store at −70°C until use.
2. Transfer 100 µg of tissue to a sterile 1.5-mL microcentrifuge tube and homogenize with a
184 Nikiforov and Howles
Table 1
Effect of Fixatives on Nucleic Acid Preservation in Paraffin Blocks
Good fixatives Poor fixatives
10% buffered formalin B-5
Ethanol Bouin’s
Acetone Zenker’s
Omnifix Glutaraldehyde
Carnoy’s
If a stock of DNA is required for multiple PCR reactions, the extraction can be started
with 10 sections, 40 µm thick; after deparaffinization (steps 1–6), the digestion and phenol-
chloroform extraction can be performed as described above for DNA isolation from fresh
or frozen tissue (steps 3–12).
Primer Design
The choice of primer sequences is a complex process and depends partly on the
specific application. A major factor is the constraints imposed by the nature of the
target sequence (the segment to be amplified) and the purpose of the assay. If the goal
is to detect a polymorphism that consists of a short (≤50 nucleotides) insertion or
deletion, the segment to be amplified should be relatively short, 10 times the difference
in length, at most. If point mutations are to be detected by sequence analysis or restriction
enzyme analysis, the primers need to be positioned to optimize these assays. If mRNA
is being assayed, amplifying across an intron/exon junction ensures that the assay will
not generate a false-positive from contaminating genomic DNA.
The standard primer length is 20 nucleotides, although much longer ones are used
successfully. Twenty nucleotides provide enough sequence complexity to prevent ran-
dom priming but not so much as to permit formation of stable hybrids with partial
homology. When possible, the cytosine + guanine (GC) content in each primer should
be similar and between 45% and 65%, so that melting temperatures (Tm, see formula
below) are moderate and similar for the pair of primers. Since the annealing temperature
for a primer is approx 5°C below its Tm, if the GC contents of a primer pair differ
markedly, the annealing temperature needed for the more AT-rich primer may permit
significant mispairing of the GC-rich primer, resulting in nonspecific amplification.
Conversely, the temperature needed for the GC-rich primer may preclude annealing
of the AT-rich primer. For example:
Tm (in °C) = 81.5 + 16.6 × log[cation] + 0.41 × %(GC) − 500/length
Tm for 20-mer, 50% GC, 0.1 M cation = 60°C
Tm for 20-mer, 65% GC, 0.1 M cation = 67°C
Primer sequences should be free of intra- and intermolecular homology that could
result in stable stem-loop structures or interstrand hybrids forming between two strands
of the same primer or between one strand from each primer of the pair. The presence
of such structures decreases the efficiency of the reaction in each cycle since the
effective primer concentration is decreased. Such hybrids can be especially troublesome
if they result in a recessed 3′ end since it will be extended by the polymerase, corrupting
the primer(s). This corruption is amplified with each cycle, resulting in failure of the
PCR assay. Other motifs to be avoided are simple sequences such as long mononucleo-
tides (e.g., GGGGG) or dinucleotide repeats (e.g., ATATATAT) that are not flanked
by a more complex sequence. It is desirable for the 3′ terminal nucleotide to be a G
or C. The extra hydrogen bond helps stabilize the end of the hybrid, resulting in more
efficient initiation of DNA synthesis by the polymerase. This is especially true if the
primer is AT rich (≥50%).
These factors constitute a useful set of guidelines for primer design. Currently, a
number of computer programs are available to assist in primer design. However, each
primer and each primer pair must be tested empirically. It is not uncommon for primers
PROTOCOL: PCR 187
to appear satisfactory on paper but to fail in amplification reactions. Below are examples
of successful primer pairs. (All Tms are calculated based on a 0.1 M cation concentration.)
Template
on an ethidium bromide-stained agarose gel. More cycles may be needed (35 cycles
for 105 initial copies) to achieve a visible band when the template is from partially
degraded DNA isolated from fixed tissues. More cycles may also be needed when
trying to detect rare events or ones of unknown frequency, such as mutations in a
subset of cells from a tissue. However, as a general rule, the number of cycles in
one PCR should not exceed 35 to avoid problems with background and nonspecific
amplification.
Reaction Components
Reactions should be assembled on ice. Several recipes exist for PCR reactions, but
all consist of basic components (see below) that are readily available and easily prepared
in any laboratory. A volume of reaction mix sufficient for all samples, and consisting
of all components except template DNA, is prepared and then aliquoted into the reaction
tubes. The DNA is added just before the tubes are placed in the thermocycler. The
final volume of reactions varies from 20 to 100 µL and should not be changed once
successful conditions are established.
1. Buffer and salt consist of 10 mM Tris-HCl, pH 8.3, and 50 mM KCl. Tris is used as the
pH buffer, and KCl is important for optimal enzyme activity; however, it is inhibitory at
concentrations above 50 mM. Many companies supply a 10× or 5× stock of reaction buffer
containing these components, which is further optimized for their enzyme preparation or
which contains a proprietary PCR-enhancing component. Typically, they offer either magne-
sium-free buffer or one that yields 1.5 mM final of MgCl2 concentration.
2. Magnesium (MgCl2) is added at a final concentration of 1.5–6 mM. The optimum [Mg+2]
must be determined for each primer set. Control reactions are performed with [Mg+2] varying
in 1- or 2-mM increments followed by 0.5-mM steps above and below the concentration(s)
giving a positive result.
3. Primer concentrations are usually 1–5 µM each. This concentration ensures that primers will
always be at least 10-fold in excess relative to template, even at the end of the PCR. The
best concentration must be determined empirically, as for MgCl2. Since divalent cations
complex with DNA readily, the primer (and template) concentration can change the effective
Mg+2 concentration. Thus, it is often necessary to retitrate Mg+2 requirements if the primer
concentrations are changed dramatically.
4. Nucleotide concentrations (dATP, dCTP, dGTP, dTTP) vary from 50 to 200 µM each. As
for primers, the goal is that nucleotides not become a limiting reagent. Too great an excess
will effect the Mg+2 requirement as above, however.
5. Taq DNA polymerase is available from many suppliers of biotechnology reagents. It is
usually supplied at 5 U/µL and is used at 1–2.5 U per reaction. One unit is defined as the
incorporation of 10 ηmol of deoxynucleotide into acid-precipitable DNA in 30 min at 74°C.
6. Template DNA is added last. The amount varies according to the quality of the template
and purpose of the assay. A typical amount is 20–50 ng of high-quality genomic DNA or
100 ng of DNA isolated from paraffin-embedded tissue.
7. Mineral oil (only enough to cover the reaction mix) is added on top to prevent evaporation
in some thermocyclers. However, most newer models have a heated lid, which prevents
condensation and obviates the need for oil.
8. Optional components include glycerol or dimethylsulfoxide (DMSO) used at 5–10% to
reduce background and lower the Tm of GC-rich primers. These agents change the dielectric
properties of aqueous solutions and so affect base pairing of DNA. However, DMSO partially
inhibits the polymerase and should be used at the minimum concentration needed. The
amount of enzyme or the elongation time is increased when DMSO is used.
190 Nikiforov and Howles
Below is an example of a typical reaction mix for 20 samples (18 samples plus a
positive and negative control) based on a 30 µL reaction volume. Note that you should
prepare enough mix for 21 tubes to accommodate pipetting losses.
Mix and put 28 µL into each reaction tube. Add 50 ng of template DNA to each tube
in a total volume of 2 µL and proceed with the assay.
Fig. 2. Analysis of a 236-bp PCR product by electrophoresis in a standard 1.5% TBE agarose
gel. Ten microliters of each product (30 cycles of amplification starting from 50 ng high-quality
genomic DNA) were loaded into the wells. The gel was run at 10 V/cm for 60 min and washed
in 0.5 µg EthBr/mL for 20 min. L, 1 Kb Plus DNA Ladder (Gibco BRL); lanes 1–6, PCR
products; N, “no template” negative control.
lane per gel must be reserved for size standards, typically a 100-bp ladder, which are
available from most biotech supply companies.
Polyacrylamide gels may be used to resolve and detect small fragments (≤200 bp).
These gels are typically 10% acrylamide (plus 0.3% bis-acrylamide) and are run in
TBE. In this system, diffusion of small fragments is minimal, so resolution is greater
and smaller amounts of DNA can be detected. Using an acrylamide gel, differences
of as little as 10 bp are easily distinguished. Thus, they must be used for detecting
restriction site polymorphisms and microsatellite length polymorphisms in amplified
sequences where size differences are often small.
the last few cycles, since 50% of the strands can be labeled per cycle (50% labeled after
one cycle, 75% labeled after two cycles, and so forth).
should be made to 100 µM, divided into small aliquots of 20–50 µL each, and stored
at −20°C. Multiple freeze/thaw cycles degrade primers, and it is best to discard any
remainder after five uses of an aliquot. A similar approach of freezing aliquots is needed
for deoxynucleotides, which are available in very pure form as 100-mM stocks. Although
simple chemicals like KCl and Tris need not be aliquotted and frozen, their purity and
sterility must be ensured. Water used to make reagents and reaction mixes should also
be the highest quality available and unopened since being autoclaved. It is a good
practice to make a 5× or 10× buffer mix that is then divided and frozen as above.
These preparations, although tedious at first, allow one to use fresh reagents that are
identical in composition.
Contamination Control
Preventing contamination of reagents and equipment is necessarily an obsession in
laboratories engaged in extensive use of PCR. The capacity to amplify a single copy
of DNA to a routinely detectable level is a serious matter. The best prevention protocol
begins by identifying for the technologist the most likely sources and modes of transfer
of contaminating DNA so that their diligence is appropriately focused. Make sure they
understand how small an amount of DNA is needed to generate serious contamination.
Provide careful instruction for pipettor, tip, and tube handling to prevent generation of
aerosols and transfer of DNA residue.
PROTOCOL: PCR 195
1. Have one complete set of pipettors, tube racks, and solutions, as well as microcentrifuge,
that are used only for preparing the PCR reactions and another set that is used only for
manipulating samples after amplification and/or for manipulating cloned DNA of the target
sequence. Most contamination comes from using a pipettor to prepare PCR reactions after
it has been used to handle amplified or cloned DNA.
2. Have one lab bench or area designated as the PCR bench. This bench should be isolated
from areas where amplified products are manipulated. Many labs designate a fume hood or
laminar flow hood as the PCR bench or have a separate small room for the purpose. This
certainly isolates and defines the area but is probably overkill. If a hood is used, it must be
configured so all manipulations can be performed comfortably. A tired worker makes mis-
takes.
3. Regularly clean pipettors. There are a variety of procedures used for this purpose, some
probably more extreme than needed. The simplest cleaning method, and one that is fully
effective, is to disassemble the instrument and wash the barrel, plunger, Teflon parts, and O-
rings in mild detergent. Soap and water is the best cleaning agent. There are also commercially
available solutions marketed specifically for this purpose that are effective. Ethanol (70%)
is often used to sterilize surfaces but is not the best cleaning agent for this purpose. Ethanol
will precipitate DNA but not necessarily flush it away and can actually be counterproductive
in this case. Dilute bleach can be used, but it is harsh and will shorten the life of some parts
of the instrument. After cleaning, rinse the parts well in distilled water and then rinse with
70% ethanol to enhance drying. Avoid 100% ethanol. It removes plasticizers and reduces
the life of the instrument.
4. Aerosol-resistant tips are widely used and greatly reduce the risk of contaminating pipettor
barrels. Although they add significantly to the cost of each assay, they may be advisable for
the laboratory using PCR extensively as a research or diagnostic tool.
5. PCR workstations can be purchased or constructed as an alternative to a separate bench or
room. Essentially, an area is enclosed in Plexiglas with doors that open toward the user.
Racks, pipettors, and other equipment designated for PCR are kept inside, and all reactions
are assembled inside the enclosure. When not in use, ultraviolet lights are turned on to
decontaminate exposed surfaces. Pipetters still require attention, however, since only one
side can be exposed and the insides of the barrels are not exposed at all.
6. When preparing and aliquotting reagent stocks such as buffers, primers, and nucleotides,
use freshly autoclaved solutions and tubes and freshly cleaned pipettors.
7. To minimize losses due to contamination and to minimize its spread, it is important to divide
all reagents into small aliquots. By doing so, if a contamination occurs, it is a simple and
inexpensive matter to discard all current reagents and open fresh aliquots.
Limitations
The size of the fragment that can be amplified by PCR is limited in two ways.
Although it is theoretically possible to amplify several thousand base pairs, factors
intrinsic to the DNA sequence may reduce this length by an order of magnitude. Such
factors could be secondary structure—such as direct or inverted repeats—high GC
content, or regions of high AT content. These limitations may be overcome by using
modified buffers and reaction conditions.
A limitation more likely to be encountered by pathologists is that due to sample
integrity. The amount of tissue degradation that occurs before a sample is fixed as well
as chemical degradation that occurs during fixation and tissue processing can reduce
the average length of the fragments into which the genome has been broken. As
discussed above, shorter target sequences (≤200 bp) are more likely to be successfully
amplified than longer ones. Longer amplifications may be successful, but the amount
196 Nikiforov and Howles
of template DNA will have to be increased to increase the likelihood of having enough
genome copies with the desired region intact.
The error rate of Taq polymerase is about 2/105 nucleotides (23). Although very
low, this level of infidelity is enough to introduce several mutations during the course
of an amplification. Such mutations do not adversely effect most applications since
only a small proportion of the copies will be affected. Obviously, mutations introduced
in early cycles will be represented at a greater frequency than those introduced in later
cycles. If amplified fragments are to be cloned and sequenced, however, it is important
to analyze several isolates to be sure that the reported sequence is accurate.
The fidelity of amplification by PCR can be increased by maximizing the annealing
temperature, minimizing the MgCl2 and dNTP concentrations, minimizing the annealing
and extension times, and using DNA polymerase with proofreading activity (24).
Other Considerations
Importance of Initial Cycles
The fidelity and efficiency of the first 5–10 cycles critically affect the outcome of
the PCR. Any mispriming or mispairing that occurs in these cycles will be amplified
throughout the reaction. Errors occurring in later cycles affect a comparatively small
number of copies. Thus, if the annealing temperature is too permissive or if mispairing
with some extension has occurred during sample preparation, aberrant bands or a general
background smear of DNA may be produced.
Time Extensions for Later Cycles
During later cycles of the PCR, the concentrations of primers and nucleotides
decrease, and there may be some loss of enzyme integrity. The combined result is that
copying may not be completed in the designated period of time, resulting in decreased
yield and a lower signal-to-noise ratio in the final product. This problem can be overcome
by lengthening the synthesis step of later cycles. Most thermocyclers have options for
incremental extension of individual steps (e.g., 2 a cycle) to overcome this problem.
Often, a final synthesis step of 5 min is added to extend any incomplete strands to
full length.
Special Reaction Buffers and Amplification Kits
Kits are available from various biotech supply companies. Often they tout a “one
mix fits all” buffer system to which you need only add your primers and template
DNA. Many of these kits work well. Their main drawback is that if they quit working
or fail to work for a particular application, the user has difficulty trouble-shooting the
PCR because a description of the components of the kit and their concentrations is
rarely provided by the manufacturer. In the end, the investigator may have to develop
the assay from the basics anyway.
RNA Extraction
It is important to note that RNA is less stable than DNA and is susceptible to rapid
degradation by RNAases that are present in cells, tissue, and the environment. Therefore,
all steps of RNA handling should be performed using appropriate precautions. Clean
gloves should be worn at all times to avoid introduction of RNAases from hands.
Sterile, disposable nuclease-free plasticware should be used whenever possible. A set
of solutions should be reserved for RNA work only. Water and most solutions should
be treated with diethylpyrocarbonate (DEPC) to inhibit RNAase activity. (Add 0.1 mL
DEPC/100 mL, leave overnight, and autoclave.) Autoclaving inactivates the DEPC, so
avoid subsequent contamination by RNAases. Tris solution should not be treated with
DEPC; instead it should be prepared using DEPC water and powder that is certified
as nuclease free.
RNA Extraction from Fresh or Frozen Tissue
Fresh tissue represents the best source of good-quality RNA. If tissue samples or
cells from which RNA is to be extracted cannot be processed immediately, they should
be snap frozen in liquid nitrogen and stored in liquid nitrogen or at −70°C.
RNA can be isolated by many methods. Many commonly used methods use the
guanidine thiocyanate/acid phenol method as described by Chomczynski and Sacchi
(25). In this isolation procedure, tissue is lysed in the presence of guanidine thiocyanate,
a strong protein denaturant, and a mild detergent. Under acidic conditions (pH 4), in
the presence of 4 M guanidine thiocyanate and phenol, RNA remains soluble, whereas
most proteins and small fragments of DNA will be found in the organic (phenol) phase
and larger DNA fragments and some proteins remain in the interface. The method
provides nearly total recovery of RNA from small quantities of cells or tissue. Several
commercial kits offer a ready-to-use phenol-guanidine thiocyanate solution for the
single-step method. Some of them are TRIzol Reagent (Life Technologies), TRI reagent
(Molecular Research Center), RNAzol B (Cinna Scientific), or Tri-Pure Isolation
Reagent (Boehringer Mannheim).
Method
1. Homogenize tissue sample in 1 mL of TRIzol reagent per 100 mg of tissue in a glass Teflon
homogenizer or in a Falcon 2063 tube using a power homogenizer.
2. Incubate for 5 min at room temperature. Remove insoluble materials by centrifugation at
12,000g for 10 min at 4°C. Transfer the supernatant into a new tube.
3. Add 0.2 mL of chloroform per 1 mL of TRIzol reagent, shake the tube vigorously by hand
for 15 s and incubate at room temperature for 3 min.
4. Centrifuge the sample at no more than 12,000g for 10 min at 4°C. The mixture should be
separated into a lower phenol-chloroform phase (red), an interphase, and an upper aqueous
phase (colorless). RNA remains exclusively in the aqueous phase.
5. Transfer the upper aqueous phase to a 1.5-mL microcentrifuge tube and precipitate with
an equal volume of isopropanol for 10 min at room temperature.
6. Centrifuge at no more than 12,000g for 10 min at 4°C. Discard the supernate.
7. Wash the RNA pellet once with cold 70% ethanol. Mix pellets with ethanol by vortexing,
and spin at no more than 7500g for 5 min at 4°C.
198 Nikiforov and Howles
8. Pipet off the ethanol, and briefly dry the RNA pellet (air-dry or vacuum-dry) for 5–10 min.
Avoid overdrying the pellet.
9. Dissolve RNA in 50 mL of DEPC water by passing the solution a few times through a
pipet tip and incubating for 10 min at 55°C.
10. Quantitate the RNA concentration by spectrophotometry (in a 1-cm-long light path, [RNA] =
A260 × 40 µg/mL × dilution factor) and store at −70°C. Use 1–5 µg for the RT reaction.
*Alternatively, after digesetion with proteinase K (step 7), isolation of RNA can be
performed in a single step using TRIzol reagent or equivalent as follows:
8. Divide the contents of each tube into two separate 1.5-mL microcentrifuge tubes. Add to
each tube 1 mL of TRIzol reagent and 200 µL of chloroform; shake tubes vigorously by
hand for 15 s and incubate at room temperature for 3 min.
9. Spin at no more than 12,000g for 10 min at 4°C. Transfer the upper aqueous phase to a
new tube.
10. Add 2 µL of glycogen to each tube. Precipitate RNA with 0.1 vol of 3 M sodium acetate
and 0.8 vol of isopropanol at −20°C overnight or at −70°C for 2 h.
11. Spin at 12,000g for 15 min at 4°C, wash pellet with 70% ethanol, and air-dry or vacuum-
dry pellet (avoid overdrying).
12. Dissolve pellets in 15 µL of DEPC water. Store at −70°C. Use 5 µL for each RT reaction.
in the extracts are templates for first-strand cDNA synthesis. This is particularly important
when the integrity of RNA is compromised or unknown. This is the only method to prime
RNA isolated from fixed tissues.
2. Oligo(dT) primers (12–18 nucleotides) bind to 3′ polyA tails of mRNA to prime synthesis
of full-length cDNAs. This is a more specific priming method to prime selectively mRNA
that constitutes only 1–2% of total RNA. The amount of cDNA is considerably less, but the
complexity is greater (ribosomal RNA is not copied) than when random hexamers are used.
This method may give an RT-PCR product more consistently.
3. Gene-specific primer (GSP) contains a sequence of the target cDNA and therefore primes
selective transcription of the RNA of interest. This is the most specific priming method;
however, some GCPs fail to prime first-strand cDNA synthesis even though they function
in PCR of a DNA template.
The protocol for reverse transcription of RNA with random hexamer primers using the
Gibco-BRL kit is as follows:
1. Prepare RNA/primer mixtures in a sterile, nuclease-free 0.5-mL microcentrifuge tube as
follows:
1–5 µg total RNA (or 5 µL of RNA isolated from paraffin-embedded tissue)
random hexamers (50 ng/µL) 4 µL
10 mM dNTP mix 1 µL
DEPC water to 10 µL
2. Incubate at 65°C for 5 min and place on ice for at least 1 min.
3. Prepare the reaction mixture for each sample as follows:
10× RT buffer 2 µL
25 mM MgCl2 4 µL
0.1 M DTT 2 µL
RNAase Inhibitor 1 µL
4. Add 9 µL of reaction mixture to each RNA/primer mixture, mix, centrifuge briefly, and
incubate at 25°C for 10 min.
5. Add 1 µL (50 U) of SuperScript II reverse transcriptase (Gibco-BRL), mix, and incubate at
25°C for 10 min.
6. Transfer to 42°C and incubate for 50 min.
7. Terminate the reaction at 70°C for 15 min. Chill on ice. Store cDNA at −20°C. Use 2 µL
for PCR.
PCR Step
After the reverse transcription reaction is complete, the amplification of cDNA by
PCR is not different from amplification of double-strand DNA templates. All the general
and technical considerations for PCR apply to RT-PCR as well. However, it is important
to note that RT-PCR is a method that detects the expression of a gene, not its presence
in the genome. Therefore, in some situations it may be important to ensure that PCR
is amplifying cDNA and not genomic DNA molecules that can be present in RNA
preparations. This can be achieved by designing primers for cDNA amplification that
are complementary to different exons; the amplification will be cDNA specific because
many introns are too long to be amplified by standard PCR, or it will result in a much
larger size of PCR product. The risk of amplification of contaminant DNA can be
eliminated by treating RNA extracts with DNAase I prior to the RT reaction.
To increase chances for successful amplification of RNA targets extracted from
paraffin-embedded tissue, the smallest possible target size should be selected—typically
PROTOCOL: PCR 201
in a range of 80–200 bp. Sensitivity of the PCR can be enhanced by Southern blotting
of the product and hybridization with a radiolabeled internal oligonucleotide probe
(28). However, the use of a second round of PCR amplification or nested RT-PCR
should be discouraged because of a high risk of cross-contamination or nonspecific
amplification that would lead to false-positive results.
The important controls for each RT-PCR reaction are as follows: a negative control
at the reverse transcription step in which the RT enzyme is omitted (a control for the
cDNA specificity of the reaction); a standard PCR “no template” negative control; and
a positive control in which RNA extracted from tissue with known expression of the
target gene is added to the RT reaction. In addition, for each sample studied by RT-
PCR for expression of a particular mRNA target, an amplification control is necessary
to ensure that the template RNA is of sufficient integrity to allow the amplification of
the target product. Usually, a portion of an ubiquitously expressed gene of a similar
or larger size should be amplified by PCR using the same cDNA template to ensure
that amplification is possible. Examples of such control reactions are amplification of
phosphoglycerate kinase cDNA (29) or N-RAS cDNA (see below).
A positive result can be concluded only if the negative controls are clean. A negative
result can be interpreted as absence of a target mRNA fragment in a specimen only
when both conditions are met: 1) the expected level of amplification is detected in the
positive control, and 2) the template mRNA showed successful amplification of another
similar-size RNA fragment (positive amplification control). All other results must be
interpreted as inconclusive and should be repeated.
QUANTITATIVE PCR
One of the most exciting applications of PCR for the molecular biologist has been
its use to quantitate the amount of starting DNA and, especially, RNA in a reaction.
The important areas of clinical application of quantitative PCR are the assessment of
residual disease after treatment of leukemia or lymphoma, the detection of viral load,
and the assay for gene amplification or for loss of heterozygosity.
The principle behind quantitative PCR is simple. Reactions are “spiked” with a
known amount of standard DNA of some definition. The standard is amplified at the
same time as the test sequence, the PCR products are fractionated by the usual methods,
the amount of amplified test DNA is compared with the amount of amplified standard
DNA, and the number of initial test copies is calculated based on the relative signal
intensities. Typically, a standard curve is generated using different initial numbers of
standard DNA molecules to correct for nonlinear standard curves due to decreasing
efficiency with increasing cycle number. Initial experiments are required to establish
the number of cycles through which amplification remains logarithmic as well as the
lower limit of sensitivity.
Choice of standards is a complicated process. It is best to use the same primer set
so the PCR conditions are optimal for both test and standard. This requires that the
standard curve be generated in separate reaction tubes (30). If, however, there are
factors in the test samples that inhibit PCR or if there is tube-to-tube variation, the
standard curve will not adequately represent the test reactions. Standards can be ampli-
fied in the same tubes as the samples by constructing a sequence that is slightly different
202 Nikiforov and Howles
in size (50–100 bp) than the test sequence when amplified. Engineering a restriction
site polymorphism (31), an insertion (32), or a deletion into a clone of the test sequence
is used in a variation termed competitive PCR (31,32). A clever variation of this
principle whereby the same standard DNA can be used for a set of genes (primer pairs)
of interest has also been described (33).
The most accurate method—and the most expensive—is real-time PCR. With spe-
cially designed thermocyclers, attached optical systems, and fluorescently labeled prim-
ers or intercalating dyes, the amplification process is monitored in each tube as it
occurs. A complete description of this method is presented in another chapter of
this book.
APPLICATIONS OF PCR
The versatility of PCR has resulted in multiple PCR-based applications, the number
of which continues to increase rapidly. Many clinical diagnostic and research fields
take advantage of the high sensitivity and specificity of PCR to generate millions of
identical copies of specific portions of individual genes from only a few cells or from
archival tissue.
The PCR amplification followed by direct sequencing or another analytic method
is a common method to diagnose point mutations in genes associated with inherited
diseases or in oncogenes and tumor suppressor genes associated with different types
of human neoplasms. The RT-PCR, on the other hand, has become a standard method
for detection of specific chromosomal rearrangements that are common in soft tissue
sarcomas, hematopoetic malignancies, and some epithelial cancers. Another important
area of PCR application is in the detection of infections by amplification of a defined
region of DNA or reverse-transcribed RNA using primers unique for the pathogen.
Some of the practical applications of PCR were briefly discussed throughout this
chapter. Here, we describe in detail two protocols of the PCR-based screening and
detection of point mutations in a gene of interest and of detection of chromosomal
rearrangement using RT-PCR.
Fig. 4. SSCP analysis of PCR-amplified tumor DNA for mutations in exon 5 of the p53
gene. N, normal DNA (negative control); J6, tumor sample with known point mutation in p53
exon 5 (positive control); lanes 3–18, tumor samples. Samples C9 and C2 show abnormal
pattern of migration in the polyacrylamide gel, suggesting the presence of mutations.
Fig. 5. Sequence analysis of tumor sample C9, suspected to harbor a mutation by SSCP
analysis, reveals a heterozygous somatic mutation A→G in codon 160 of p53. N, normal
DNA sequence.
6. Detect the nature of the shifted band by sequencing. Precipitate with ethanol the PCR product
from samples that displayed a band shift by SSCP analysis and sequence using the fmol
sequencing kit (Promega, Madison, WI) by following the manufacturer’s instruction. The
sequence analysis of two positive samples from the example SSCP gel shown above revealed
a heterozygous somatic mutation of A→G in codon 160 of p53 in sample C9 (Fig. 5) and
C→T mutation in the third position of codon 182 in sample C2.
denaturation 30 s at 94°C
annealing 0 s at 59°C
elongation 1 min at 72°C (with final elongation for 5 min at 72°C)
5. Assemble another PCR to test the integrity of the RNA template. For this purpose, use the
same cDNA samples to amplify a 236-bp sequence of the c-N-RAS using a pair of primers
spanning intron 1 of the gene:
upstream primer 5′-ATGACTGAGTACAAACTGGT-3′
downstream primer 5′-AGGAAGCCTTCGCCTGTCCT-3′
The conditions of the assay, including optimal MgCl2 concentration and annealing tempera-
ture, are identical to those used for the ELE1/H4 amplification.
6. Load 10 µL of each PCR product onto a 1.5% agarose gel. Samples that show a 242-bp
band are considered to have the rearrangement (Fig. 6). Samples that show no amplification
for ELE1/RET, but successfully amplify N-RAS cDNA, are considered negative for the
rearrangement. Positive cases can be selectively sequenced to confirm the presence of a
chimeric transcript.
206 Nikiforov and Howles
REFERENCES
1. Saiki, R. K., Scharf, S. J., Faloona, F. A., et al. (1985) Enzymatic amplification of β-globin
genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science
230, 1350–1354.
2. Mullis, K. B. and Faloona, F. A. (1987) Specific synthesis of DNA in vitro via a polymerase-
catalyzed chain reaction. Methods Enzymol. 155, 335–350.
3. Engelke, D. R., Hoener, P. A., and Collins, F. S. (1988) Direct sequencing of enzymatically
amplified human genomic DNA. Proc. Natl. Acad. Sci. USA 85, 544–548.
4. Wong, C., Dowling, C. E., Saiki, R. K., Higuchi, R. G., Erlich, H. A., and Kazazian, H.
H. (1987) Characterization of β-thalassemia mutations using direct genomic sequencing of
amplified single copy DNA. Nature 330, 384–386.
5. Saiki, R. K., Gelfand, D. H., Stoffel, S., et al. (1988) Primer-directed enzymatic amplification
of DNA with a thermostable DNA polymerase. Science 239, 487–491.
6. Greer, C. E., Peterson, S. L., Kiviat, N. B., and Manos, M. M. (1991) PCR amplification
from paraffin-embedded tissues: effects of fixative and fixation time. Am. J. Clin. Pathol.
95, 117–124.
7. Shibata, D. (1994) Extraction of DNA from paraffin-embedded tissue for analysis by
polymerase chain reaction: new tricks from an old friend. Hum. Pathol. 25, 561–563.
8. Ben-Ezra, J., Johnson, D. A., Rossi, J., Cook, N., and Wu, A. (1991) Effect of fixation on
the amplification of nucleic acids from paraffin-embedded material by the polymerase chain
reaction. J. Histochem. Cytochem. 39, 351–354.
9. Greer, C. E., Lund, J. K., and Manos, M. M. (1991) PCR amplification from paraffin-
embedded tissues: recommendations on fixatives for long-term storage and prospective
studies. PCR Methods Appl. 1, 46–50.
10. Shibata, D., Martin, W. J., and Arnheim, N. (1988) Analysis of DNA sequences in forty-
year old paraffin-embedded thin-tissue sections: a bridge between molecular biology and
classic histology. Cancer Res. 48, 4564–4566.
11. Kallio, P., Syrjanen, S., Tervahauta, A., and Syrjanen, K. (1991) A simple method for
isolation of DNA from formalin-fixed paraffin-embedded samples for PCR. J. Virol. Meth.
35, 39–47.
12. Heller, M. J., Robinson, R. A., Burgart, L. J., TenEyck, C. J., and Wilke, W. W. (1992)
DNA extraction by sonication: a comparison of fresh, frozen, and paraffin-embedded tissue
extracted for use in polymerase chain reaction assays. Mod. Pathol. 5, 203–206.
13. Sepp, R., Szabo, I., Uda, H., and Sakamoto, H. (1994) Rapid techniques for DNA extraction
from routinely processed archival tissue for use in PCR. J. Clin. Pathol. 47, 318–323.
14. Frank, T. S., Svoboda-Newman, S. M., and Hsi, E. D. (1996) Comparison of methods for
extracting DNA from formalin-fixed paraffin sections for nonisotopic PCR. Diagn. Mol.
Pathol. 5, 220–224.
15. Birch, D. E. (1996) Simplified hot start PCR. Nature 381, 445–446.
16. Bassam, B. J., Caetano-Anolles, G., Gresshoff, P. M. (1991) Fast and sensitive silver
staining of DNA in polyacrylamide gels. Anal. Biochem. 196, 80–83.
17. Peats, S. (1984) Quantitation of protein and DNA in silver-stained agarose gels. Anal.
Biochem. 140, 178–182.
18. Liu, X. Y., Nelson, D., Grant, C., Morthland, V., Goodnight, S. H., Press, R. D. (1995)
Molecular detection of a common mutation in coagulation factor V causing thrombosis via
hereditary resistance to activated protein C. Diagn. Mol. Pathol. 4, 191–197.
19. Sutcharitchan, P., Saiki, R., Huisman, T. H., et al. (1995) Reverse dot-blot detection of the
African-American beta-thalassemia mutations. Blood 86, 1580–1585.
20. Kant, J. A., Mifflin, T. E., McGlennen, R., Pice, E., Naylor, E., and Cooper, D. L. (1995)
Molecular diagnosis of cystis fibrosis. Clin. Lab. Med. 15, 877–898.
PROTOCOL: PCR 207
21. Orita, M., Suzuki, Y., Sekiya, T., and Hayashi, K. (1989) Rapid and sensitive detection of
point mutations and DNA polymorphisms using the polymerase chain reaction. Genomics
5, 874–879.
22. Ainsworth, P. J., Surh, L. C., and Coulter-Mackie, M. B. (1991) Diagnostic single strand
conformation polymorphism (SSCP): a simplified non-radioisotopic method as applied to
a Tay-Sachs B1 variant. Nucleic Acids Res., 19, 405–406.
23. Lundberg, K. S., Shoemaker, D. D., Adams, M. W., Short, J. M., Sorge, J. A., Marthur,
E. J. (1991) High-fidelity amplification using a thermostable DNA polymerase isolated
from Pyrococcus furiosus. Gene 108, 1–6.
24. Cariello, N. F., Swenberg, J. A., Skopek, T. R. (1991) Fidelity of Thermococcus litoralis
DNA polymerase (Vent) in PCR determined by denaturing gradient gel electrophoresis.
Nucleic Acids Res. 19, 4193–4198.
25. Chomczynski, P. and Sacchi, N. (1987) Single-step method of RNA isolation by acid
guanidine thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156–159.
26. Foss, R. D., Guha-Thakurta, N., Conran, R. M., Gutman, P. (1994) Effect of fixative time
on the extraction and polymerase chain reaction amplification of RNA from paraffin-
embedded tissue. Diagn. Mol. Pathol. 3, 148–155.
27. Ladanyi, M. and Bridge, J. A. (2000) Contribution of molecular genetic data to the classifica-
tion of sarcomas. Hum. Pathol. 31, 532–538.
28. Nikiforov, Y. E., Rowland, J. M., Bove, K. E., Monforte-Munoz, H., and Fagin, J. A.
(1997) Distinct pattern of ret oncogene rearrangements in morphologic variants of radiation-
induced and sporadic thyroid papillary carcinomas in children. Cancer Res. 57, 1690–1694.
29. Argani, P., Zakowski, M. F., Klimstra, D. S., Rosai, J., and Ladanyi, M. (1998) Detection
of the SYT-SSX chimeric RNA of synovial sarcoma in paraffin-embedded tissue and its
application in problematic cases. Mod. Pathol. 11, 65–71.
30. Muthuchamy, M., Pajak, L., Howles, P., Doetschman, T., and Wieczorek, D. (1993) Devel-
opmental analysis of tropomyosin expression in embryonic stem cells and mouse embryos.
Mol. Cell. Biol. 13, 3311–3323.
31. Becker-Andre, M. and Hahlbrock, K. (1989) Absolute mRNA quantification using the
polymerase chain reaction (PCR). A novel approach by a PCR aided transcript titration
assay (PATTY). Nucleic Acids Res. 17, 9438–9446.
32. Gilliland, G., Perrin, S., Blanchard, K., and Bunn, F. (1990) Analysis of cytokine mRNA
and DNA: detection and quantitation by competitive polymerase chain reaction. Proc. Natl.
Acad. Sci. USA 87, 2725–2729.
33. Wang, A. M., Doyle, M. V., and Mark, D. F. (1989) Quantitation of mRNA by the
polymerase chain reaction. Proc. Natl. Acad. Sci. USA 86, 9717–9721.
PROTOCOLS
12
Detection of Nucleic Acids in Cells and Tissues
by In Situ Polymerase Chain Reaction
INTRODUCTION
The solution-based polymerase chain reaction (PCR) method for amplification of
defined gene sequences has proved a valuable tool not only for basic researchers but
also for clinical scientists. Using even a minute amount of DNA or RNA and choosing
a thermostable enzyme from a large variety of sources, one can enlarge the amount of
the gene of interest, which can be analyzed and/or sequenced. Thus genes or segments
of gene sequences present only in a small sample of cells or small fraction of mixed
cellular populations can be examined. However, one of the major drawbacks of the
solution-based PCR technique is that the procedure does not allow the association of
amplified signals of a specific gene segment with the histologic cell type(s) (1–24).
For example, it would be advantageous to determine what types of cells in the peripheral
blood circulation or in a pathologic specimens carry the human immunodeficiency virus
type 1 (HIV-1) gene or a vector used for gene therapy or an aberrant gene in a
leukemia patient and what percent of leukemia cells is present after various forms of
antitumor therapies.
The ability to identify under the microscope individual cells expressing or carrying
specific genes of interest in a latent form in a tissue section provides a great advantage
in determining various aspects of normal, as opposed to pathologic, conditions. For
example, this technique could be used in determination of viral burden (11), degree of
gene expression (13,21,22), or tumor burden (25,26), before and after therapy (24,27,32),
in lymphomas or leukemias, in which specific aberrant gene translocations are associated
with certain types of malignancy (28). In the case of HIV-1 or other viral infections,
one can determine the effects of therapy or putative antiviral vaccination by evaluating
the number of cells still infected with viral agent post chemotherapy or vaccination
(2,5,6,20). Similarly, one can potentially determine the preneoplastic lesions by examin-
ing tumor suppresser genes (i.e., p53 mutations associated with certain tumors or
oncogenes or other aberrant gene sequences known to be associated with certain types
of tumors) (26,28,29). In the area of diagnostic pathology, determination of origin of
metastatic tumors is a perplexing problem. By utilizing the proper primers for genes
209
210 Bagasra, Bobroski, and Amjad
expressed by certain histologic cell types, one can potentially determine the origin of
metastatic tumors by performing reverse transcriptase (RI)-initiated in situ PCR (26).
One can also determine the subcellular localization of a gene or virus inside the
cytoplasmic or nuclear compartments (2–7,12,14,18,27,30–32). Both RNA and DNA
signal can be detected simultaneously by utilizing primers that anneal to the splicing
junctions of exons of mRNA (8,10,13,16,17,21,22,25,33,34).
Our laboratories have been utilizing in situ PCR techniques for several years, and
we have developed simple, sensitive protocols for both RNA and DNA in situ PCR
that have proved reproducible in multiple double-blinded studies (2–7,10,17,18,22).
One can use this method for amplification of both DNA or RNA gene sequences. By
use of multiple labeled probes, one can detect various signals in a single cell (14,22).
In addition, under special circumstances, one can perform immunohistochemistry and
RNA and DNA amplification at a single cell level (so-called triple labeling) (10,23).
To date, we have successfully amplified and detected HIV, simian immunodeficiency
virus (SIV), human papillomavirus (HPV), hepatitis B virus (HBV), cytomegalovirus
(CMV), Epstein-Barr virus (EBV), human herpes virus (HHV)-6, HHV-8, herpes sim-
plex virus (HSV), and p53 and its mutations, as well as mRNA for surfactant protein
A, estrogen receptors, and inducible nitrous oxide synthesis (iNOS) gene sequences
associated with multiple sclerosis, by DNA and/or RNA (RT in-situ PCR). We have
used various tissues, including peripheral blood mononuclear cells (PBMCs), lymph
node, spleen, skin, breast, lungs, placenta, sperm, cytologic specimens, Kaposi’s sarcoma
(KS), cultured cells, numerous others formalin-fixed, paraffin-embedded tissues and
various cellular populations in the frozen sections of AIDS brains, infected with HIV-
1 (1–14,17,18,22,23,25,30). In the following pages, we present a detailed protocol
currently being utilized in our laboratory. For more details, the readers are referred to
ref. 33.
allowing for reaction chambers about 20 µm in height to form between. Second, the
Teflon border helps keep the nail polish from entering the reaction chambers when the
polish is being applied. This is important, because any leakage of nail polish into a
reaction chamber can compromise the results in that chamber. Even if one is using an
advanced thermal cycler with humidification, use of the Teflon-coated slides is still
recommended, as the hydrophobicity of the Teflon combined with the pressure applied
by a cover slip helps spread small volumes of reaction cocktail over the entire sample
region, without forcing much fluid out the periphery.
To prepare glass slides properly, follow this procedure:
1. Prepare the following 2% AES solution just prior to use: 3-aminopropyltriethoxysilane (AES;
Sigma A-3648) 5 mL and acetone 250 mL.
2. Put solution into a Coplin jar or glass staining dish and dip glass slides in 2% AES for 60
s. Allow to dry for 1–2 h at room temperature or overnight, if not in a hurry.
3. Dip slides five times into a different vessel filled with 1000 mL of distilled water.
4. Repeat step 3 three to four times, changing the water each time.
5. Air-dry in a laminar flow hood from a few hours to overnight, and then store slides in a
sealed container at room temperature. Try to use slides within 15 days of silanization; 250
mL of AES solution is sufficient to treat 200 glass slides.
PREPARATION OF TISSUE
Cell Suspensions
To use peripheral blood leukocytes, first isolate cells on a Ficoll-Hypaque density
gradient. Tissue culture cells or other single-cell suspensions can also be used. Prepare
all cell suspensions with the following procedure:
1. Wash cells with 1× phosphate-buffered saline (PBS) twice.
2. Resuspend cells in PBS at 2 × 106 cells/mL.
3. Add at least 10 µL of cell suspension to each well of the slide using a P20 micropipet, and
spread across surface of slides.
4. Air-dry slide in a laminar flow hood.
Paraffin-Fixed Tissue
Routinely fixed paraffin tissue sections can be amplified quite successfully. This
permits the evaluation of individual cells in the tissue for the presence of a specific
RNA or DNA sequence. For this purpose, tissue sections are placed on specially
designed slides that have single wells (described further in the Materials and Methods
section, page 224). In our laboratory, we routinely use placental tissues, central nervous
system (CNS) tissues, cardiac tissues, and so on, which are sliced to a 5–6 µm thickness.
212 Bagasra, Bobroski, and Amjad
Other laboratories prefer to use sections up to 10-µm thickness, but in our experience,
amplification is often less successful with the thicker sections, and multiple cell layers
can often lead to difficult interpretation due to superposition of cells. However, if one
is using tissues that contain particularly large cells—such as ovarian follicles—then
thicker sections may be appropriate.
1. Place tissue section on the glass surface of the slide.
2. Incubate the slides in an oven at 60–80°C (depending on type of paraffin used to embed the
tissue) for 1 h, to melt the paraffin.
3. Dip the slides in EM grade xylene solution for 5 min, and then in EM grade 100% ethanol
for 5 min. (EM grade reagents are benzene-free.) Repeat these washes two or three times,
to rid the tissue of paraffin completely.
4. Dry the slides in an oven at 80°C for 1 h.
Frozen Sections
It is possible to use frozen sections for in situ amplification; however, the morphology
of the tissue following the amplification process is generally not as good as with paraffin
sections. It seems that the cryogenic freezing of the tissue, combined with the lack of
paraffin substrate during slicing, compromises the integrity of the tissue. Usually, thicker
slices must be made, and the tissue “chatters” in the microtome. As any clinical
pathologist will relate, definitive diagnoses are made from paraffin sections, and this
rule of thumb seems to extend to the amplification procedure as well.
It is very important to use tissues that were frozen in liquid nitrogen or placed on
dry ice immediately after they were harvested before autolysis began to take place. If
tissues were frozen slowly by placing them at −70°C, then eventually ice crystals will
form inside the tissues, creating a gap that will distort the morphology.
To use frozen sections, use as thin a slice as possible (down to 4–6 µm), apply to
the slide, dehydrate for 10 min in 100% methanol (the exception to methanol is when
surface antigens are lipoprotein, which will denature in methanol; then use 2% PFA
or other reagent), and air-dry in a laminar flow hood. Then proceed with the heat
treatment described below.
Heat Treatment
Place the slides with attached tissue or cells on a heat block at 105°C for 0.5–2
min, to stabilize the cells or tissue on the glass surface of the slide. This step is absolutely
critical, and one may need to experiment with different periods to optimize the heat
treatment for specific tissues. Our laboratory routinely uses 90 s for DNA target
sequences and 30–45 s for RNA sequences.
PROTOCOL: In Situ PCR 213
Fig. 1. Overview of in situ PCR protocol: amplification and hybridization of DNA target
sequence. See text for details and modifications.
4. At this point, slides with attached tissue can be stored at −80°C until use. Before storage,
dehydrate with 100% ethanol.
If biotinylated probes or peroxidase-based color development are to be used, the
samples should further be treated with a 0.3% solution of hydrogen peroxide in PBS,
to inactivate any endogenous peroxidase activity. Once again, incubate the slides over-
night—either at 37°C or at room temperature. Then, wash the slides once with PBS.
If other probes are to be used, proceed directly to the following proteinase K digestion,
which is perhaps the most critical step in the protocol.
Proteinase K Treatment (The Most Rate-Limiting Step)
1. Treat samples with 6 µg/mL proteinase K in PBS for 5–60 min at room temperature or at
55°C. To make a proper solution, dilute 1.0 mL of proteinase K (1 mg/mL) in 150 mL of
1× PBS.
2. After 5 min, look at the cells under the microscope at 400×. If the majority of the cells of
interest exhibit uniform-appearing, small, round “salt and pepper” dots on the cytoplasmic
membrane, then stop the treatment immediately with step 3. Otherwise, continue treatment
for another 5 min, and reexamine.
3. After proper digestion, heat slides on a block at 95°C for 2 min to inactive the proteinase K.
4. Rinse slides in 1× PBS for 10 s.
5. Rinse slides in distilled water for 10 s.
6. Air-dry.
Optimizing Digestion
The time and temperature of incubation should be optimized carefully for each cell
line or tissue section type. Too little digestion, and the cytoplasmic and nuclear mem-
branes will not be sufficiently permeable to primers and enzyme, and amplification
will be inconsistent. Too much digestion, and the membranes will either deteriorate
during repeated denaturation or worse, signals will leak out. In the first case cells will
not contain the signal, and high background will result. In the last case, many cells
will show pericytoplasmic staining or “rim staining” representing the leaked signals
going into the cells not containing the signals. Attention to detail here can often mean
the difference between success and failure, and this procedure should be practiced
rigorously with extra sections before continuing on to the amplification steps.
In our laboratory, proper digestion parameters vary considerably with tissue type.
Typically, lymphocytes will require 5–10 min at 25°C or room temperature, CNS tissue
will require 12–18 min at room temperature, and paraffin-fixed tissue will require
15–30 min at room temperature. However, these periods can vary widely, and the
appearance of the “salt and pepper dots” is the important factor. Unfortunately, the
appearance of the “salt and pepper dots” is less prominent in paraffin sections.
The critical importance of these dots should not be underestimated, since a extra
2–3 min of treatment after the appearance of dots will result in leakage of signals.
An alternative to the observation of “dots” method is to select a constant time and
treat slides in varying amounts of proteinase K. For example, treat slides for 15 min
in 1–6 µg/mL of proteinase K.
RT Variation: In Situ RNA Amplification
One has two choices in order to detect an RNA signal. The first and more elegant
method is simply to use primer pairs that flank spliced sequences of mRNA, as these
PROTOCOL: In Situ PCR 215
particular sequences will be found only in RNA will be split into sections in the DNA
(see Fig. 3). Thus, by using these RNA-specific primers, one can skip the following
DNase step and proceed directly to reverse transcription.
The second, more brutal, yet often necessary approach is to treat the cells or tissue
with a DNase solution subsequent to the proteinase K digestion. This step destroys all
of the endogenous DNA in the cells so that only RNA survives to provide signals for
amplification.
Note: All reagents for RT in situ amplification should be prepared with RNase-free
water (i.e., diethylpyrocarbonate [DEPC]-treated water). In addition, the silanized glass
slides and all glassware should be RNase-free, which we ensure by baking the glassware
overnight in an oven before use in the RT amplification procedure.
DNase Treatment
Prepare an RNase-free DNase solution:
40 mM Tris HCl, pH 7.4.
6 mM MgCl2.
2 mM CaCl2.
100 U/µL final vol. of DNase (use RNase-free DNase, such as 1000 U/µL RQ1 DNase
cat. no. 776785 from Boehringer).
1. Add 10–15 µL of solution to each well.
2. Incubate the slides at 37°C in a humidified chamber for 1 h. If one is using liver tissue, this
incubation should be extended for an additional hour.
3. After incubation, rinse the slides with a similar solution that was prepared without the
DNase enzyme.
4. Wash the slides twice with DEPC-treated water.
Note: Some cells are particularly rich in ribonuclease; in this circumstance, add the
following ribonuclease inhibitor to the DNase solution: 1000 U/mL placental ribo-
nuclease inhibitor (e.g., RNAsin) plus 1 mM dithiotheitol (DTT). Also, some investiga-
tors prefer to use a long incubation period with a lower concentration of DNase (1 U/
mL for 18 h).
Reverse Transcriptase Reaction
Next, one wishes to make DNA copies of the targeted RNA sequence so that the
signal can be amplified. The following are typical cocktails for the RT reaction:
1. If using AMVRT or MMLVRT enzyme:
10× reaction buffer* 4.0 µL
10 mM dATP 2.0 µL
10 mM dCTP 2.0 µL
10 mM dGTP 2.0 µL
10 mM dTTP 2.0 µL
RNasin at 40 U/µL (Promega)** 0.5 µL
20 µM downstream primer 1.0 µL
AMVRT 20 U/µL 0.5 µL
DEPC water 6.0 µL
Total volume 20.0 µL
*5× reaction buffer: 250 mM Tris-HCl pH 8.3, 375 mM KCl, 15 mM MgCl2.
216 Bagasra, Bobroski, and Amjad
RT Enzymes
Avian myoblastosis virus reverse transcriptase (AMVRT) and Moloney murine leuke-
mia virus reverse transcriptase (MMLVRT) give comparable results in our laboratory.
Other RT enzymes will probably work also. However, it is important to read the
manufacturer’s descriptions of the RT enzyme and to make certain that the proper
buffer is used.
An alternative RT enzyme is available that lacks RNase activity. Called Superscript
II, it is available from BRL Lifesciences, and it is suitable for RT of long mRNAs. It
is also suitable for routine RT amplification, and in our laboratory it has proved to be
more efficient than the two enzymes described above.
Primers and Target Sequences
In our laboratory, we simply use antisense downstream primers for our gene of
interest, as we already know the sequence of most genes we study. However, one can
alternatively use oligo (dT) primers first to convert all mRNA populations into cDNA
and then to perform the in situ amplification for a specific cDNA. This technique may
be useful when one is performing amplification of several different gene transcripts at
the same time in a single cell. For example, if one is attempting to detect various
cytokine expressions, one can use an oligo (dT) primer to reverse transcribe all the
mRNA copies in a cell or tissue section. Then one can amplify more than one type of
cytokine and detect the various types with different probes that develop into different
colors (see Multiple Signals and Multiple Labels in Individual Cells, page 220.)
In all RT reactions, it is advantageous to reverse-transcribe only relatively small
fragments of mRNA (<1500 bp). Larger fragments may not completely reverse-
transcribe due to the presence of secondary structures. Furthermore, the RT enzymes—
AMVRT and MMLVRT, at least—are not very efficient in transcribing large mRNA
fragments. However, this size restriction does not apply to amplification reactions that
PROTOCOL: In Situ PCR 217
are exclusively DNA, for the polymerase enzyme copies nucleotides better. In in situ
DNA reactions, we routinely amplify genes up to 300 bp.
The following are several additional points one should keep in mind:
1. The length for both sense and antisense primers should be 14–22 bp.
2. At the 3′ ends, primers should contain GC-type base pairs (e.g., GG, CC, GC, or CG) to
facilitate complementary strand formation.
3. The preferred GC content of the primers is 45–55%.
4. Try to design primers so they do not form intra- or interstrand base pairs. Furthermore, the
3′-ends should not be complementary to each other, or they will form primer dimers.
5. One can design an RT primer so that it does not contain secondary structures.
Annealing Temperatures
Annealing temperatures for RT and for DNA amplification can be chosen according
to the following formula:
Tm of the primers = 81.5°C + 16.6 (log M) + 0.41 (G + C%) − 500/n
where n = length of primers and M = molarity of the salt in the buffer, usually 0.047
M for DNA reactions and 0.070 M for RT reactions (see below).
If using AMVRT, the value will be lower, according to the following formula:
Tm of the primers = 62.3°C + 0.41 (G + C%) − 500/n
A simpler version could be used for primers of 18 bp or less:
Tm = 4°C (no. of GC pairs) + 2°C (no. of AT pairs)
Usually, primer annealing is optimal at 2°C above its Tm. However, this formula
provides only an approximate temperature for annealing, since base stacking, near-
neighbor effect, and buffering capacity may play a significant role for a particular primer.
Optimization of the annealing temperature should be carried out first with solution-
based reactions. It is important to know the optimal temperature before attempting to
conduct in situ amplification, as in situ reactions are simply not as robust as solution-
based ones. We hypothesize that this is because primers do not have easy access to
DNA templates inside cells and tissues, as numerous membranes, folds, and other small
structures can keep primers from binding homologous sites as readily as they do in
solution-based reactions.
There are two addition ways to determine the real annealing temperatures:
1. To utilize a recently developed thermocycler designed for determination of actual annealing
temperature called the Robocycler (Stratagene, La Jolla, CA).
2. To utilize so-called touchdown PCR (22).
The logic for determining the correct annealing temperature for in situ PCR is as follows:
Recently, M.J. Research (Watertown, MA) have devised a thermocycler that has the
capacity to perform in situ gene amplification both in slides and in solution (tubes)
simultaneously, in the same block. Also, their DNA-Engine Twin-Tower MS 16×2 can
perform multiple functions, and in situ hybridization could be carried out if one inserts
a wet Kimwipe paper in one of the slide slots. In addition, Perkin-Elmer also has
recently introduced an excellent thermocycler. These kinds of thermocyclers can be
very useful in the in situ amplification of a gene of interest.
218 Bagasra, Bobroski, and Amjad
completely seal the cover slip-slide assembly in order to form a small reaction “chamber”
that can contain the water vapor during thermal cycling. For effective sealing, do not
use colored polish or any other nail polish that is especially “runny”—our laboratory
prefers to use Wet & Wild Clear nail polish. Proper sealing is very important, for this
keeps reaction concentrations consistent through the thermal cycling procedure, and
concentrations are critical to proper amplification. However, be certain to apply the
nail polish very carefully so that none of the polish gets into the actual chamber where
the cells or tissues reside. If any nail polish does enter the chamber, discard that slide—
the results will be questionable. Please bear in mind that the painting of nail polish is
truly a learned skill; therefore, it is strongly recommended that researchers practice
this procedure several times with mock slides before attempting an experiment.
In the case of thick tissue sections, it is best to use another identical blank slide for
the cover instead of a cover slip. Apply the amplification cocktail to the appropriate
well of the blank slide, place an inverted tissue-containing slide atop the blank slide,
and seal the edges as described. Invert the slide once again so that the tissue-containing
slide is on the bottom. This technique can be modified to accommodate a hot start
(see below).
Hot Start Technique
There is much debate as to whether a hot start helps to improve the specificity and
sensitivity of amplification reactions. In our laboratory, we find the hot start adds no
advantage in this regard, rather, it adds only technical difficulty to the practice of the
in situ technique. However, recently a variation of the hot start has been reported
(42,75). In this procedure, one simply uses anti-Taq antibody in the PCR cocktail
(containing Taq), which keeps the Taq enzyme in the cocktail “blocked” until the first
cycle of 92°C when anti-Taq antibody is denatured and restores the full Taq activity.
This modification essentially serves the same function as the hot start procedure but
without its difficulties (15).
Thermal Cyclers
Various thermocycler technologies will work in this application; however, some
instruments work much better than others. In our laboratory, we use two types: a
standard, block-type thermocycler that normally holds 60 0.5-mL tubes but that can
be adapted with aluminum foil, paper towels, and a weight to hold four to six slides.
We also use dedicated thermocyclers that are specifically designed to hold 12 or 16
slides. We understand that other labs have used stirred-air, oven-type thermocyclers
quite successfully; however, we have also heard that there are sometimes problems
with the cracking of glass slides during cycling. Thermocyclers dedicated to glass slides
are now available from several vendors, including Barnstead Thermolyne of Iowa, Coy
Corporation of Minnesota, Hybaid of England, Perkin-Elmer of California, and M.J.
Research of Massachusetts. Our laboratory has used an M.J. Research PTC-100-16MS,
PTC-100-16MS, and DNA-Engine Twin-Tower 16×2 quite successfully. Recently, this
company has combined the slide and tubes into a single block, allowing the simultaneous
confirmation of in situ amplification in a tube. Furthermore, newer designs of thermal
cycler incorporate humidification chambers. However, we do not yet have sufficient
experience with this technology to verify whether they can eliminate the need for
220 Bagasra, Bobroski, and Amjad
sealing the slides with nail polish during thermal cycling. Nonetheless, the humidified
instruments are especially useful in the RT and hybridization steps, for which a humidi-
fied incubator is otherwise needed.
We suggest that you follow the manufacturer’s instructions on the use of your own
thermocycler, bearing in mind the following points:
1. Glass does not easily make good thermal contact with the surface on which it rests. Therefore,
a weight to press down the slides and/or a thin layer of mineral oil to fill in the interstices
will help thermal conduction. If using mineral oil, make certain that the oil is well smeared
over the glass surface so that the slide is not merely floating on air bubbles beneath it.
2. The top surfaces of slides lose heat quite rapidly through radiation and convection; therefore,
use a thermocycler that envelopes the slide in an enclosed chamber (as in some dedicated
instruments), or insulate the tops of the slides in some manner. Insulation is particularly
critical when using a weight on top of the slides, for the weight can serve as an unwanted
heat sink if it is in direct contact with the slides.
3. Good thermal uniformity is imperative for good results—poor uniformity or irregular thermal
change can result in cracked slides, uneven amplification, or completely failed reactions. If
adapting a thermocycler that normally holds plastic tubes, use a layer of aluminum foil to
spread out the heat.
DNA and RT amplification can be carried out simultaneously; of course one still needs
to perform the RT step but this time without pre-DNAse treatment. Subsequently,
products can be labeled with different kinds of probes, resulting in different colors of
signal. For example, proteins can have a rhodamine-labeled probe, mRNA can show
a fluorescein isothiocyanate (FITC) signal (FITC-conjugated probe, >20 different fluo-
rochromes are available), and DNA can be labeled with a biotin-peroxidase probe or
a fluorochrome with different color emissions. Each will show a different signal within
an individual cell (14,22).
HYBRIDIZATION
The in situ hybridization (ISH) technique has been successfully applied in both the
research and clinical settings. However, one single, easy to use universal procedure
has not been developed. Therefore, specific needs of the diagnostic or research goals
must be considered in choosing a suitable protocol.
In comparing ISH with other methods, one has to realize what is being detected.
For example, the immunocytochemical method localizes protein within a cell or on a
cell surface and therefore identifies gene expression. However, these assays cannot yield
useful information on posttranslational processing of the gene product or differentiate
between the uptake and storage of the protein and the site of synthesis of the protein.
In addition, one needs several hundred copies of the proteins to be able to identify an
expression signal. Also, most of the protein is destroyed by formalin fixation methods,
making it difficult to identify the protein in most of the pathologic specimens. In the
mRNA extraction methods that utilize the isolation of nucleic acids from cells (filter
hybridization assays), there can be a dilution of the target found in only a few cells
by many cells with little or no target. These methods provide no distribution information.
They provide only an average measurement of the nucleic acid target present in the
mixed cell population. Therefore ISH is a very powerful technique when the target is
focally distributed within a single cell or a certain histologic cell type within a tissue.
Consequently, ISH has greater sensitivity than filter assays if the gene expression is
taking place in a small subpopulation of cells. The major limitation of ISH is relative
insensitivity, compared with in situ PCR. By utilizing ISH, one can detect as few as
20 copies of mRNA. However, success to that degree of sensitivity is limited to only
few highly specialized laboratories and to a few specific genes. More realistically,
detection of >100 copies/cell would be an achievable goal for a laboratory not specialized
in ISH. To detect a single copy of an integrated gene or to detect a very low level of
gene expression (few copies of mRNA), one can amplify the gene sequences in situ
by DNA or RNA (RT) in situ PCR and then utilize ISH to detect the amplicans.
Analyzing gene expression by ISH after RT can provide information on the site of
mRNA synthesis, which provides information about the cellular origin of protein
synthesis and demonstrates the amounts of synthesis (level of gene expression). This
permits an understanding of the cell types involved in the synthesis of certain proteins
in certain gene regulation and cell types infected by various viral or other infectious
agents. In addition, by combining immunohistochemistry and RT, differential expression
of a gene in different cell types or different stages of development can be analyzed at
the microscopic level.
222 Bagasra, Bobroski, and Amjad
ISH Procedure
Prepare a solution containing: 20–50 pM/µL of the appropriate probe, 50% deionized
formamide, 2× SSC buffer, 10× Denhardt’s solution, 0.1% sonicated salmon sperm
DNA, and 0.1% sodium dodecyl sulfate (SDS). The following is a convenient recipe:
probe (biotinylated, or digoxigenin) 2 µL
deionized formamide 50 µL
20× SSC* 10 µL
50× Denhardt’s solution 20 µL
10 mg/ml ssDNA* 10 µL
10% SDS 1 µL
H2O 7 µL
Total volume 100 µL
*See Materials and Methods, page 224, for preparation of 20× SSC buffer; the
salmon sperm should be denatured at 94°C for 10 min, before it is added to the
hybridization buffer.
Note: 2% bovine serum albumin (BSA) can be added if one is observing nonspecific
binding. For this purpose, one can add 10 µL of 20% BSA solution and reduce the
amount of water.
1. Add 10 µL of hybridization mixture to each well and add cover slips.
2. Heat slides on a block at 95°C for 5 min to denature the double-stranded DNA.
3. Incubate slides at 48°C for 2–4 h in a humidified atmosphere.
Note: The optimal hybridization temperature is a function of the Tm (melting temperature)
of the probe. This must be calculated for each probe, as described earlier. However, the
hybridization temperatures used should not be too high. If that circumstance occurs, then
the formula for the hybridization solution should be modified and instead of 50% formamide,
40% formamide should be substituted (described further in the in situ hybridization section
of ref. 15.)
PROTOCOL: In Situ PCR 223
Proteinase K
Dissolve powder from Sigma in water to obtain 1 mg/mL concentration. Aliquot
and store at −20°C.
Working Solution
Dilute 1 mL of stock (1 mg/mL) into 150 mL of 1× PBS.
20× SSC
Dissolve 175.3 g of NaCl and 88.2 g of sodium citrate in 800 mL of water. Adjust
the pH to 7.0 with a few drops of 10N solution of NaOH. Adjust the volume to 1 L
with water. Sterilize by autoclaving.
PROTOCOL: In Situ PCR 225
2× SSC
Dilute 20× SSC; 100 mL of 20× SSC and 900 mL of water.
Streptavidin Peroxidase
Dissolve powder from Sigma in PBS to make a stock of 1 mg/mL. Just before use,
dilute stock solution in sterile PBS at a 1:30 ratio.
Color Solution
Dissolve one AEC (Sigma) tablet in 2.5 mL of N,N,-dimethyl formamide. Store at
4°C in the dark.
Working solution:
50 mM acetate buffer, pH 5.0 5 mL
AEC solution 250 µL
30% H2O2 25 µL
REFERENCES
1. Bagasra, O. (1990) Polymerase chain reaction in situ. Amplifications March, 20–21.
2. Bagasra, O., Hauptman, S. P., Lischner, H. W., Sachs, M., and Pomerantz, R. J. (1992)
Detection of HIV-1 provirus in mononuclear cells by in situ PCR. N. Engl. J. Med. 326, 1385–
1391.
3. Bagasra, O., Seshamma, T., and Pomerantz, R. J. (1993) Polymerase chain reaction in situ:
intracellular amplification and detection of HIV-1 proviral DNA and other specific genes.
J. Immunol. Methods 158, 131–145.
4. Bagasra, O. and R. J. Pomerantz. (1993) HIV-1 provirus is demonstrated in peripheral
blood monocytes in vivo: a study utilizing an in situ PCR. AIDS Res. Hum. Retroviruses
9, 69–76.
5. Bagasra, O., Seshamma, T., Oakes, J., and Pomerantz, R. J. (1993) Frequency of cells
positive for HIV-1 sequences assessed by in situ polymerase chain reaction. AIDS 7, 82–86.
6. Bagasra, O., Seshamma, T., Oakes, J., and Pomerantz, R. J. (1993) High percentages of
CD4-positive lymphocytes harbor the HIV-1 provirus in the blood of certain infected
individuals. AIDS 7, 1419–1425.
226 Bagasra, Bobroski, and Amjad
7. Bagasra, O., Farzadegan, H., Seshamma, T., Oakes, J., Saah, A., and Pomerantz, R. J. (1994)
Human immunodeficiency virus type 1 infection of sperm in vivo. AIDS 8, 1669–1674.
8. Bagasra, O. and Pomerantz, R. J. (1994) In situ polymerase chain reaction and HIV-1, in
Clinics of North America (Pomerantz, R.J., ed.), W. B. Saunders, publishers, Philadelphia,
pp. 351–366.
9. Bagasra, O., Hui, Z., Bobroski, L., Seshamma, T., Saikumari, P., and Pomerantz, R. J.
(1995) One step amplification of HIV-1 mRNA and DNA at a single cell level by in situ
polymerase chain reaction. Cell Vision 2, 425–1573.
10. Bagasra, O., Michaels, F., Mu, Y., et al. (1995) Activation of the inducible form of nitric
oxide synthetase in the brains of patients with multiple sclerosis. Proc. Natl. Acad. Sci.
USA 92, 12041–12045.
11. Bagasra, O. (1996) Use of in situ PCR for measuring viral burden. AIDS Reader 6, 43–47.
12. Bagasra, O. and Pomerantz, R. J. (1997) Human herpesvirus 8 DNA sequences in CD8 T-
cells [Letter]. J. Infect. Dis. 176, 541.
13. Bagasra, O. and Pomerantz, R. J. (1995) Detection of HIV-1 in the brain tissue of individuals
who died from AIDS, in (Sarkar, G., ed.), PCR in Neuroscience, Academic, San Diego,
pp. 339–357.
14. Bagasra, O., Seshamma, T., Pastanar, J. P., and Pomerantz, R. (1995) Detection of HIV-1
gene sequences in the brain tissues by in situ polymerase chain reaction, in Technical
Advances in AIDS Research in the Nervous System (Majors, E., ed.), Plenum, New York,
pp. 251–266.
15. Bagasra, O., Seshamma, T., Hansen, J., and Pomerantz, R. (1995) In situ polymerase chain
reaction and hybridization to detect low abundance nucleic acid targets, in Current Protocols
in Molecular Biology (Ausubel, et al.), Chapter 14, pp. 1–49.
16. Bagasra, O., Michaels, F., Zheng, Y. M., et al. (1995) Activation of the inducible form of
nitric oxide synthetase in the brains of patients with multiple sclerosis. Proc. Natl. Acad.
Sci. USA 92, 12041–12045.
17. Bagasra, O., Lavi, U., Bobroski, L., Khalili, K., Pestaner, J. P., and Pomerantz, R. J.
(1996) Cellular reservoirs of HIV-1 in the central nervous system of infected individuals:
identification by the combination of in situ PCR and immunohistochemistry. AIDS 10, 573–
585.
18. Bagasra, O. and Amjad, M. (2000) Protection against retroviruses are owing to a different
form of immunity: an RNA-based molecular immunity hypothesis. Appl. Immunochem.
Mol. Morphol. 8, 133–146.
19. Bobroski, L., Alexander, U., Bagasra, O., et al. (1999) Mechanism of vertical transmission
of HIV-1: role of intervillous space, Appl. Immunochem. Mol. Morphol. 7, 271–279.
20. Bobroski, L., Alexander, U., Bagasra, O., et al. (1999) Localization of human herpes virus
type 8 (HHV-8) in the Kaposi’s sarcoma tissues and the semen specimens of HIV-1 infected
and uninfected individuals by utilizing in situ polymerase chain reaction. J. Reprod. Immunol.
41, 149–160.
21. Embretson, J., Zupanic, M., Beneke, T., et al. (1993) Analysis of human immunodeficiency
virus-infected tissues by amplification and in situ hybridization reveals latent and permissive
infections at single-cell resolution. Proc. Natl. Acad. Sci. USA 90, 357–361.
22. Embretson, J., Zupancic, M., Ribas, J. L., et al. (1993) Massive covert infection of helper
T lymphocytes and macrophages by HIV during the incubation period of AIDS. Nature
62, 359–362.
23. Hooper, D. C., Bagasra, O., Marini, J. C., et al. (1997) Prevention of experimental allergic
encephalomyelitis by targeting nitric oxide and peroxynitrite: implications for the treatment
of multiple sclerosis. Proc. Natl. Acad. Sci. USA 94, 2528–2533.
24. Patterson, B. K., Till, M., Otto, P., et al. (1993) Detection of HIV-I DNA and messenger
PROTOCOL: In Situ PCR 227
RNA in individual cells by PCR-driven in situ hybridization and flow cytometry. Science
260, 976–979.
25. Hsu, T-C., Scott, K., Seshamma, T., Bagasra, O., and Walsh, P. N. (1998) Molecular cloning
of platelet factor XI, an alternative splicing product of the plasma factor XI. J. Biol. Chem.
273, 13,787–13,793.
26. Pereira, R. F., Halford, K. W., O’Hara, M. D., et al. (1995) Cultured stromal cells from
marrow serve as stem cells for bone, lung and cartilage in irradiated mice. Proc. Natl.
Acad. Sci. USA 92, 4857–4861.
27. Qureshi, M. N., Barr, C. E., Seshamma, T., Pomerantz, R. J., and Bagasra, O. (1994)
Localization of HIV-1 proviral DNA in oral mucosal epithelial cells. J. Infect. Dis. 171, 190–
193.
28. Pellett, P. E., Spira, T., Bagasra, O., et al. (1999) Multi-center comparison of polymerase
chain reaction detection of human herpesvirus 8 DNA in semen. J. Clin. Microbiol. 37, 1293–
1301.
29. Mehta, A., Maggioncalda, J., Bagasra, O., et al. (1995) In situ PCR and RNA hybridization
detection of herpes simplex virus sequences in trigeminal ganglia of latently infected mice.
Virology 206, 633–640.
30. Lattime, E. C., Mastrangelo, M. J., O. Bagasra, and Berd D. (1995) Expression of cytokine
mRNA in human melanoma tissue. Cancer Immunol. & Immunother. 41, 151–156.
31. Pestaner, J. P., Bibbo, M., Bobroski, L., Seshamma, T., and Bagasra, O. (1994). Potential
of in situ polymerase chain reaction in diagnostic cytology. Acta Cytol. 38, 676–680.
32. Sullivan, D. E., Bobroski, L. E., Bagasra, O., Finney, M. (1997) Self-seal reagent: evapora-
tion control for molecular histology procedure without chambers, clips or fingernail polish.
Biotechniques 23, 320–225.
33. Bagasra, O. and Hansen, J. (1997) In Situ PCR Techniques. John Wiley & Sons, New York.
34. Nuovo, G. J. (1994) PCR In Situ Hybridization Protocols and Applications, 2nd ed., Raven,
New York.
13
Immunohistochemistry: Theory and Practice
INTRODUCTION
In the daily practice of surgical pathology, immunohistochemistry (IHC) is essential
for the diagnosis and classification of neoplasms. The technique is no longer limited
to large university-based or reference laboratories and has become widespread in com-
munity hospital laboratories. The propagation of commercially available, automated
immunostainers has played a major role in increasing the capability to perform IHC
and has enabled smaller laboratories to produce quality immunostains without prior
technical expertise and without significant increases in personnel. The availability of
a large selection of primary antibodies that can be used on formalin-fixed, paraffin-
embedded tissue sections has also contributed to increased use and sensitivity of the
method.
The ruling of the Food and Drug Administration (FDA) on the classification of
immunohistochemical reagents and kits (1) is a recent and very significant event familiar
to most who deal with IHC on a daily basis. Implementation of the ruling resulted in
the classification of the majority of immunohistochemical reagents as Analyte Specific
Reagents” (ASRs) (2,3) and class I medical devices, thus exempting them from premar-
ket notification. The rationale for this decision is based on the fact that IHC staining
results are incorporated into a surgical pathology report as one part of the entire
diagnostic evaluation. The immunostains are not usually “stand-alone” results. Estrogen
receptor and progesterone receptor, however, are considered class II devices. These
immunostains do not have routine morphologic correlates but do have substantial and
widely accepted scientific validation. Class III devices include those immunostains that
are not part of the surgical pathology diagnostic process and may result in an independent
report to a physician. These stains require premarket notification and specific FDA
approval. The HercepTest for determination of Her2 protein overexpression as an
indication for Herceptin (trastuzumab) therapy is an example.
The FDA ruling allows IHC laboratories to continue operating essentially as they
have in the past. Reagent manufacturers must now label the antibodies they sell for
diagnostic use (and which do not have FDA clearance) as ASRs. The onus is on the
manufacturers to follow “good manufacturing practices” and to make certain that
antibody reagents have consistent high quality and the specificity that is claimed.
229
230 Roche and Hsi
Surgical pathology reports incorporating IHC results, however, are now obligated to
include a statement indicating that responsibility for assuring quality of the immunostains
rests with the individual laboratory and not the manufacturer of the reagent (4). The
laboratory director therefore has the ultimate responsibility to generate high-quality
immunostains and to document their performance.
In this chapter, current and practical procedures essential for performing quality
immunostains in the modern automated IHC laboratory are discussed.
Tissue fixation is perhaps the single most important variable influencing the outcome
of paraffin section immunostains, and its influence can not be overstated. The routine
fixative of choice in most pathology laboratories is neutral-buffered formalin. It is
232 Roche and Hsi
to as antigen retrieval. The impact of this innovative and revolutionary technique has
been profound, and the reader is referred to Chapter 14 for an in-depth discussion of
antigen retrieval methods in immunohistochemistry.
DETECTION SYSTEMS
Numerous methods and reagents are commercially available for detection of bound
primary antibodies by IHC, but enzyme-based techniques using either horseradish
peroxidase or alkaline phosphatase predominate. For most automated stainers in clinical
laboratories, peroxidase methods are favored. Examples of peroxidase-based immuno-
histochemistry include the peroxidase-antiperoxidase (PAP) method, the enzyme-
labeled streptavidin-biotin (LSAB) technique, and the avidin-biotin complex (ABC)
method. These are all three-layer procedures involving a primary antibody, a secondary
antibody with appropriate specificity for the primary, and the enzyme-containing tertiary
reagent. In the case of the PAP method, the secondary antibody is unlabeled and added
in excess so that it binds both the primary antibody and the soluble enzyme-antibody
(PAP) complex that comprises the third layer. Consequently, with the PAP technique
the antibody species in PAP complex must match the primary antibody. For the LSAB
and ABC techniques, the secondary antibody is covalently conjugated with biotin
(biotinylated) and need only recognize and bind the primary antibody. The third layer
of enzyme-labeled streptavidin or avidin-biotin complex then binds tightly to the biotin.
The long-established “indirect method” utilizes a secondary antibody that is directly
conjugated with peroxidase or alkaline phosphatase to bind to the primary antibody.
This two-step method is simple and quick but also considerably less sensitive than the
three-step detection systems described above. More recently, secondary antibodies
conjugated with a dextran polymer backbone containing a large number of enzyme
molecules (about 100) have been introduced (EnVision Systems, Dako, Carpinteria,
CA). With this new generation of secondary antibody conjugates, it is possible to
generate two-layer immunostaining protocols that have sensitivities approaching the
three-layer methods (16). An added advantage is that the system is biotin free and
therefore does not display the nonspecific staining associated with biotin-binding pro-
teins or endogenous biotin.
There are also catalyzed signal amplification or catalyzed reporter deposition detec-
tion methods that generate extreme sensitivity and to which most clinical immunohi-
stochemists are not accustomed (17). Typically these are peroxidase- and tyramide-
based amplification techniques; they are discussed in detail in Chapter 15.
Selection of a chromogen is also an important factor to consider in immunohisto-
chemistry. Peroxidase systems routinely use 3,3′-diaminobenzidine (DAB) (brown) or
3-amino-9-ethylcarbazole (AEC) (red) as chromogens. For alkaline phosphatase sys-
tems, 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium (BCIP/NBT) may
be used to yield a blue-purple product or fast red/naphthol AS-TR phosphate to produce
a red/pink product. Chromogen choice can be influenced by personal preference (red
vs blue vs brown) and degree of contrast with the counterstain (hematoxylin vs methyl
green). Other considerations include incubation time and sensitivity (peroxidase/DAB
reacts faster, with a shorter incubation time, than phosphatase/FAST red), safety issues
(DAB is a carcinogen), or technical issues (AEC is soluble in organic solvents and
234 Roche and Hsi
PRIMARY ANTIBODIES
Selecting a primary antibody for IHC use can be a difficult task. Frequently, different
antibodies from several different sources are available against a particular antigen;
often all of them claim to work for IHC. Frequently, though, there are assessments in
the literature that can provide guidance in this decision. For example, different antibodies
directed against p53 have been compared for their reactivity in formalin-fixed tissues
and have demonstrated differences in reactivity that impact the prognostic significance
of p53 antigen expression in breast cancer (18). The polyclonal antibody pAb1801 and
the monoclonal antibody DO7 were reported to be more effective than others tested.
A comprehensive review of differences in antibody reactivity to particular antigens
is beyond the scope of this discussion. However, it is critically important that the
immunohistochemist be familiar with the recent literature specific to an individual
antibody clone to avoid, or at least be informed of, potential problems.
Most antibodies used in a clinical immunohistochemistry laboratory are commercially
available, and their specificity is well characterized. As ASRs, which most IHC reagents
are, good manufacturing practices are assumed to have been followed, and most antibod-
ies will perform “as advertised.” Usually, peer-reviewed literature also exists and can
be used to gather a realistic review of antibody performance. Such information will
usually also provide a good reference point from which to start primary antibody
dilutions and antigen retrieval methods. Many manufacturers now offer their ASR
antibodies as prediluted reagents “optimized” for paraffin section IHC. Predilutes are
becoming a common practice but can limit flexibility when developing a new stain in
a particular laboratory. The main concern is that the antibody dilution will be such
that it only works on optimally fixed tissues with the manufacturer’s own detection
system. For cases in which suboptimally processed tissue is evaluated or a less sensitive
detection system is used, the preset dilution can be too great. Whenever possible,
undiluted primary antibody is preferable.
The immunohistochemist should also pay attention to whether the primary antibody is
polyclonal or monoclonal and note the species and immunoglobulin isotype. Polyclonal
antibodies generally have high-affinity clones within them and can be significantly
diluted (often >1:1000). On the other hand, polyclonals are also more likely to generate
background staining and can require extra or prolonged blocking steps. Knowing the
species and isotype of monoclonals will avoid problems in which an inappropriate
secondary antibody is used. An example in which isotype makes a difference is the
Immunohistochemistry: Theory and Practice 235
LeuM1 clone (anti-CD15), an IgM mouse monoclonal antibody that has been shown
to have superior performance when an anti-IgM secondary antibody is used.
OPTIMIZING CONDITIONS
(26,27). Blocking kits are commercially available from several vendors, including
Dako and Vector Laboratories (Burlingame, CA). Endogenous peroxidase activity in
horseradish peroxidase-based systems and endogenous phosphatase in alkaline phospha-
tase-based systems are other potential causes of false-positive staining. Both activities
can be eliminated by blocking steps that incorporate incubation with hydrogen peroxide
or levamisole, respectively (28,29).
CONCLUSIONS
This chapter has presented practical methods and considerations that are essential
for the production of quality immunohistochemical stains. Protocols for tissue freezing
and formalin fixation have been discussed since consistent and proper execution of
these basic procedures is critical for maintaining tissue morphology and antigenicity.
The increasing acceptance and utilization of automated immunostainers is promoting
standardization of IHC, both within an individual laboratory and among laboratories
using similar reagents and equipment. National proficiency testing programs, such as
those conducted by the College of American Pathologists, also encourage and support
standardization of staining procedures (including issues such as specimen processing,
antigen retrieval, detection systems, and antibody selection) and interpretation. Immuno-
histochemistry laboratories can only benefit from participation in these programs and
surveys.
REFERENCES
1. Food and Drug Administration. (1998) Medical devices; classification/reclassification of
immunohistochemistry reagents and kits—FDA. Final rule. Fed. Reg. 63, 30132–30142.
2. Taylor, C. R. (1999) FDA issues final rule for classification and reclassification of immuno-
histochemistry reagents and kits [Editorial]. Am. J. Clin. Pathol. 111, 443–444.
3. Taylor, C.R. (1998) Report from the Biological Stain Commission: FDA issues final rule
for classification/reclassification of immunohistochemistry (IHC) reagents and kits. Biotech.
Histochem. 73, 175–177.
4. Swanson, P. E. (1999) Labels, disclaimers, and rules (oh, my!). Analyte-specific reagents
and practice of immunohistochemistry [editorial]. Am. J. Clin. Pathol. 111, 445–448.
5. Rentrop, M., Knapp, B., Winter, H., and Schweizer. (1986) Aminoalkylsilane-treated glass
slides as support for in situ hybridization of keratin cDNAs to frozen tissue sections under
varying fixation and pretreatment conditions. Histochem. J. 18, 271–276.
6. Jacobson, B. S. and Branton, D. (1977) Plasma membrane: rapid isolation and exposure
of the cytoplasmic surface by use of positively charged beads. Science 195, 302–304.
7. Werner, M., Von Wasielewski, R., and Komminoth, P. (1996) Antigen retrieval, signal
amplification and intensification in immunohistochemistry. Histochem. Cell Biol. 105, 253–
260.
8. Morgan, J. M., Navabi, H., Schmnid, K. W., and Jasani, B. (1994) Possible role of tissue-
bound calcium ions in citrate-mediated high-temperature antigen retrieval. J. Pathol.
174, 301–307.
9. Bostwick, D. G., al Annouf, N., and Choi, C. (1994) Establishment of the formalin-free
surgical pathology laboratory. Utility of an alcohol-based fixative. Arch. Pathol. Lab. Med.
118, 298–302.
10. Battifora, H. (1991) Assessment of antigen damage in immunohistochemistry. The vimentin
internal control. Am. J. Clin. Pathol. 96, 669–671.
Immunohistochemistry: Theory and Practice 237
11. Battifora, H., and Kopinski, M. (1986) The influence of protease digestion and duration of
fixation on the immunostaining of keratins. A comparison of formalin and ethanol fixation.
J. Histochem. Cytochem. 34, 1095–100.
12. Fox, C. H., Johnson, F. B., Whiting, J., and Roller, P. P. (1985) Formaldehyde fixation. J.
Histochem. Cytochem. 33, 845–853.
13. Roche, P. C. (1998) Antigen stability in stored paraffin sections. CAP Today 12, 59–61.
14. Prioleau, J. E., and Schmitt, S. J. (1995) p53 antigen loss in stored paraffin slides. N. Engl.
J. Med. 332, 1521–1522.
15. Jacobs, T. W., Prioleau, J. E., Stillman, I. E., and Schmitt, S. J. (1996) Loss of tumor
marker-immunostaining intensity on stored paraffin slides of breast cancer. J. Natl. Cancer
Inst. 88, 1054–1059.
16. Sabattini, E., Bisgaard, K., Ascani, S., et al. (1998) The EnVision++ system: a new immuno-
histochemical method for diagnostics and research. Critical comparison with the APAAP,
ChemMate, CSA, LABC, and SABC techniques. J. Clin. Pathol. 51, 506–511.
17. King, G., Payne, S., Walker, F., and Murray, G. I. (1997) A highly sensitive detection
method for immunohistochemistry using biotinylated tyramine. J. Pathol. 183, 237–241.
18. Horne, G. M., Anderson, J. J., Tiniakos, D. G., et al. (1996) p53 protein as a prognostic
indicator in breast carcinoma: a comparison of four antibodies for immunohistochemistry.
Br. J. Cancer 73, 29–35.
19. Shi, S. R., Cote, R. J., Yang, C., et al. (1997) Development of an optimal protocol for
antigen retrieval: a ‘test battery’ approach exemplified with reference to the staining of
retinoblastoma protein (pRB) in formalin-fixed paraffin sections. Histopathology 31, 400–
407.
20. Shi, S. R., Cote, R. J., and Taylor, C. R. (1996) Antigen retrieval immunohistochemistry:
past, present, and future. Biotech. Histochem. 71, 190–196.
21. Shi, S. R., Key, M. E., and Kalra, K. L. (1991) Antigen retrieval in formalin-fixed, paraffin-
embedded tissues: an enhancement method for immunohistochemical staining based on
microwave oven heating of tissue sections. J. Histochem. Cytochem. 39, 741–748.
22. Cuevas, E. C., Bateman, A. C., Wilkins, B. S., et al. (1994) Microwave antigen retrieval
in immunohistochemistry: a study of 80 antibodies. J. Clin. Pathol. 47, 448–452.
23. Pileri, S. A., Roncador, G., Ceccarelli, C., et al. (1997) Antigen retrieval techniques in
immunohistochemistry: comparison of different methods. J. Pathol. 183, 116–123.
24. Cattoretti, G., Pileri, S., Parravicini, C., et al. (1993) Antigen unmasking on formalin-fixed,
paraffin-embedded tissue sections. J. Pathol. 171, 83–98.
25. von Wasielewski, R., Werner, M., Nolte, M., Wilkens, L., and Georgii, A. (1994) Effects
of antigen retrieval by microwave heating in formalin-fixed tissue sections on a broad panel
of antibodies. Histochemistry 102, 165–172.
26. Bussolati, G., Gugliotta, P., Volante, M., Pace, M., and Papotti, M. (1997) Retrieved
endogenous biotin: a novel marker and a potential pitfall in diagnostic immunohistochemis-
try. Histopathology 31, 400–407.
27. Wood, G. S. and Warnke, R. (1981) Suppression of endogenous avidin-binding activity
in tissues and its relevance to biotin-avidin detection systems. J. Histochem. Cytochem.
29, 1196–1204.
28. Li, C. Y., Ziesmer, S. C., and Lazcano-Villareal, O. (1987) Use of azide and hydrogen
peroxide as an inhibitor for endogenous peroxidase in the immunoperoxidase method.
J. Histochem. Cytochem. 35, 1457–1460.
29. Ponder, B. A., and Wilkinson, M. M. (1981) Inhibition of endogenous tissue alkaline
phosphatase with the use of alkaline phosphatase conjugates in immunohistochemistry.
J. Histochem. Cytochem. 29, 981–984.
14
Antigen Retrieval Technique for
Immunohistochemistry
Principles, Protocols, and Further Development
INTRODUCTION
Immunohistochemistry (IHC) has a long history stemming from the work of Coons
and colleagues in 1941 (1), who developed an immunofluorescence technique to detect
bacteria. In the early 1970s, extensive efforts were made to render immunoperoxidase
methods more widely available to routine formalin-fixed, paraffin-embedded tissues
through a series of technical developments, including increasingly sensitive detection
systems, and various pretreatment methods prior to the immunostaining procedure
(2–9). The basic philosophy of applying IHC for routine formalin-fixed tissues was
set forth by Taylor and Kledzik (2) in 1981.
However, IHC methods that included unmasking procedures such as enzymatic
digestion, although widely applied, did not improve IHC staining of the majority of
antigens, as pointed out by Leong et al. (10). A major drawback of enzymatic digestion
was that it proved difficult to control the optimal digestion conditions for individual
tissue sections when they were stained with different antibodies. Difficulties in standard-
izing this method provided a powerful incentive to the development of a new technique,
with the requirements that it should be more effective, more widely applicable, and
easier to use than enzymatic digestion. In addition, to be acceptable, any new method
should enhance immunohistochemical staining of routine paraffin sections in a reproduc-
ible and reliable manner.
In this context, the early experience of one of the authors (3–5), led to the realization
that the application of IHC to “routine formalin paraffin sections” would represent a
new era for the demonstration of antigens in the types of “routinely” processed tissues
that abound in university pathology laboratories. Indeed, the method opened a new
vista of research and diagnostic possibilities, yielding a level of morphologic detail
that allowed pathologists to correlate immunologic findings directly with traditional
cytologic and histologic criteria. This advance, in turn, provided Shi with the impetus for
development of an antigen retrieval (AR) technique. The first seeds in the development of
a retrieval approach for routinely processed celloidin-embedded human temporal bone
239
240 Taylor and Shi
sections sprouted in the early 1980s when Shi studied under Dr. Schuknecht’s guidance
at the Eastern National Temporal Bone Bank in Boston. Attempts to perform IHC study
on archival formalin-fixed tissue sections were actively supported by Dr. Shuknecht.
Subsequently the work involved BioGenex for a period and then extended into a broader
range of applications, culminating in our studies at the University of Southern California
in the early 1990s, which provide much of the basis of this chapter.
A key point that we considered in developing the AR technique was whether the
crosslinkages of protein caused by formalin fixation are reversible or irreversible. The
answer was, in essence, already available in the serial studies of Fraenkel-Conrat and
co-workers published in 1940s (11–13); we had only to rediscover it! Their work
indicated that hydrolysis of crosslinkages between formalin and protein may be limited
by certain amino acid side chains, such as imidazol and indol. However, these crosslink-
ages could be disrupted by high-temperature heating (120°C) or strong alkaline treat-
ment. These observations formed the basis for the development of AR in both heating
and nonheating methods. A dramatic enhancement of immunostaining on archival
formalin-fixed, paraffin-embedded tissue sections could be achieved either by boiling
the tissue sections in water solution, or by employing sodium hydroxide-methanol
solutions for immersing celloidin-embedded human temporal bone sections prior to
immunostasining (14,15).
The high-temperature heating AR technique has been widely adopted as a simple,
effective, and reliable pretreatment for routine IHC in surgical pathology as well as
analytical morphology (16–27). Various modifications in terms of heating method and
AR solution have been proposed since 1991. In addition to microwave and conventional
heating methods (14), Shin et al. (28) described a hydrated autoclaving method to
enhance the immunoreactivity of tau protein in brain tissue fixed in formalin. This
work was followed by other studies (29–32) with good results. Subsequently, domestic
pressure cookers and steamers were applied for the high-temperature heating procedure
(33–37). Metal salt solutions and distilled water were used as the earliest AR solutions
(14). The increase of immunostaining intensity resulting from the use of metal salt AR
solutions has subsequently been demonstrated by several studies (16,22,24,38–49). A
single article (50) that reported a poor result using a metal salt solution for AR-IHC
appears to have been based on inappropriate comparisons, as analyzed in detail elsewhere
(51). However, a major drawback of metal salts, particularly lead salts, is the potentially
toxic effect, as emphasized by Suurmeijer and Boon (16). It is, therefore, preferable
to avoid using lead or other toxic metal salts, if other kinds of solution may serve with
equal effect in AR-IHC staining.
The use of citrate buffer solution for the AR heating method was documented by
Cattoretti et al. (52), followed by Gown et al. (18), Suurmeijer and Boon (16), Leong
and Milios (19), Taylor et al. (22), Swanson (53), Cuevas et al. (23), and others
(51,54–64). Cattoretti et al. (17) extended their investigation to 256 antibodies using
a citrate buffer of pH 6.0 as the AR solution, with a microwave heating protocol as
previously documented (14). Meanwhile, other chemicals, such as urea, glycine-HCl
buffer, Tris-HCl, phosphate buffer (PBS), periodic acid, aluminum chloride, and other
metal salts, as well as detergent solutions, have subsequently been essayed as AR
solutions (17,22,42,44,65–71). In addition, commercial AR solutions such as target
unmasking fluid (TUF) have also been proposed (72). The chemical nature of the
Antigen Retrieval Technique for Immunohistochemistry 241
solution has been credited by some as a major factor that influences the effect of AR-
IHC. This idea is a particularly interesting issue for commercial products, for which
key ingredients are kept secret. Confronted with this form of limited “anarchy” (73),
a critical question must be addressed in any effort to achieve greater uniformity of the
AR technique, namely, what are the major factors that influence the effectiveness of
AR-IHC? The following discussion of major factors that influence the effectiveness of
AR-IHC is presented to promote a better understanding of effects, in order to standardize
and further develop the AR technique using a more scientific approach.
Since the AR technique was developed in 1991, the heating AR technique has been
widely applied as a simple and extremely effective pretreatment for IHC on routinely
formalin-fixed, paraffin-embedded tissues (14). During this period, hundreds of original
articles and more than a dozen comprehensive reviews have been published worldwide.
An international symposium on AR-IHC was held on July 25th, 1998 at the Fifth
Joint Meeting of The Japan Society of Histochemistry and Cytochemistry and The
Histochemical Society in San Diego, California. At this symposium, speakers and
audiences from Japan, Europe, and the United States reviewed the major issues concern-
ing the development, clinical application, and factors that influence the effectiveness
of AR-IHC.
Studies of the major factors that influence the effect of AR-IHC provide a basis for
establishing a “test battery” approach, by which various factors may be rapidly evaluated
in order to identify the optimal protocols for “maximal retrieval” level. This simple
approach may allow some standardization of AR-IHC on routinely processed tissues,
as discussed below.
HEATING AR METHOD
Currently, the high-temperature heating AR method is the predominant technique
for enhancement of IHC staining, particularly for routinely formalin-fixed, paraffin-
embedded tissue sections. Extensive use of the method has, however, raised a number
of interesting issues. The major factors that influence the effects of AR-IHC are summa-
rized below.
pH of the AR Solution
The advocacy of various investigators for a variety of buffer solutions as AR solutions
has nevertheless provided food for thought. One hypothesis proposed by our group in
1993 stated that the effectiveness of AR-IHC may be influenced by the pH value of
the AR solution; in essence, the pH value may be more critical than the chemical
composition of the AR solution. We tested the hypothesis that the pH of the AR solution
may influence the quality of immunostaining by using seven different AR buffer
solutions at a series of different pH values ranging from 1 to 10. We evaluated the
staining of monoclonal antibodies to cytoplasmic antigens (AE1, HMB45, neuron-
specific enolase [NSE]), nuclear antigens (MIB1, proliferating cell nuclear antigen
[PCNA], estrogen receptor [ER]), and cell surface antigens (MT1, L26, epithelial
membrane antigen [EMA] on routinely formalin-fixed, paraffin-embedded sections
under different pH conditions with microwave AR heating for 10 min. The pH value
of the AR buffer solution was carefully measured before, immediately after, and 15
min after the AR procedure. From this study, we drew the following conclusions:
1. There were three types of patterns reflecting the influence of pH. First, several antigens
showed no significant variation utilizing AR solutions with pH values ranging from 1.0 to
10.0 (L26, PCNA, AE1, EMA, and NSE); second, other antigens (MIB1, ER) showed a
dramatic decrease in the intensity of the AR-IHC at middle-range pH values (pH 3.0–6.0)
but strong AR-IHC results above and below these critical zones; and third, still other antigens
(MT1, HMB45) showed negative or very weak focally positive immunostaining with a low
pH (1.0–2.0) but excellent results in the high-pH range.
2. Among the seven buffer solutions at any given pH value, the intensity of AR-IHC staining
was very similar, with the single exception that Tris-HCl buffer tended to produce better
results at higher pH, compared with other buffers.
3. Optimization of the AR system should include optimization of the pH of the AR solution.
4. A high-pH AR solution, such as Tris-HCl or sodium acetate buffer at pH 8.0–9.0, may be
suitable for most antigens.
5. Low-pH AR solutions are most useful for nuclear antigens such as retinoblastoma protein
(RB), ER, and androgen receptor.
6. Focal weak false-positive nuclear staining may be found when using low pH AR solution;
the use of negative control slides is important to exclude this possibility (79).
Evers and Uylings (80) also found that AR-IHC is pH and temperature dependent.
They tested two antibodies, microtubule-associated protein-2 (MAP-2) and SMI-32,
and indicated that the optimal pH values were pH 4.5 for MAP-2 and pH 2.5 for SMI-
32. In addition, they demonstrated that the use of 4% AlCl3 as the AR solution for
SMI-32 could achieve a similar result to that obtained by using citrate buffer of pH
Antigen Retrieval Technique for Immunohistochemistry 243
2.5, from which they concluded that it is not important what kind of solution is used
as long as the pH is at an appropriate level. They subsequently demonstrated that
microwave AR in Tris-HCl buffer of pH 9.0 at full-power heating yielded the best
results for most antibodies tested in neuroscience research (81), a finding that supported
our previous conclusion (58,60,65). Unfortunately, the issue is complicated by the
observation that not all antigen/antibody pairings in IHC are affected optimally for the
same pH value; indeed, as noted above, we have reported three separate groupings,
with nuclear antigens, for example, showing greatest recovery at low pH (58,79,82).
Fig. 1. One of the factors that influence the effect of non-heating AR-IHC used for celloidin-
embedded human tissues: correlation among immunostaining reactivity, the concentration of
NaOH-methanol solution, and the immersion time of celloidin sections. The stronger the concen-
tration, the shorter the immersion time. (Reproduced with permission from ref. 62.)
NONHEATING AR METHOD
Penido, Tseng, and Kao (unpublished data) conducted a study of standardization
and development of AR-IHC using routinely celloidin-embedded tissues of human
temporal bone, kidney, liver, spleen, adenoids, lung, skin, and intestine and baboon
temporal bone. A total of 603 celloidin sections were used. Factors that influenced the
results of the nonheating AR method may be summarized as follows:
1. Concentration of AR solution (NaOH-methanol) and immersion time. A series of concentra-
tions of NaOH-methanol solutions ranging from 1:3 to 1:160 diluted by 100% methanol was
carefully analyzed. The results are summarized in Fig. 1. In general, the immunostaining
reactivity was correlated with both the concentration and the immersion time in NaOH-
methanol solution. Since the stronger concentration of NaOH-methanol may damage the fine
structure of inner ear tissue, more diluted AR solution with elongated immersion time may
be preferable. In addition, the stronger NaOH-methanol solution may result in detachment
of the sections from the slides.
2. Methanol is the preferred solvent. One comparative study of AR-IHC on celloidin tissue
sections employed three solutions; NaOH-methanol, NaOH-ethanol, and NaOH-water at the
same ratio of 1:5. After immersing the slides in each AR solution for 30 min, the slides
were rinsed for 20 min in two changes of methanol, ethanol, or distilled water, respectively.
The immunostaining reactivity obtained by using the NaOH-methanol solution showed the
strongest intensity, and the reactivity obtained by ethanol showed a moderate intensity.
Immunostaining reactivity was not detected for the celloidin tissue sections treated by the
Antigen Retrieval Technique for Immunohistochemistry 245
NaOH-water solution. However, a similar strong intensity was achieved by rinsing the slides
in methanol with the three different AR solutions.
Methanol plays an important role in removing celloidin from the celloidin-embedded
tissue, an important factor in improving access of antibody to antigen.
3. Optimal incubation time of immunostaining on celloidin tissue section after AR. A comparison
of AR-IHC for celloidin tissue sections using different incubation times is summarized in
Fig. 2. The study demonstrated that an elongated time of incubation is necessary to achieve
satisfactory results, a finding similar to that incorporated in our protocol documented pre-
viously (15,74,84,85), i.e., overnight incubation for primary antibody and 60 min for link
and label reagents, respectively. One reason why a longer incubation time is necessary may
relate to the thickness of the celloidin section (20 µm).
the temperature, the better the staining for most nuclear and cytoplasmic antibodies
that are tested. In contrast, when using a low-pH solution, overly intense heating (such
as in a pressure cooker) may yield a poor result. For example, the strongest intensity
of staining for MIB1 or pRB on archival paraffin sections was achieved either with
regular microwave heating and low-pH solution or high-pH solution with intense heating
(pressure cooker or elongated regular microwave heating method). Low pH with intense
heating gave poor results (86).
The test battery thus serves as a rapid screening approach to identify an optimal
protocol for each antibody/antigen to be tested. The goal is to establish the “maximal
retrieval” level for formalin-masked antigens with a variety of fixation times, in order
to standardize immunostaining results. In one study of 14 antibodies using the test
battery method, we demonstrated that the strongest intensity of AR-IHC for most
antibodies tested was achieved by using either a low-pH buffer solution with regular
microwave heating conditions or a high pH buffer with intense heating conditions (such
as autoclave heating or microwave heating for elongated heating time) (86). In addition,
the use of a test battery may identify some previously “false-negative” AR-IHC staining,
as demonstrated by a special AR protocol for thrombospondin-IHC staining (60,87).
at 90°C yielded a weaker signal, around 10% compared with the frozen section. In
this study the protocol of microwave heating pretreatment involved immersing sections
in 10 mM citrate buffer, pH 6.0, and microwaving at full power in an 800-W microwave
oven for three 5-min periods, with washes in diethylpyrocarbonate (DEPC)-treated
water.
Extraction of Protein
Ikeda et al. (107) developed a protocol for extraction of diagnostically useful proteins
from formalin-fixed, paraffin-embedded tissue sections. By heating the deparaffinized
tissue sections at 100°C for 20 min before incubation at 60°C for 2 h, in RIPA solution,
121.5 µg of protein may be extracted from 5 mm2 × 50 µm-thick tissue. They have
successfully detected a variety of membrane-bound, cytosolic and nuclear proteins in
archival paraffin-embedded tissues to obtain valuable information for molecular biology.
One potential cause of increased background after AR heating may be the unmasking
of endogenous biotin, which does account for some positive staining in frozen tissue
sections, particularly for liver, kidney, adrenal cortex, and thyroid tissues (115,116).
In this instance, the cytoplasmic biotin reaction appears as a fine granular staining
pattern in the cytoplasm. Kashima et al. (116) found a high incidence (93/208) of this
form of positive cytoplasmic biotin staining in thyroid lesions, especially in papillary
carcinoma. This effect can be avoided by using a routine blocking procedure or using
a detection system free of avidin-biotin complex. It is important to keep in mind that
any background staining that may be observed in frozen tissue sections may also be
found in microwave heated archival paraffin sections; the use of appropriate controls
is helpful in identifying this problem.
Care must be taken when testing new antibodies using different AR protocols, as
emphasized recently by Swanson (73). Indeed variable patterns of immunostaining
have been reported when using different AR protocols (117). The incorporation of
frozen sections into a test battery, employed to select an optimal AR protocol, is
extremely helpful in identifying any false positivity or unexpected change of staining
pattern, as exemplified by the monoclonal antibody to thrombospondin (TSP) immuno-
localization in prostate and bladder mentioned above.
may yield clean background and sharp contrast for both paraffin- or resin-embedded
tissue sections.
ACKNOWLEDGMENTS
Parts of this chapter were adapted with permission from refs. 51 and 62.
REFERENCES
1. Coons, A. H., Creech, H. J., and Jones, R. N. (1941) Immunological properties of an
antibody containing a fluorescent group. Proc. Soc. Exp. Biol. Med. 47, 200–202.
2. Taylor, C. R. and Kledzik, G. (1981) Immunohistologic techniques in surgical pathology.
A spectrum of new special strains. Hum. Pathol. 12, 590–596.
3. Taylor, C. R. and Burns. J. (1974) The demonstration of plasma cells and other immuno-
globulin containing cells in formalin-fixed, paraffin-embedded tissues using peroxidase
labelled antibody. J. Clin. Pathol. 27, 14–20.
4. Taylor, C. R. (1979) Immunohistologic studies of lymphomas: new methodology yields
new information and poses new problems. J. Histochem. Cytochem. 27, 1189–1191.
5. Taylor, C. R. (1980) Immunohistologic studies of lymphoma: past, present and future.
J. Histochem. Cytochem. 28, 777–787.
6. Pinkus, G. S. (1982) Diagnostic immunocytochemistry of paraffin-embedded tissues. Hum.
Pathol. 13, 411–415.
7. Elias, J. M. (1990) Immunohistopathology: A Practical Approach to Diagnosis. 1st ed.
ASCP Press, Chicago, pp. 1–9.
8. Bhan, A. K. (1995) Immunoperoxidase, 2nd ed. Raven, New York, pp. 711–723.
9. DeLellis, R. A. (1988) Advances in Immunohistochemistry. Raven, New York, pp. 1–45.
10. Leong, A. S.-Y., Milios, J., and Duncis, C. G. (1988) Antigen preservation in microwave-
irradiated tissues: a comparison with formaldehyde fixation. J. Pathol. 156, 275–282.
11. Fraenkel-Conrat, H., Brandon, B. A., and Olcott, H. S. (1947) The reaction of formaldehyde
with proteins. IV. Participation of indole groups. J. Biol. Chem. 168, 99–118.
12. Fraenkel-Conrat, H. and Olcott, H. S. (1948) Reaction of formaldehyde with proteins. VI.
Antigen Retrieval Technique for Immunohistochemistry 253
Cross-linking of amino groups with phenol, imidazole, or indole groups. J. Biol. Chem.
174, 827–843.
13. Fraenkel-Conrat, H. and Olcott, H. S. (1948) The reaction of formaldehyde with proteins.
V. Cross-linking between amino and primary amide or guanidyl groups. J. Am. Chem.
Soc. 70, 2673–2684.
14. Shi, S. R., Key, M. E., and Kalra, K. L. (1991) Antigen retrieval in formalin-fixed, paraffin-
embedded tissues: an enhancement method for immunohistochemical staining based on
microwave oven heating of tissue sections. J. Histochem. Cytochem. 39, 741–748.
15. Shi, S. R., Cote, C., Kalra, K. L., Taylor, C. R., and Tandon, A. K. (1992) A technique
for retrieving antigens in formalin-fixed, routinely acid-decalcified, celloidin-embedded
human temporal bone sections for immunohistochemistry. J. Histochem. Cytochem. 40,
787–792.
16. Suurmeijer, A. J. and Boon, M. E. (1993) Notes on the application of microwaves for
antigen retrieval in paraffin and plastic tissue sections. Eur. J. Morphol. 31, 144–150.
17. Cattoretti, G., Pileri, S., Parravicini, C., et al. (1993) Antigen unmasking on formalin-
fixed, paraffin-embedded tissue sections. J. Pathol. 171, 83–98.
18. Gown, A. M., de Wever, N., and Battifora, H. (1993) Microwave-based antigenic unmask-
ing. A revolutionary new technique for routine immunohistochemistry. Appl. Immunohisto-
chem. 1, 256–266.
19. Leong, A. S.-Y. and Milios, J. (1993) An assessment of the efficacy of the microwave
antigen-retrieval procedure on a range of tissue antigens. Appl. Immunohistochem. 1, 267–
274.
20. Swanson, P. E. (1993) Methodologic standardization in immunohistochemistry. A doorway
opens [Editorial]. Appl. Immunohistochem. 1, 229–231.
21. Dookhan, D. B., Kovatich, A. J., and Miettinen, M. (1993) Non-enzymatic antigen retrieval
in immunohistochemistry. Comparison between different antigen retrieval modalities and
proteolytic digestion. Appl. Immunohistochem. 1, 149–155.
22. Taylor, C. R., Shi, S. R., Chaiwun, B., Young, L., Imam, S. A., and Cote, R. J. (1994)
Strategies for improving the immunohistochemical staining of various intranuclear prog-
nostic markers in formalin-paraffin sections: androgen receptor, estrogen receptor, proges-
terone receptor, p53 protein, proliferating cell nuclear antigen, and Ki-67 antigen revealed
by antigen retrieval techniques. [See comments]. Hum. Pathol. 25, 263–270.
23. Cuevas, E. C., Bateman, A. C., Wilkins, B. S., et al. (1994) Microwave antigen retrieval
in immunocytochemistry: a study of 80 antibodies. J. Clin. Pathol. 47, 448–452.
24. Gu, J., Forte, M., Xenachis, C., Tarazona, N., Windsor, J., and Santoian, E. C. (1994)
Immunohistochemical demonstration of PCNA and Ki 67 positivities in stimulated myocha-
riocytes indicates dividing potential for cardiac muscle cells in adult heart. Cell Vision
1, 91–92.
25. Brown, R. W. and Chirala, R. (1995) Utility of microwave-citrate antigen retrieval in
diagnostic immunohistochemistry. Mod. Pathol. 8, 515–520.
26. Cote, R. J., Shi, S.-R., and Taylor, C. R. (1998) Development and standardization of
antigen retrieval (AR) for immunohistochemistry (IHC): current issues. J. Histochem.
Cytochem. 46, A13.
27. Elias, J. M. (1998) Technical considerations in immunocytochemistry. Cell Vision 5, 35–36.
28. Shin, R.-W., Iwaki, T., Kitamoto, T., and Tateishi, J. (1991) Hydrated autoclave pretreat-
ment enhances TAU immunoreactivity in formalin-fixed normal and Alzheimer’s disease
brain tissues. Lab. Invest. 64, 693–702.
29. Bankfalvi, A., Navabi, H., Bier, B., Bocker, W., Jasani, B., and Schmid, K. W. (1994)
Wet autoclave pretreatment for antigen retrieval in diagnostic immunohistochemistry.
J. Pathol. 174, 223–228.
30. Igarashi, H., Sugimura, H., Maruyama, K., et al. (1994) Alteration of immunoreactivity
254 Taylor and Shi
in papillary thyroid cancers is associated with two established indicators of poor prognosis.
Exp. Mol. Pathol. 62, 52–62.
49. Stieber, A., Mourelatos, Z., and Gonatas, N. K. (1996) In Alzheimer’s disease the Golgi
apparatus of apopulation of neurons without neurofibrillary tangles is fragmented and
atrophic. Am. J. Pathol. 148, 415–426.
50. Momose, H., Mehta, P., and Battifora, H. (1993) Antigen retrieval by microwave irradiation
in lead thiocyanate. Appl. Immunohistochem. 1, 77–82.
51. Shi, S.-R., Cote, R. J., Young, L. L., and Taylor, C. R. (1997) Antigen retrieval immunohis-
tochemistry: practice and development. J. Histotechnol. 20, 145–154.
52. Cattoretti, G., Becker, M. H. G., Key, G., et al. (1992) Monoclonal antibodies against
recombinant parts of the Ki-67 antigen (MIB 1 and MIB 3) detect proliferating cells in
microwave-processed formalin-fixed paraffin sections. J. Pathol. 168, 357–363.
53. Swanson, P. E. (1994) Microwave antigen retrieval in citrate buffer. Lab. Med. 25, 520–522.
54. Hopwood, D. (1994) Epitope retrieval—survey and prospect. Eur. J. Morphol. 32, 317–
324.
55. Shi, S.-R., Gu, J., Kalra, K. L., Chen, T., Cote, R. J., and Taylor, C. R. (1995) Antigen
retrieval technique: a novel approach to immunohistochemistry on routinely processed
tissue sections. Cell Vision 2, 6–22.
56. Boon, M. E. and Kok, L. P. (1995) Breakthrough in pathology due to antigen retrieval.
Mal. J. Med. Lab. Sci. 12, 1–9.
57. Cattoretti, G. and Suurmeijer, A. J. H. (1995) Antigen unmasking on formalin-fixed
paraffin-embedded tissues using microwaves: a review. Adv. Anat. Pathol. 2, 2–9.
58. Taylor, C. R., Shi, S.-R., and Cote, R. J. (1996) Antigen retrieval for immunohistochemistry.
Status and need for greater standardization. Appl. Immunohistochem. 4, 144–166.
59. Werner, M., Von Wasielewski, R., and Komminoth, P. (1996) Antigen retrieval, signal
amplification and intensification in immunohistochemistry. Histochem. Cell Biol. 105, 253–
260.
60. Shi, S. R., Cote, R. J., and Taylor, C. R. (1997) Antigen retrieval immunohistochemistry:
past, present, and future. J. Histochem. Cytochem. 45, 327–343.
61. Shi, Y., Li, G.-D., and Liu, W.-P. (1997) Recent advances of the antigen retrieval technique.
Linchuang Yu Shiyan Binglixue Zazhi (J. Clin. Exp. Pathol.). 13, 265–267.
62. Shi, S. R., Cote, R. J., and Taylor, C. R. (1998) Antigen retrieval immunohistochemistry
used for routinely processed celloidin-embedded human temporal bone sections: standard-
ization and development. Auris Nasus Larynx 25, 425–443.
63. McNicol, A. M. and Richmond, J. A. (1998) Optimizing immunohistochemistry: antigen
retrieval and signal amplification. Histopathology 32, 97–103.
64. Evers, P., Uylings, H. B., and Suurmeijer, A. J. (1998) Antigen retrieval in formaldehyde-
fixed human brain tissue. Methods 15, 133–140.
65. Shi, S. R., Cote, R. J., Young, L., Imam, S. A., and Taylor, C. R. (1996) Use of pH 9.5
Tris-HCl buffer containing 5% urea for antigen retrieval immunohistochemistry. Biotech.
Histochem. 71, 190–196.
66. Shi, S. R., Chaiwun, B., Young, L., Cote, R. J., and Taylor, C. R. (1993) Antigen retrieval
technique utilizing citrate buffer or urea solution for immunohistochemical demonstration
of androgen receptor in formalin- fixed paraffin sections. J. Histochem. Cytochem. 41,
1599–1604.
67. Shi, S. R., Chaiwun, B., Young, L., Imam, A., Cote, R. J., and Taylor, C. R. (1994)
Antigen retrieval using pH 3.5 glycine-HCl buffer or urea solution for immunohistochemi-
cal localization of Ki-67. Biotech. Histochem. 69, 213–215.
68. Imam, S. A., Young, L., Chaiwun, B., and Taylor, C. R. (1995) Comparison of two
microwave based antigen-retrieval solutions in unmasking epitopes in formalin-fixed tissue
for immunostaining. Anticancer Res. 15, 1153–1158.
256 Taylor and Shi
69. Hazelbag, H. M., van den Broek, L. J., van Dorst, E. B., Offerhaus, G. J., Fleuren, G. J.,
and Hogendoorn, P. C. (1995) Immunostaining of chain-specific keratins on formalin-
fixed, paraffin-embedded tissues: a comparison of various antigen retrieval systems using
microwave heating and proteolytic pre-treatments. J. Histochem. Cytochem. 43, 429–437.
70. Kwaspen, F., Smedts, F., Blom, J., et al. (1995) Periodic acid as a nonenzymatic enhance-
ment technique for the detection of cytokeratin immunoreactivity in routinely processed
carcinomas. Appl. Immunohistochem. 3, 54–63.
71. Pileri, S. A., Roncador, G., Ceccarelli, C., et al. (1997) Antigen retrieval techniques in
immunohistochemistry: comparison of different methods. J. Pathol. 183, 116–123.
72. van den Berg, F. M., Baas, I. O., Polak, M. M., and Offerhaus, G. J. A. (1993) Detection
of p53 overexpression in routinely paraffin-embedded tissue of human carcinomas using
a novel target unmasking fluid. Am. J. Pathol. 142, 381–385.
73. Swanson, P. E. (1997) HIERanarchy: the state of the art in immunohistochemistry. Am.
J. Clin. Pathol. 107, 139–140.
74. Taylor, C. R. and Cote, R. J. (1994) Immunomicroscopy: A Diagnostic Tool for the Surgical
Pathologist, 2nd ed. W. B. Saunders, Philadelphia.
75. Taylor, C. R. (1994) The current role of immunohistochemistry in diagnostic pathology.
Adv. Pathol. Lab. Med. 7, 59–105.
76. Katoh, A. and Breier, S. (1994) Nonspecific antigen retrieval solutions. J. Histotechnol.
17, 378.
77. O’Reilly, P. E., Raab, S. S., Niemann, T. H., Rodgers, J. R., and Robinson, R. A. (1997)
p53, proliferating cell nuclear antigen, and Ki-67 expression in extrauterine leiomyosarco-
mas. Mod. Pathol. 10, 91–97.
78. Shi, S.-R., Cote, R. J., Chaiwun, B., et al. (1998) Standardization of immunohistochemistry
based on antigen retrieval technique for routine formalin-fixed tissue sections. Appl. Immu-
nohistochem. 6, 89–96.
79. Shi, S. R., Imam, S. A., Young, L., Cote, R. J., and Taylor, C. R. (1995) Antigen retrieval
immunohistochemistry under the influence of pH monoclonal antibodies. J. Histochem.
Cytochem. 43, 193–201.
80. Evers, P. and Uylings, H. B. (1994) Microwave-stimulated antigen retrieval is pH and
temperature dependent. J. Histochem. Cytochem. 42, 1555–1563.
81. Evers, P. and Uylings, H. B. (1997) An optimal antigen retrieval method suitable for
different antibodies on human brain tissue stored for several years in formaldehyde fixative.
J. Neurosci. Methods. 72, 197–207.
82. Taylor, C. R., Shi, S.-R., Chaiwun, B., et al. (1994) Correspondence. Standardization and
reproducibility in diagnostic immunohistochemistry. Hum. Pathol. 25, 1107–1109.
83. Jones, M. D., Banks, P. M., and Caron, B. L. (1981) Transition metal salts as adjuncts to
formalin for tissue fixation. Lab. Invest. 44, 32A.
84. Shi, S. R., Tandon, A. K., Cote, C., and Kalra, K. L. (1992) S-100 protein in human inner
ear: use of a novel immunohistochemical technique on routinely processed, celloidin-
embedded human temporal bone sections. Laryngoscope 102, 734–738.
85. Shi, S. R., Tandon, A. K., Haussmann, R. R., Kalra, K. L., and Taylor, C. R. (1993)
Immunohistochemical study of intermediate filament proteins on routinely processed,
celloidin-embedded human temporal bone sections by using a new technique for antigen
retrieval. Acta Otolaryngol. (Stockh.) 113, 48–54.
86. Shi, S. R., Cote, R. J., Yang, C., et al. (1996) Development of an optimal protocol for
antigen retrieval: a ‘test battery’ approach exemplified with reference to the staining of
retinoblastoma protein (pRB) in formalin-fixed paraffin sections. J. Pathol. 179, 347–352.
87. Grossfeld, G. D., Shi, S. R., Ginsberg, D. A., et al. (1996) Immunohistochemical detection
of thrombospondin-1 in formalin-fixed, paraffin-embedded tissue. J. Histochem. Cytochem.
44, 761–766.
Antigen Retrieval Technique for Immunohistochemistry 257
88. Lan, H. Y., Hutchinson, P., Tesch, G. H., Mu, W., and Atkins, R. C. (1996) A novel
method of microwave treatment for detection of cytoplasmic and nuclear antigens by flow
cytometry. J. Immunol. Methods 190, 1–10.
89. Leers, M. P., Schutte, B., Theunissen, P. H., Raemaekers, F. C., and Nap, M. (1999) Heat
pretreatment increases resolution in DNA flow cytometry of paraffin-embedded tumor
tissue. Cytometry 35, 260–266.
90. Overton, W. R., Catalano, E., and McCoy, J. P. (1996) Method to make paraffin-embedded
breast and lymph tissue mimic fresh tissue in DNA analysis. Cytometry 26, 166–171.
91. Redkar, A. A. and Krishan, A. (1999) Flow cytometric analysis of estrogen, progesterone
receptor expression and DNA content in formalin-fixed, paraffin-embedded human breast
tumors. Cytometry 38, 61–69.
92. Boon, M. E., Beck, S., and Kok, L. P. (1995) Semiautomatic PAPNET analysis of proliferat-
ing (MiB-1-positive) cells in cervical cytology and histology. Diagn. Cytopathol. 13, 423–
428.
93. Brorson, S. H. (1998) The combination of high-accelerator epoxy resin and antigen retrieval
to obtain more intense immunolabeling on epoxy sections than on LR-White sections for
large proteins. Micron 29, 89–95.
94. Hand, N. M. and Church, R. J. (1998) Superheating using pressure cooking: its use and
application in unmasking antigens embedded in methyl methacrylate. J. Histotechnol.
21, 231–236.
95. Barou, O., Laroche, N., Palle, S., Alexandre, C., and Lafage-Proust, M. H. (1997) Pre-
osteoblastic proliferation assessed with BrdU in undecalcified, epon-embedded adult rat
trabecular bone. J. Histochem. Cytochem. 45, 1189–1195.
96. Blythe, D., Hand, N. M., Jackson, P., Barrans, S. L., Bradbury, R. D., and Jack, A. S.
(1997) Use of methyl methacrylate resin for embedding bone marrow trephine biopsy
specimens. J. Clin. Pathol. 50, 45–49.
97. Krenacs, T. and Rosendaal, M. (1998) Connexin43 gap junctions in normal, regenerating,
and cultured mouse bone marrow and in human leukemias: their possible involvement in
blood formation [See comments]. Am. J. Pathol. 152, 993–1004.
98. Sormunen, R. and Leong, A. S.-Y. (1998) Microwave-stimulated antigen retrieval for
immunohistology and immunoelectron microscopy of resin-embedded sections. Appl.
Immunohistochem. 6, 234–237.
99. D’Ambra-Cabry, K., Deng, D. H., Flynn, K. L., Magee, K. L., and Deng, J. S. (1995)
Antigen retrieval in immunofluorescent testing of bullous pemphigoid [See comments].
Am. J. Dermatopathol. 17, 560–563.
100. Fariss, R. N., Apte, S. S., Olsen, B. R., Iwata, K., and Milam, A. H. (1997) Tissue inhibitor
of metalloproteinases-3 is a component of Bruch’s membrane of the eye. Am. J. Pathol.
150, 323–328.
101. Al-Rifai, I., Kanitakis, J., Faure, M., and Claudy, A. (1997) Immunofluorescence diagnosis
of bullous dermatoses on formalin-fixed tissue sections after antigen retrieval [Letter;
comment]. Am. J. Dermatopathol. 19, 103–105.
102. Merz, H., Malisius, R., Mannweiler, S., et al. (1995) ImmunoMax. A maximized immuno-
histochemical method for the retrieval and enhancement of hidden antigens. Lab. Invest.
73, 149–156.
103. Lan, H. Y., Mu, W., Ng, Y. Y., Nilolic-Paterson, D. J., and Atkins, R. C (1996) A simple,
reliable, and sensitive method for nonradioactive in situ hybridization: use of microwave
heating to improve hybridization efficiency and preserve tissue morphology. J. Histochem.
Cytochem. 44, 281–287.
104. McMahon, J. and McQuaid, S. (1996) The use of microwave irradiation as a pretreatment
to in situ hybridization for the detection of measles virus and chicken anaemia virus in
formalin-fixed paraffin-embedded tissue. Histochem. J. 28, 157–164.
258 Taylor and Shi
105. Sibony, M., Commo, F., Callard, P., and Gasc, J.-M. (1995) Enhancement of mRNA in
situ hybridization signal by microwave heating. Lab. Invest. 73, 586–591.
106. Oliver, K. R., Heavens, R. P., and Sirinathsinghji, D. J. S. (1997) Quantitative comparison
of pretreatment regimens used to sensitive in situ hybridization using oligonucleotide
probes on paraffin-embedded brain tissue. J. Histochem. Cytochem. 45, 1707–1713.
107. Ikeda, K., Monden, T., Kanoh, T., et al. (1998) Extraction and analysis of diagnostically
useful proteins from formalin-fixed, paraffin-embedded tissue sections. J. Histochem.
Cytochem. 46, 397–404.
108. Baas, I. O., van den Berg, F. M., Mulder, J.-W. R., et al. (1996) Potential false-positive
results with antigen enhancement for immunohistochemistry of the p53 gene product in
colorectal neoplasms. J. Pathol. 178, 264–267.
109. Binks, S., Clelland, C. A., Ronan, J., and Bell, J. (1997) p53 gene product expression in
resected non-small cell carcinoma of the lung, with studies of concurrent cytological
preparations and microwave antigen retrieval. J. Clin. Pathol. 50, 320–323.
110. Daidone, M. G., Benini, E., Rao, S., Pilotti, S., and Silvestrini, R. (1998) Fixation time
and microwave oven irradiation affect immunocytochemical p53 detection in formalin-
fixed paraffin sections. Appl. Immunohistochem. 6, 140–144.
111. McCluggage, G., McBride, H., Maxwell, P., and Bharucha, H. (1997) Immunohistochemi-
cal detection of p53 and bcl-2 proteins in neoplastic and non-neoplastic endocervical
glandular lesions. Int. J. Gynecol. Pathol. 16, 22–27.
112. Sebenik, M. and Wieczorek, R. (1995) Nonspecific nuclear staining (NNS) after antigen
retrieval (AR). Lab. Invest. 72, 168A.
113. Wieczorek, R., Stover, R., and Sebenik, M. (1997) Nonspecific nuclear immunoreactivity
after antigen retrieval using acidic and basic solutions. J. Histotechnol. 20, 139–143.
114. Guiter, G. E., Kwan, P. W., and DeLellis, R. A. (1995) Unwanted tissue immunoreactivities
following microwave antigen retrieval: a critical analysis. Lab. Invest. 72, 165A.
115. Bussolati, G., Gugliotta, P., Volante, M., Pace, M., and Papotti, M. (1997) Retrieved
endogenous biotin: a novel marker and a potential pitfall in diagnostic immunohistochemis-
try [See comments]. Histopathology 31, 400–407.
116. Kashima, K., Yokoyama, S., Daa, T., Nakayama, I., Nickerson, P. A., and Noguchi, S.
(1997) Cytoplasmic biotin-like activity interferes with immunohistochemical analysis
of thyroid lesions: a comparison of antigen retrieval methods. Mod. Pathol. 10, 515–
519.
117. Mighell, A. J., Robinson, P. A., and Hume, W. J. (1995) Patterns of immunoreactivity
to an anti-fibronectin polyclonal antibody in formalin-fixed, paraffin-embedded oral
tissues are dependent on methods of antigen retrieval. J. Histochem. Cytochem.
43, 1107–1114.
118. Pateraki, M. and Kontogeorgos, G. (1997) Postadhesive technique for archival paraffin
sections. Biotech. Histochem. 72, 168–170.
119. Mote, P. A., Leary, J. A., and Clarke, C. L. (1998) Immunohistochemical detection of
progesterone receptors in archival breast cancer. Biotech. Histochem. 73, 117–127.
120. Krenacs, T. and Krenacs, L. (1999) Immunogold-silver staining (IGSS) for single and
multiple antigen detection in archived tissues following antigen retrieval. Appl. Immuno-
histochem. Mol. Morphol. 7, 93–94.
121. Grabau, K. A., Nielsen, O., Hansen, S., et al. (1998) Influence of storage temperature and
high-temperature antigen retrieval buffers on results of immunohistochemical staining in
sections stored for long periods. Appl. Immunohistochem. 6, 209–213.
122. Jacobs, T. W., Prioleau, J. E., Stillman, I. E., and Schnitt, S. J. (1996) Loss of tumor
marker-immunostaining intensity on stored paraffin slides of breast cancer. J. Natl. Cancer
Inst. 88, 1054–1059.
Antigen Retrieval Technique for Immunohistochemistry 259
123. Kato, J., Sakamaki, S., and Niitsu, Y. (1995) More on p53 antigen loss in stored paraffin
slides. N. Engl. J. Med. 333, 1507–1508.
124. Prioleau, J. and Schnitt, S. I. (1995) p53 antigen loss in stored paraffin slides. N. Engl.
J. Med. 332, 1521–1522.
125. Cote, R. J., Shi, Y., Groshen, S., et al. (1998) Association of p27kip1 levels with recurrence
and survival in patients with stage C prostate carcinoma. J. Natl. Cancer Inst. 90, 916–920.
126. Prento, P. and Lyon, H. (1997) Commercial formalin substitutes for histopathology. Bio-
tech. Histochem. 72, 273–282.
127. Zhang, P. J., Wang, H., Wrona, E. L., et al. (1998) Effects of tissue fixatives on antigen
preservation for immunohistochemistry: a comparative study of microwave antigen
retrieval on Lillie fixative and neutral buffered formalin. J. Histotechnol. 21, 101–106.
128. Allison, R. T. and Best, T. (1998) p53, PCNA and Ki-67 expression in oral squamous-
cell carcinomas—the vagaries of fixation and microwave enhancement of immunocyto-
chemistry. J. Oral Pathol. Med. 27, 434–440.
129. Carson, N. E., Gu, J., and Ianuzzo, C. D. (1998) Detection of myosin heavy chain in
skeletal muscles using monoclonal antibodies on formalin fixed, paraffin embedded tissue
sections. J. Histotechnol. 21, 19–24.
130. Elias, J. M. and Margiotta, M. (1997) Low temperature antigen restoration of steroid
hormone receptor proteins in routine paraffin sections. J. Histotechnol. 20, 155–158.
131. Koopal, S. A., Coma, M. I., Tiebosch, A. T. M. G., and Suurmeijer, A. J. H. (1998) Low-
temperature heating overnight in Tris-HCl buffer pH 9 is a good alternative for antigen
retrieval in formalin-fixed paraffin-embedded tissue. Appl. Immunohistochem. 6, 228–233.
132. Peston, D. and Shousha, S. (1998) Low temperature heat mediated antigen retrieval for the
demonstration of oestrogen and progesterone receptors in formlin-fixed paraffin sections.
J. Pathol. 186, A21.
133. Malisius, R., Merz, H., Heinz, B., Gafumbegete, E., Koch, B. U., and Feller, A. C. (1997)
Constant detection of CD2, CD3, and CD5 in fixed and paraffin-embedded tissue using
the peroxidase-mediated deposition of biotin-tyramide. J. Histochem. Cytochem. 45, 1665–
1672.
134. Butmarc, J. R., Koureau, H. P., Levi, E., and Kadin, M. E. (1998) Improved detection of
CD5 epitope in formalin-fixed paraffin-embedded sections of benign and neoplastic lymph-
oid tissues by using biotinylated tyramine enhancement after antigen retrieval. Am. J. Clin.
Pathol. 109, 682–688.
135. Dorfman, D. M. and Shahsafaei, A. (1997) Usefulness of a new CD5 antibody for the
diagnosis of T-cell and B-cell lymphoproliferative disorders in paraffin sections. Mod.
Pathol. 10, 859–863.
136. Kawai, K. and Osamura, R. Y. (1998) Epitope retrieval vs. amplification techniques in
diagnostic immunohistochemistry. J. Histochem. Cytochem. 46, A14.
137. Moll, B. and Jelveh, Z. (1998) Antigen recovery for vimentin immunostaining by reducing
agents for two cases of spindle cell tumors. J. Histotechnol. 21, 45–48.
138. Shi, S.-R., Gu, J., Kalra, K. L., Chen, T., Cote, R. J., and Taylor, C. R. (1997) Antigen
retrieval technique: a novel approach to immunohistochemistry on routinely processed
tissue sections. In: Go, J. (ed.). Analytical Morphology: Theory, Application, and Protocols,
Eaton Publishing, Natick, MA, pp. 1–40.
139. Tacha, D. E. and Chen, T. (1994) A modified antigen retrieval method. A calibration
technique for microwave ovens. J. Histotechnol. 17, 365–366.
140. Lan, H. Y., Mu, W., Nikolic-Patterson, D. J., and Atkins, C. A. (1995) A novel, simple,
reliable, and sensitive method for multiple immunoenzyme staining: use of microwave oven
heating to block antibody crossreactivity and retrieve antigens. J. Histochem. Cytochem. 43,
97–102.
260 Protocols
141. Stirling, J. W. and Graff, P. S. (1995) Antigen unmasking for immunoelectron microscopy:
labeling is improved by treating with sodium ethoxide or sodium metaperiodate, then
heating on retrieval medium. J. Histochem. Cytochem. 43, 115–123.
142. Wilson, D. F., Jiang, D.-J., Pierce, A. M., and Wiebkin, O. W. (1996) Antigen retrieval
for electron microscopy using a microwave technique for epithelial and basal lamina
antigens. Appl. Immunohistochem. 4, 66–71.
143. Strater, J., Gunthert, A. R., Bruderlein, S., and Moller, P. (1995) Microwave irradiation
of paraffin-embedded tissue sensitizes the TUNEL method for in situ detection of apoptotic
cells. Histochemistry 103, 157–160.
144. Lucassen, P. J., Labat-Moleur, F., Negoescu, A., and van Lookeren Campagne, M. (2000)
Microwave-enhanced in situ end-labeling of apoptotic cells in tissue sections; pitfalls and
possibilities. In: Shi, S.-R., Gu, J., and Taylor, C. R., eds., Antigen Retrieval Techniques:
Immunohistochemistry and Molecular Morphology. Eaton, Natick, MA, pp. 71–91.
145. Harkins, L. E. and Grizzle, W. E. (1995) Advanced techniques in immunohistochemistry,
Workshop #35, National Society for Histotechnology Annual Symposium 1995, Buffalo,
New York, October 7–13, 1995.
146. Shi, S.-R., Cote, R. J., and Taylor, C. R. (1999) Standardization and further development
of antigen retrieval immunohistochemistry: strategies and future goals. J. Histotechnol.
22, 177–192.
PROTOCOLS
METHODS
Requirements for Routinely Formalin-Fixed, Paraffin-Embedded Tissue
Sections Prior to AR Treatment
1. Mount the tissue sections on slides coated with either poly-l-lysine or APES, or charged
slides provided by Fisher Scientific (Pittsburgh, PA), dry the slides at 60°C for at least 1 h
in order to make the tissue sections adhere to the slides. Overnight incubation of the slides
produces optimal adhesion.
2. Deparaffinize by using the routine procedure: histoclear or xyline, rehydration by graded
alcohols, and blocking of endogenous peroxidase by an H2O2-methanol solution.
3. Rinse the slides in distilled water, 3 changes for 15 min.
setting for AR (S) is: S = 250/800 × 10 = 3.1. Therefore, it should be set at 3 and heated
for 7–10 min.
Steam Heating AR
For details, see refs. 35 and 37.
1. Use Steam HIER (Heat-Induced Epitope Retrieval) provided by BioTek (Ventura, CA) for
convenient use with the BioTek autostainer.
2. Set the slides in the TechMate slide holder in the regular face-to-face orientation, maintaining
the capillary gap.
3. Fill the steamer with distilled water to the top line, and heat it to boiling point. Then turn
the dial to 30 min for the next step.
4. Add 10 mL to each of the 10-well trays in the gray tile provided by BioTek, and place the
slide holder on the gray tile with the tips of the slide pairs in the AR solution located in
AR Technique 263
the 10-well trays. Then place the gray tile with the slide holder into the center of the
steam chamber.
5. Place the whole steam chamber base on the steamer, while boiling vigorously for 20 min.
6. After cooling the heated slides for 15 min, follow the remaining procedure for IHC as in
the microwave heating method.
7. If a third or fourth primary antibody is required for multiple immunolabeling, step 5 (micro-
wave heating procedure) should be repeated between each of the immunostaining procedures.
For paraffin sections, the same procedure (i.e., step 5 microwave heating) is used
between each immunostaining sequence, except that: 1) microwave heating is used
prior to the first primary antibody staining as in AR (i.e., before step 1); and 2) as
some surface antigens may not tolerate the two 5-min microwave heating periods, this
method may be limited to nuclear and cytoplasmic antigens.
AR Immunoelectron Microscopy
For details, see ref. 141.
1. Cut routinely processed epoxy resin-embedded tissue block, flatten with chloroform vapor,
and collect on Formvar-coated nickel grids.
2. Etch the grids with either a saturated aqueous solution of sodium metaperiodate for 1 h or
10% fresh saturated solution of sodium ethoxide (prepared overnight) diluted with anhydrous
ethanol for 2 min.
3. Heat the grids in a preheated microwave oven to the boiling point; heat the sections at
95–100°C by a hotplate for 10 min. (Citrate buffer pH 6.0 is recommended as the AR solution.)
4. After heating, allow the grids to cool for 15 min before immunostaining.
INTRODUCTION
Immunohistochemistry is an important technique in both investigative and diagnostic
pathology, for correlating the localization of antigens and tissue morphology. Since
the advent of immunocytochemistry in the late 1940s, significant progress has been made
in methods for labeling, visualizing, and detecting cell constituents. Two techniques have
been widely accepted for the demonstration of protein: the peroxidase anti-peroxidase
(PAP) (1) technique and the avidin-biotin-complex (ABC) (2) method. The PAP tech-
nique uses peroxidase as the labeling agent, and ABC is based on the noncovalent
binding between avidin and biotin. The production of highly specific polyclonal and
monoclonal antibodies improves sensitivity as well as specificity of immunocytochemi-
cal reactions and allows visualization of specific proteins. More recently, heat-induced
epitope retrieval (3,4) has become one of the most effective techniques for enhancing
the sensitivity of immunohistochemistry. This method significantly improves the number
of antigens that can be demonstrated in routine formalin-fixed paraffin sections.
Another amplification system called catalyzed reporter deposition (CARD) or the
tyramide amplification technique (TAT) has been reported recently. In 1989, Bobrow
et al. (5) described a novel application of horseradish peroxidase (HRP) as an analyte-
dependent reporter enzyme to catalyze the deposition of additional peroxidase molecules
on the surface in a solid phase immunoassay. With the use of a biotinylated tyramide
amplification step, this technique produced about a 200-fold increase in sensitivity in
solid phase assay (5). CARD has been successfully incorporated into immunohistochem-
ical and in situ hybridization techniques in many fields of investigation.
PRINCIPLE
This method relies on the ability of HRP to catalyze the dimerization of biotinylated
tyramine (tyramide), followed by the deposition of a large number of avidin-biotin-
peroxidase complexes or peroxidase-labeled streptoavidin molecules on the complex
(Fig. 1). The deposition of immune complex oxidizes biotinylated tyramine in the
presence of hydrogen peroxide to produce radical species that react with electron
267
268 Sanno et al.
Tyramide Amplification in Immunohistochemistry 269
moieties (e.g., tyrosine, tryptophans, and others) of protein molecules present in the
vicinity of the HRP label (5). The net effect is that many peroxidase molecules surround
a single HRP label.
Table 1
Comparison of the Immunohistochemical Sensitivities for Detecting Follicle-
Stimulating Hormone- (FSH-) by the CARD System, the Indirect Immuno-
peroxidase Method, and the ABC Method Using Graded Dilution of Primary
Antibodya
Dilution of anti-FSH-β mouse monoclonal antibody 1
Method 1 × 107 1 × 106 5 × 105 1 × 105 5 × 104 1 × 104 5 × 103 1 × 103 5 × 102 1 × 102
CSA − ++ + +++ +++ HBG HBG HBG HBG HBT
ID − − − − − − − + + ++
ABC − − − − − ++ ++ ++ ++ ++
a
CARD, catalyzed reporter deposition; ID, indirect immunoperoxidase method; ABC, avidin-biotin
complex method; HBG, high background. Grading: +, weakly positive; ++, positive, +++, strongly positive.
Fig. 2. (A,B) The catalyzed reporter deposition system for normal pituitary gland using anti-
FSH-β antibody diluted 1:100,000. Immunoreactivity for FSH-β is clearly demonstrated. (C,D)
Avidin-biotin complex (ABC) method using anti-FSH-β antibody diluted 1:100,000. Immuno-
reactivity for FSH-β was not detected. (E,F) Indirect immunoperoxidase method and original
magnification ×400. The sensitivity of CARD revealed a 1000-fold increase compared with the
indirect immunoperoxidase method and a 100-fold increase over the standard ABC method.
Fig. 3. Results of the catalyzed signal amplification system in the analysis of nonfunctioning
adenomas. (A) By standard ABC method on the same adenoma tissue, immunoreactivity for
FSH-β is observed only in scattered cells (case #28). (B) Using catalyzed signal amplification,
immunoreactivity for FSH-β is seen in the cytoplasm of the adenoma cells.
The fixation and preservation of tissues may also influence the results. Four percent
paraformaldehyde or 10% paraffin-embedded tissues usually work well (9,10).
Another advantage of CARD for immunohistochemistry is that lower concentrations
of antibodies can be used. In studies with antibodies available in limited amounts, this
may be beneficial.
for detection of DNA sequences and mRNA (20–23). Although recent advances in
evaluation of nonradioactive ISH led to greater detection sensitivity, sometimes a more
sensitive detection method for low copy number of gene products may be required.
In this procedure we use antisense and sense probes that are also employed for
conventional ISH. The probes are labeled with biotin 11dUTP (Boehringer Mannheim)
by the terminal deoxyribonucleotidyl transferase reaction. Target retrieval can be per-
formed by heating the frozen tissue sections in 10 mmol/L citric acid (pH 6.0) in a
microwave oven for 5 min (up to 95°C) and digested with 1 mg/mL proteinase K at
23°C for 10 min. The retrieval and digestion differ from tissues and expected signals.
Sections are then treated with 0.2 N HCl for 20 min followed by incubation with
0.25% (vol/vol) acetic anhydride in triethanolamine for 10 min. To reduce endogenous
peroxidase background, slides are immersed in 3% H2O2 in methanol for 30 min and
covered with prehybridization buffer for 1 h at room temperature. Thereafter, the
sections can be hybridized with the probe at 42°C for 18 h. After stringent washing
for 10 min at 42°C, slides are subject to amplification with reagents including primary
streptoavidin HRP (1:400) for 15 min, biotinyl-tyramide solution for 15 min, and
secondary streptavidin HRP for 15 min, with fresh 1× Tris-buffered saline/Tween
(TBST) washing after each step. The reaction product is visualized by developing the
slides in DAB chromogen/H2O2 solution for 3–5 min.
Using this amplification method, a positive reaction has been seen with antibodies
previously described as unsuitable for paraffin section immunostaining. Although the
CARD method involves multiple steps and is time consuming, it is a more sensitive
and specific method for the detection of proteins by immunostaining in many laboratory
investigations. The CARD system is one of the most powerful amplification methods
in immunochemistry developed to date.
REFERENCES
1. Sternberger, L. A., Hardy, P. H. Jr., Cuculis, J. J., and Meyer, H. G. (1970) The unlabeled
antibody enzyme method of immunocytochemistry. Preparation and properties of soluble
antigen-antibody complex (horseradish peroxidase-antiperoxidase) and its use in identifica-
tion of spirochetes. J. Histochem. Cytochem. 18, 315–333.
2. Hsu, S. M., Raine, L., and Fanger, H. (1981) Use of avidin-biotin-peroxidase complex (ABC)
in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP)
procedures. J. Histochem. Cytochem. 29, 577–581.
3. Merz, H., Malisius, R., Mannweiler, S., et al. (1995) ImmunoMax. A maximized immunohis-
tochemical method for the retrieval and enhancement of hidden antigens. Lab. Invest.
73, 149–156.
4. Shi, S. R., Key, M. E., and Kalra, K. L. (1991) Antigen retrieval in formalin-fixed, paraffin-
embedded tissues: an enhancement method for immunohistochemical staining based on
microwave oven heating of tissue sections. J. Histochem. Cytochem. 39, 741–748.
5. Bobrow, M. N., Harris, T. D., Shaughnessy, K. K., and Litt, G. J. (1989) Catalyzed
reporter deposition, a novel method of signal amplification. Application to immunoassays.
J. Immunol. Methods 125, 279–285.
6. Kerstens, M. J., Poddighe, P. J., and Hanselaar, A. G. J. M. (1995) A novel in situ
hybridization amplification method based on the deposition of biotinylated tyramine. J.
Histochem. Cytochem. 43, 347–352.
Tyramide Amplification in Immunohistochemistry 275
7. Dako Corporation. (1998) Instructions for DAKO catalyzed signal amplification (CSA)
system, peroxidase for mouse primary antibodies. Dako Corporation, Carpinteria, CA.
8. Sanno, N., Teramoto, A., Sugiyama, M., Itoh, Y., and Osamura, R. Y. (1996) Application
of catalyzed signal amplification in immunodetection of gonadotropin subunits in clinically
nonfunctioning pituitary adenomas. Am. J. Clin. Pathol. 106, 16–21.
9. von Wasielewski, R., Mengel, M., Gignac, S., Wilkens, L., Werner, M., and Georgii A.
(1997) Tyramine amplification technique in routine immunohistochemistry J. Histochem.
Cytochem. 45, 1455–1459.
10. King, G., Payne, S., Walker, F., and Murray, G. I. (1997) A highly sensitive detection
method for immunohistochemistry using biotinylated tyramine. J. Pathol. 183, 237–241.
11. Mengel, M., Werner, M., and von Wasielewski. (1999) Concentration dependent and adverse
effects in immunohistochemistry using the tyramine amplification technique. Histochem.
J. 31, 195–200.
12. Kaufmann, O., Baume, H., and Dietel, H. (1998) Detection of oestrogen receptors in
non-invasive and invasive transitional cell carcinomas of the urinary bladder using both
conventional immunohistochemistry and the thyramide staining amplification (TSA) tech-
nique. J. Pathol. 186, 165–168.
13. Tajima, A., Osamura, R. Y., Takekoshi, S., et al. (1999) Distribution of adrenomedullin
(AM), proadrenomedullin N-terminal 20 peptide, and AM mRNA in the rat gastric mucosa
by immunocytochemistry and in situ hybridization. Histochem. Cell Biol. 112, 139–146.
14. Kohler, A., Lauritzen, B., and Van Noorden, J. F. C. (2000) Signal amplification in immuno-
histochemistry at the light microscopic level using biotinylated tyramide and nanogold-
silver staining. J. Histochem. Cytochem. 48, 933–942.
15. Hunyady, B., Krempels, K., Harta, G., and Meqey, E. (1996) Immunohistochemical signal
amplification by catalyzed reporter deposition and its application in double immunostaining.
J. Histochem. Cytochem. 44, 1353–1362.
16. Teramoto, N., Szekely, L., Pokrovskaja, K., et al. (1998) Simultaneous detection of two
independent antigens by double staining with two mouse monoclonal antibodies. J. Virol.
Methods 73, 89–97.
17. Mayer, G. and Bendayan, M. (1997) Biotinyl-tyramide: a novel approach for electron
microscopic immunocytochemistry. J. Histochem. Cytochem. 45, 1449–1454.
18. Schmidt, B. F., Chao, J., Zhu, Z., DeBiasio, R. L., and Fisher, G. (1997) Signal amplification
in the detection of single-copy DNA and RNA by enzyme-catalyzed deposition (CARD)
of the novel fluorescent reporter substrate tyramide. J. Histochem. Cytochem. 45, 365–373.
19. Matsuno, A., Sanno, N., Tahara, S., et al. (1999) Silent somatotroph adenoma, detected by
catalyzed signal amplification and non-radioisotopic in situ hybridization. Endocr. J. 46
(suppl.), S81–84.
20. Sanno, N. and Osamura, R. Y. (1998) Catalyzed reporter deposition method for amplifying
endocrine products. Endocr. Pathol. 9, 195–199.
21. von Gijlswijk, R. P. and Zijlmans, H. J. (1997) Fluorochrome-labeled tyramides: use in
immunohistochemistry and fluorescence in situ hybridization. J. Histochem. Cytochem.
45, 375–382.
22. Koji, T., Kanemitsu, Y., Hoshino, A., and Nakane, P. K. (1997) A novel amplification
method of nonradioactive in situ hybridization signal for specific RNA with biotinylated
tyramine. Acta Histochem. Cytochem. 30, 401–406.
23. Oka, H., Jin, L., Scheithauer, B. W., et al. (1999) Growth hormone-releasing hormone
receptor (GHRH-R) mRNA expression in human pituitary adenomas: a study by catalyzed
reporter deposition in situ hybridization (CARD-ISH). Endocr. Pathol. 10, 27–36.
276 Protocols
PROTOCOLS
CARD Immunohistochemistry
1. Place 3% hydrogen peroxide on the tissue specimen. Incubate for 5 min and rinse with
buffer (0.05 M Tris-HCl, pH 7.6, containing 0.3 M NaCl and 0.1% Tween 20).
2. Protein block: serum-free protein in phosphate-buffered saline (PBS) with 0.015 M sodium
azide. Incubate for 5 min. Do not rinse off protein block.
3. Primary antibody: mouse monoclonal or rabbit polyclonal. Incubate for 15 min and rinse
with buffer.
4. Link antibody: biotinylated rabbit anti-mouse (for mouse monoclonal primary antibody) or
mouse anti-rabbit (for rabbit polyclonal primary antibody) immunoglobulins in Tris-HCl
buffer containing carrier protein and 0.015 M sodium azide. Incubate for 15 min and rinse
with buffer.
5. Streptoavidin-biotin complex: horseradish peroxidase labeled. Incubate for 15 min and rinse
with buffer.
6. Amplification reagent: biotinyl tyramide and hydrogen peroxide. Incubate for 15 min and
rinse with buffer.
7. Streptoavidin-peroxidase. Incubate for 15 min and rinse with buffer.
8. Substrate-chromogen solution: DAB. Incubate for 1–5 min and rinse with DW.
9. Hematoxylin counterstain. Rinse with DW.
10. Mounting: use aqueous-based mounting medium.
16
Application of Immunohistochemistry in the
Diagnosis of Lymphoid Lesions
Paul J. Kurtin, MD
INTRODUCTION
The value of phenotypic studies to aid in the diagnosis and classification of reactive
and neoplastic disorders of lymphocytes and myeloid cells has been established. In
general, immunophenotyping has five main goals. First, it can be used to distinguish
reactive from neoplastic lymphoid proliferations in lymph nodes and in extranodal
sites. Second, it can help to determine that a poorly differentiated malignant neoplasm
is a lymphoma. Third, once a neoplasm has been determined to be a lymphoma or
leukemia, the marker studies can establish the lineage: B-cell, T-cell, NK cell, myeloid,
or histiocytic. These distinctions are clinically important and form the basis for lym-
phoma classification by the World Health Organization classification of neoplastic
diseases of lymphoid tissues (1). Fourth, phenotypic data can be used to help classify
lymphomas because some lymphomas express a characteristic constellation of antigens.
Finally, phenotypic data can be used to help distinguish Hodgkin’s disease from non-
Hodgkin’s lymphomas. Because many different phenotypic modalities are available
(flow cytometry, paraffin section immunohistochemistry, frozen section immunohisto-
chemistry, molecular genetics studies) there is often confusion about the best modality
to employ to solve specific differential diagnostic problems. The techniques are comple-
mentary, and none is perfect for every application. In this chapter, the phenotypic
characterization of malignant lymphomas and leukemias by immunohistochemistry
is discussed.
277
278 Kurtin
Table 1
Frozen Section Immunohistochemistry: Useful Markers of Hematolymphoid Cells
Routine B-cell-associated antigens Routine T-cell-associated antigens
CD20 CD2
CD23 CD3
CD10 CD5
κ light chains CD7
λ light chains
Other B-cell-associated antigens Other T-cell-associated antigens
CD19 CD1
CD21 CD4
CD22 CD8
IgG TCR-αβ
IgA TCR-γδ
IgM CD56
IgD CD103
CD103 (hairy cell leukemia)
Myeloid lineage-associated antigens Antigens for special applications
CD15 CD45
CD33 CD30
CD13 bcl-2
CD14 Terminal deoxynucleotidyl transferase
c-kit CD34
Table 2
Paraffin Section Immunohistochemistry: Useful Markers of
Hematolymphoid Cellsa
B-cell-associated antigens T-cell associated antigens
CD20 (L-26) CD3 (polyclonal anti-ε chain)
CD79a CD5
CD10 CD45R0 (UCHL-1)
CD23 CD43
CD45RA (4KB5) CD4
κ light chains CD8
λ light chains βF1
Immunoglobulin heavy chains CD1a
Myeloid lineage antigens Special application antigens
CD34 (blasts) CD45
Myeloperoxidase CD15
Lysozyme CD30
CD68 (KP-1 and PGM-1) Terminal deoxynucleotidyl transferase
CD15 bcl-2
Factor VIII Cyclin D1 (bcl-1)
CD61 p80/ALK-1
Hemoglobin Fascin
NK and cytotoxic lymphocyte antigens Dendritic cell antigens
CD56 CD21
CD57 CD35
Tia-1 CD1a
Granzyme B S-100
Fascin
a
Antigens listed in bold face type are the first line markers used to distinguish B- and T-cell phenotype
in non-Hodgkin’s lymphomas. All others are selectively applied on a case-by-case basis based on the
morphologic differential diagnosis.
(Fig. 1) (3,6–9). The remaining 15–20% lack sufficient surface or cytoplasmic immuno-
globulin to be detected by frozen section immunohistochemistry. Using the immunoper-
oxidase technique in paraffin sections of well-fixed tissue with antigen retrieval or
protease digestion, properly titered primary antibodies, and a sensitive detection system,
κ or λ immunoglobulin light chain restriction can be demonstrated in approximately
70% of B-cell lymphomas (Fig. 2) (2,10).
There are several pitfalls in staining frozen or paraffin sections for immunoglobulin.
Here are a few of which to be aware. Because immunoglobulin is ubiquitous in serum
and plasma, background staining of interstitial immunoglobulin can obscure cell-related
immunoglobulin. Certain extranodal sites, such as skin and tissues in which the lymphoid
infiltrates are associated with collagen sclerosis, also present a consistent problem with
background staining because the antiimmunoglobulin antibodies have a tendency to
bind to collagen. These problems are unavoidable but can be minimized by optimally
titered antisera and by using detergents such as Tween 20 in the immunohistochemistry
buffers and antibody diluents.
Diagnosis of Lymphoid Lesions 281
Fig. 1. Immunoperoxidase stains for κ (A) and λ (B) immunoglobulin light chains on frozen
sections of a case of B-cell small lymphocytic lymphoma. Note the distinct membrane staining
for κ light chains in the small lymphocytes and only hazy background staining for λ light chains
within the intermixed histiocytes. This is a monoclonal staining pattern for κ and supports a
diagnosis of malignant lymphoma of B-cell lineage. (See Color Plate 6 following page 208.)
282 Kurtin
Fig. 2. Immunoperoxidase stains for κ (A) and λ (B) immunoglobulin light chains on paraffin
sections of a case of mantle cell lymphoma. Note the distinct, perinuclear cytoplasmic staining
for κ, but not for λ light chains. (See Color Plate 6 following page 208.)
Diagnosis of Lymphoid Lesions 283
Fig. 3. B-cell small lymphocytic lymphoma. (A) Hematoxylin and eosin-stained section
demonstrating the typical cytologic features of this neoplasm. Frozen section immunoperoxidase
stains for (B) CD20, (C) CD5, (D) CD3, and (E) CD23 are also illustrated here. The neoplastic
cells are CD20-positive B-cells that express the T-cell lineage-associated antigen CD5 but that
are negative for CD3. Thus they exhibit an “aberrant phenotype.” In addition, they are positive
for CD23. Thus, they have a CD20, CD5, and CD23 positive phenotype that is characteristic
for B-cell small lymphocytic lymphoma (see also Table 7). (See Color Plate 6 following
page 208.)
Diagnosis of Lymphoid Lesions 285
Fig. 3. Continued
Fig. 3. Continued
other B-cell lymphomas will express CD43. CD43 can be demonstrated in both frozen
sections and in paraffin sections. In paraffin sections, CD43 is best preserved with B5
fixation, but can be unmasked in paraffin sections by using EDTA antigen retrieval.
Its expression is weaker on neoplastic B-lymphocytes than on normal T-cells.
Because there is no phenotypically demonstrable clonal marker for T-lymphocytes,
one of the most useful applications of immunophenotyping is to demonstrate an aberrant
phenotype in morphologically atypical proliferations of T-lymphocytes (3,23–27). An
aberrant phenotype will support a diagnosis of T-cell lymphoma, but a normal phenotype
does not exclude lymphoma. More than 70% of peripheral T-cell lymphomas have
aberrant phenotypes, and the presence of an aberrant phenotype correlates closely with
the ability to demonstrate clonal T-cell antigen receptor gene rearrangements using
molecular genetics techniques (28,29). In T-cell lymphomas, phenotypic aberrancy
usually manifests by loss of one of the pan-T-cell antigens that is expected to be present
on post-thymic T-lymphocytes: CD2, CD3, CD5, and/or CD7 (Fig. 4). The antigen
most frequently lost from neoplastic T-cells is CD7, followed in decreasing frequency
by CD5, CD3, and CD2. Other less common manifestations of phenotypic aberrancy
in T-cell populations are either absence of staining for both CD4 and CD8 or presence
of staining for both CD4 and CD8.
There are anatomic site- and disease-specific exceptions to the criteria for phenotypic
aberrancy. Reactive T-cell populations in the skin, such as those that occur in chronic
dermatitis and lichen planus as well as in other conditions, will frequently lack expression
of CD7 (30–32). Therefore, absence of CD7 from a T-cell population in the skin does
not constitute evidence in favor of the diagnosis of cutaneous T-cell lymphoma even
though most cases of cutaneous T-cell lymphoma lack CD7 expression (33–35). The
normal thymus (36) and the T-cell infiltrates in thymomas (37,38) contain substantial
numbers of normal dual-expressing CD4- and CD8-positive T-cells and T-cells that
Diagnosis of Lymphoid Lesions 287
Fig. 4. Angioimmunoblastic T-cell lymphoma stained in frozen sections for (A) CD2 and
(B) CD7. The neoplastic T-cells are positive for CD2 (as well as for CD3 and CD5; not
illustrated), but they are negative for CD7. This is an “aberrant phenotype” because the T-cells
are negative for an antigen (CD7) that is normally expressed on nonneoplastic T-cells. These
findings would support a diagnosis of a peripheral T-cell lymphoma. (See Color Plate 6
following page 208.)
288 Kurtin
are negative for both CD4 and CD8. Therefore, dual CD4 and CD8 positivity or double
CD4 and CD8 negativity should not be used as a criterion for lymphoma in the thymus
or in lymphocytes within mediastinal masses until normal thymus and thymoma are
considered in the differential diagnosis. As more blood and tissue specimens are studied
from patients with viral infections and immunodeficiencies (particularly the acquired
immunodeficiency syndrome [AIDS]), it has become clear that phenotypically unusual
peripheral blood- and tissue-based T-cell populations can be expanded in these condi-
tions. Finally, almost all the published data on phenotypic aberrancy in peripheral T-
cell lymphomas come from immunoperoxidase studies performed on frozen tissue
sections in which the immunoarchitecture of the process can be assessed. It is not clear
that flow cytometry gives similar results and it is not clear what the criteria for T-cell
phenotypic aberrancy should be when flow cytometry is employed as the phenotyping
modality. Therefore, supporting the diagnosis of a T-cell lymphoma on the basis of
phenotypic aberrancy is only valid in frozen tissue sections where direct correlation
between phenotype and morphology can be made and should be done only after carefully
considering the clinical situation, the anatomic site from which the specimen is obtained,
and the morphology of the process.
Demonstrating Protein Products of Oncogenes or Abnormally Expressed Genes
One final approach to supporting a diagnosis of malignant lymphoma appears to be
evolving with the advent of antibodies that recognize protein products of oncogenes
or abnormally translocated genes in certain malignant lymphomas. Three types of
malignant lymphoma have characteristic chromosomal translocations that result in the
overexpression of genes with the production of protein products that are characteristi-
cally not expressed in the normal cell counterparts of the neoplasm. These include bcl-
2 overexpression as the result of the t(14;18) in follicular lymphomas, cyclin D1 (bcl-
1) overexpression as the result of the t(11;14) in mantle cell lymphomas, and p80/
ALK-1 expression as the result of the t(2;5) in CD30-positive anaplastic large cell
lymphomas. bcl-2, cyclin D1, and p80/ALK-1 can all be recognized by paraffin section
immunohistochemistry using commercially available antibodies.
The best established application of immunohistologic demonstration of an oncogene
product to support a diagnosis of malignant lymphoma is showing bcl-2 protein expres-
sion in the intrafollicular lymphocytes of follicular lymphomas (39–43). In normal and
reactive lymph nodes stained with antibodies to bcl-2, the germinal center B-lympho-
cytes are negative for bcl-2. There is positivity for bcl-2 in mantle zone B-cells and a
subset of inter- and intrafollicular T-cells. By contrast, 80% of follicular lymphomas
show bcl-2 expression in the neoplastic follicular center cells (Fig. 5). Thus, faced with
the morphologic differential diagnosis of follicular lymphoma vs follicular hyperplasia,
demonstrating bcl-2 positivity in the intrafollicular B-cells strongly supports a diagnosis
of follicular lymphoma. On the other hand, absence of bcl-2 staining of intrafollicular
lymphocytes, although characteristic of follicular hyperplasia, does not exclude follicular
lymphoma because 20% of follicular lymphomas are bcl-2 negative. Recently, it has
been suggested that normal monocytoid B-lymphocytes in reactive lymph nodes do
not express bcl-2. By contrast, neoplastic monocytoid B-cells in nodal marginal zone
B-cell lymphomas are bcl-2 positive. Therefore, demonstrating bcl-2 expression by
Diagnosis of Lymphoid Lesions 289
Fig. 5. Follicular lymphoma stained for bcl-2 in paraffin sections. Note that the intrafollicular
lymphocytes strongly express bcl-2. This finding supports the diagnosis of follicular lymphoma.
(See Color Plate 6 following page 208.)
Fig. 6. Mantle cell lymphoma stained for cyclin D1 in paraffin sections. The neoplastic
lymphocytes exhibit nuclear cyclin D1 positivity. (See Color Plate 7 following page 208.)
speaking, if one finds a substantial number of small lymphocytes in any tissue site that
has nuclear staining for cyclin D1, one can support a diagnosis of malignant lymphoma,
specifically mantle cell type. Cyclin D1 overexpression also occurs in a subset of cases
of multiple myeloma (50,51) and is also observed in some adenocarcinomas and
squamous cell carcinomas.
Finally, there is no normal lymphoid cell population expressing the p80 protein that
can be demonstrated in some cases of CD30-positive anaplastic large cell lymphoma
(52–58). p80 is an 80-kDa chimeric protein that is produced as a result of the t(2;5)
as portions of the nucleophosmin gene from chromosome 5 and the anaplastic lymphoma
kinase gene (ALK-1) from chromosome 2 are deleted, and the remaining portions are
fused. The protein is antigenically distinct and is recognized by antibodies to p80.
Anaplastic lymphoma kinase can also be detected immunohistochemically. Because
the protein recognized by the anti-p80 antibodies is only present in the presence of the
t(2;5), demonstrating nuclear and cytoplasmic staining for p80 (Fig. 7 and Color Plate
7 following page 208) is tantamount to recognizing the t(2;5) and is strong supporting
evidence in favor of a diagnosis of CD30-positive anaplastic large cell lymphoma or one
of its variants (small cell, monomorphous, lymphohistiocytic, and so forth). Likewise,
anaplastic lymphoma kinase is not identifiable in normal lymphocytes, and demonstra-
ting nuclear and cytoplasmic staining for this protein can also be used to support a
diagnosis of CD30-positive anaplastic large cell lymphoma. Other chromosome partners
can be translocated to the anaplastic lymphoma kinase locus on chromosome 2p and
can result in overexpression of anaplastic lymphoma kinase. In these cases, cytoplasmic,
but not nuclear, ALK-1 positivity is observed.
Diagnosis of Lymphoid Lesions 291
Fig. 7 (Color Plate 7). CD30-positive anaplastic large cell lymphoma stained for p80 in
paraffin sections. The neoplastic cells exhibit both cytoplasmic and nuclear staining for p80.
Table 3
Paraffin Section Immunoperoxidase Approach to Undifferentiated
Large Cell Neoplasms
Tumor Type CD45 Keratin S-100/HMB45 PLAPa
Lymphoma +b − − −
Carcinoma − + ± ±
Melanoma − − + −
Seminoma − ±c − +
a
PLAP, placental alkaline phosphatase.
b
Some cases of CD30-positive anaplastic large cell lymphoma are CD45 negative. Utilization of T-
and B-cell lineage-specific antibodies, and antibodies to CD30 can help to identify these cases.
c
Some seminomas are positive for keratin using antibodies to low molecular weight keratin, such as
Cam 5.2. in a dot-like or wispy-appearing filamentous pattern.
Table 4
Paraffin Section Immunoperoxidase Approach to Small Blue Cell Tumors
Tumor type CD45 CD99 Neurofilaments Synaptophysin Desmin Keratin
Lymphoma + +a − − − −
Ewing sarcoma − + − ± − −
PNETb − + − + − −
Neuroblastoma − − + + − −
Rhabdomyosarcoma − − − − + −
Small cell carcinoma − − − ± − +
a
T-precursor lymphoblastic lymphoma in particular.
b
Peripheral neuroectodermal tumor.
Table 5
Paraffin Section Immunoperoxidase Approach to Blastic
Hematolymphoid Neoplasms
Disorder TdT CD34 CD43 MPO/Lys CD3 CD79a CD20 Cyclin D1
T-lymphoblastic + − + − + − − −
B-lymphoblastic + + ± − − + ± −
Myeloid/monocytic ± + + + − − − −
Burkitt’s lymphoma − − + − − + + −
Blastoid mantle cell − − + − − + + +
frozen section immunohistochemistry. Because many B-cell, T-cell, and NK-cell lineage
antigens can now be reliably detected in paraffin sections, a very small panel of
antibodies employed in paraffin sections will give accurate information about cell
lineage in >95% of cases of malignant lymphomas. The most useful B-cell lineage
antigens are CD20 (L-26) (59–63) and CD79a (64). CD20 is expressed by a high
percentage of cases of B-cell lymphomas (Fig. 8) and almost all of the remainder
will be positive for CD79a. Notable exceptions are the rare lymphomas that express
Diagnosis of Lymphoid Lesions 293
Fig. 8. Diffuse large B-cell lymphoma involving a lymph node. (A) Hematoxylin and eosin-
stained section and (B) immunoperoxidase stain for CD20 performed on a paraffin section of
the tumor. The morphologic features including effacement of the lymph node architecture and
the monomorphous population of large atypical lymphoid cells are diagnostic of malignant
lymphoma. The strong staining of the tumor cells for CD20 indicates that the lymphoma is of
the B-cell lineage. (See Color Plate 7 following page 208.)
294 Kurtin
immunoglobulin light or heavy chains as the only markers of B-cell lineage. They
usually arise in immunosuppressed hosts. T-cell lymphomas can almost all (75%) be
reliably identified by antibodies to CD3 (polyclonal) (62,63,65–67) (Fig. 9), and the
remainder will be positive for CD45RO (UCHL-1) (62,65,68). Because antibodies to
CD45RO lack absolute T-cell lineage specificity, and because NK lymphocytes can
express the CD3ε epitope recognized by the paraffin-reactive polyclonal CD3 antibodies,
βF1 is another excellent marker for T-cell lineage in paraffin sections (69–72). Therefore,
by using antibodies to CD20, CD79a, CD3, CD45RO, and the T-cell receptor β chain
(βF1), almost all cases of lymphoma can be reliably determined to be of the B- or T-
cell type in paraffin sections.
If frozen section immunoperoxidase stains are used to determine cell lineage, a
limited panel will also accurately help to determine phenotype. A typical frozen section
immunohistochemistry panel might include antibodies to the B-cell lineage-associated
antigens CD19, CD20, CD22, CD10, CD23, and κ and λ light chains and to the T-
cell lineage-associated antigens CD2, CD3, CD5, and CD7. This panel will successfully
determine B- or T-cell lineage in almost all lymphoma cases.
Recognizing NK cells by phenotypic characteristics can be problematic because they
have phenotypes that overlap with cytotoxic T-cells and granulocytes. They are separable
by immunophenotype using frozen tissue, as illustrated in Table 6. NK cell and cytotoxic
T-cell antigens that can be assessed in paraffin sections include CD3ε (polyclonal anti-
CD3 recognizes the CD3ε chain, which is expressed in both NK and T-cells) (73–75),
CD56 (76,77), CD8, Tia-1 (78), and granzyme B (79). In addition, some NK cell
neoplasms (e.g., nasal type NK/T cell lymphoma) exhibit strong, uniform Epstein-Barr
virus (EBV) expression using in situ hybridization techniques with probes that recognize
small EBV-encoded RNAs (EBER) (80). Because NK cells do not rearrange T-cell
receptor genes, an additional diagnostic criterion for NK cell malignancies is to demon-
strate absence of clonal T-cell receptor gene rearrangements (76).
Classifying Lymphomas Based on Phenotypic Characteristics
A detailed discussion of the diagnostic criteria for lymphoma classification is beyond
the scope of this chapter. An excellent synopsis of the morphologic and phenotypic
criteria by which lymphomas are classified can be found in the description of the
Revised European American Lymphoma Classification (81). Although lymphomas
should be classified by a combined morphologic and phenotypic approach, certain
lymphomas of B-cell lineage express characteristic constellations of antigens.
As a neoplasm of immature B-cells, the cells of B-lymphocyte precursor lymphoblas-
tic lymphoma/leukemia virtually always express terminal deoxynucleotidyl transferase
and are usually positive for CD34. Most cases express CD19 and CD10, and there is
variable expression of CD20 and CD22 (cytoplasmic and/or surface) (82–86). Examples
of B-precursor lymphoblastic lymphomas are almost always surface immunoglobulin
light chain negative. A subset of cases formerly referred to as “pre-B-cell phenotype”
is positive for cytoplasmic µ-immunoglobulin heavy chain and does not express surface
immunoglobulin heavy or light chains (87).
The phenotypic “fingerprints” of B cell small lymphocytic lymphoma/chronic lymph-
ocytic leukemia (3,13,81,88–91) (Fig. 3), mantle cell lymphoma (11,12,45,46,48,81,88,
92–94), splenic marginal zone B-cell lymphoma (81,95–99), marginal zone B-cell
Diagnosis of Lymphoid Lesions 295
Table 6
Phenotypic Characteristics of NK Cells and Cytotoxic T-Cells
Antigen NK Cells Cytotoxic T-cells
CD2 + +
CD3 − +
CD5 − +
CD7 + +
CD8 − +
CD16 + −
Tia-1/granzyme B + +
Antigen recognition receptor KIRa αβ or γδ
MHC restriction Class I Class II
Function Kill target cells TCR and ADCCb
a
Killing inhibitory receptor.
b
T-cell receptor and antibody-dependent cell-mediated cytotoxicity.
Table 7
Immunophenotypic Features of Small B-Cell Lymphoproliferative Disorders:
Frozen Section and Paraffin Section Approach
SIg Cyclin D1 CD20 CD23 CD10 CD5 CD3
Small lymphocytic/chronic Monoclonal − + + − + −
lymphocytic leukemia
Lymphoplasmacytic Monoclonal ± ± ± ± −
cIg in
plasma
cells
Mantle cella Monoclonal + + − − + −
Marginal zone: splenic, Monoclonal − + − − − −
nodal, and extranodal
Hairy cell leukemiab Monoclonal − + − − − −
Follicular lymphoma Monoclonal − + ± + − −
a
Cells in most cases of mantle cell lymphoma (>80%) exhibit nuclear cyclin D1 staining.
b
In addition to the above, the cells of hairy cell leukemia characteristically and brightly co-express
CD22 and CD11c and are positive for CD103 (frozen sections). By paraffin section immunohistochemistry,
hairy cells are also positive for tartrate-resistant acid phosphatase and DBA.44.
Fig. 10. Immunoperoxidase stain for CD10 performed on a paraffin section of a follicular
lymphoma. Note the strong cytoplasmic staining for CD10 in the neoplastic follicular center
cells. (See Color Plate 7 following page 208.)
small lymphocytic lymphoma and chronic lymphocytic leukemia. Some express CD5
without CD23, similar to mantle cell lymphoma, some are positive for CD10, and some
cases express neither CD5 nor CD23. However, a consistent defining feature of lymph-
oplasmacytic lymphoma is plasma cells within the lesion that have a monoclonal staining
pattern for κ or λ immunoglobulin light chains. Paraffin section immunohistochemistry
is best suited for this application.
Large B-cell lymphomas are almost always positive for CD19 and CD20
(13,91,106,115) (Fig. 8), a sizable subset expresses CD10 (13,91,106,108,115–118) and/
or bcl-2 (44,108,117–119), and rare examples are positive for CD5 (117,120). Surface
immunoglobulin expression occurs at a lower frequency in large B-cell lymphomas than
in the lower grade B-cell lymphomas, particularly those that arise in the mediastinum
(121). Up to 40% are immunoglobulin negative. Practically speaking, large B-cell lymph-
omas are sufficiently phenotyped in paraffin sections utilizing antibodies to CD20.
The high-grade histologic features of Burkitt’s lymphoma (intermediate cell size,
high mitotic rate, and intermixed tingible body macrophages) can render cases of
Burkitt’s lymphoma difficult to distinguish from B-precursor or T-precursor lymphob-
lastic lymphoma/leukemia by morphology alone. However, Burkitt’s lymphomas are
uniformly terminal deoxynucleotidyl transferase and CD34 negative, and they almost
always express surface immunoglobulin. In addition, they are usually positive for CD10
together with pan-B antigens, including CD19, CD20, and CD22 (13,81,91,122).
In rare cases, it is difficult by morphology alone to make the clinically important
distinction between multiple myeloma and either low-grade lymphomas with extensive
plasmacytic differentiation (lymphoplasmacytic lymphoma, MALT-type lymphoma) or
large B-cell lymphomas with plasmacytoid features (formerly termed immunoblastic
298 Kurtin
Table 8
Immunophenotypic Differences Between Multiple Myeloma and Non-Hodgkin’s
Lymphomas with Plasmacytic Differentiation: Paraffin Sectionsa
Marker Multiple myeloma Lymphoma
CD45 Negative to weak Strong positive
CD79a Positive Positive
CD20 Negative to weak Strong positive
CD138 (Syndecan) Positive Negative
Immunoglobulin heavy chain isotype Any except IgM IgM
Epithelial membrane antigen Positive Negative to weak
a
This distinction cannot always be made based solely on morphologic and phenotypic criteria; see text.
Fig. 11. Immunoperoxidase stain for CD23 performed on a paraffin section of an angioimmu-
noblastic T-cell lymphoma. Same case as Fig. 9. Note the strong CD23-positive follicular
dendritic cell meshworks that are characteristically observed in this type of T-cell lymphoma.
(See Color Plate 7 following page 208.)
Fig. 12. Immunoperoxidase stain performed on a paraffin section for the cytolytic granule
protein Tia-1 in a case of NK/T-cell lymphoma of the nasal type. The neoplastic cells exhibit
coarse granular cytoplasmic positivity, characteristic of the pattern of staining that is observed
for this antigen. (See Color Plate 7 following page 208.)
300 Kurtin
Table 9
Phenotypes of T-Lymphoblastic Lymphoma/Leukemia
Normal Counterpart CD1 CD2 CD7 CD3 CD4 CD8 TdT
Immature thymocyte a
− + + − − − +
Cortical thymocyte + + + +c + + +
Medullary thymocyte − + + +s ±b ±b ±
cCytoplasmic CD3 expression.
sSurface CD3 expression.
a
CD2, CD7, CD4, and TdT can all be expressed by myeloid lineage acute leukemias.
b
At the medullary thymocyte stage, CD4 and CD8 expression are mutually exclusive.
Fig. 13. Immunoperoxidase stain for CD30 performed on a paraffin section of a CD30-
positive anaplastic large cell lymphoma. Note the strong, uniform staining of the neoplastic
cells. The membrane and dot-like cytoplasmic staining pattern is characteristic for this marker.
Table 10
Distinction of CD30-Positive Anaplastic Large Cell Lymphoma from Hodgkin’s
Lymphoma Based on Immunophenotype and Genetic Studies
CD30-positive anaplastic Classical Hodgkin’s
Marker large cell lymphoma (%) lymphoma (%)
CD30 + (strong, all cells) + (variable, cell subset)
CD45 + −
CD15 ± (15%) + (80%)
CD43 + (70%) −
CD45RO + (50%) −
CD3 (paraffin) + (50%) −
EMA + −
p80/ALK-1 + −
Clonal T-cell receptor + (70%) −
gene rearrangements
CD3, CD45RO, and CD43 are expressed in at least 70% of cases. Frozen section
immunohistochemistry can be used to assess the tumor for other T-cell lineage antigens
such as CD2, CD5, and CD7 in cases that fail to exhibit a T-cell phenotype by paraffin
section immunohistochemistry. Recently, it has been shown that a substantial subset
of CD30-positive anaplastic large cell lymphomas expresses cytolytic granule proteins
such as Tia-1, granzyme B, and perforin (141). The remainder of the cases expresses
neither T-cell nor B-cell lineage markers. The recent World Health Organization classi-
fication of hematolymphoid malignancies does not recognize B-cell lineage CD30-
positive anaplastic large cell lymphoma as a distinct disorder. The characteristic pheno-
type of CD30-positive anaplastic large cell lymphoma and its distinction from Hodgkin’s
disease based on immunohistochemistry are listed in Table 10.
Table 11
Distinction of Classical Hodgkin’s Lymphoma from Non-Hodgkin’s Lymphoma
(NHL) Based on the Phenotype of the Large Cells
Marker Hodgkin’s lymphoma (classical) Non-Hodgkin’s lymphoma
CD45 Negative Positive
CD15 Positive (80%) Negative (90%)
CD30 Positive (90+%) Variable
CD20 Often negative if positive, Positive in most B-cell lineage NHL:
cell subset, variable intensity strong, uniform staining
CD3 Usually negative Positive in most T-cell lineage NHLs
Fascin Positive Negative to weak
sensitive marker for Reed-Sternberg cells and variants from cases of classical Hodgkin’s
lymphoma is fascin (Fig. 14), an actin bundling protein (153). Hodgkin’s cells are also
positive for activation antigens, including HLA-DR, interleukin-2 receptors (CD25),
the transferrin receptor (CD71) (150,154), and CD95 (APO-1/Fas) (155).
Reed-Sternberg cells of the classical types of Hodgkin’s lymphoma exhibit variable
expression from case to case of pan-B- and pan-T-cell markers. There have been
consistent reports of subpopulations of Reed-Sternberg cells positive for CD19, CD20,
CD22, and CD75 in up to 80% of the classical types of Hodgkin’s lymphoma (156–161).
T-cell and cytolytic lymphocyte granule-associated antigens are expressed by a small
percentage of cases of Hodgkin’s lymphoma of the nodular sclerosis and mixed cellular-
ity types (157,162–165). CD2, CD3, CD4, and the T-cell antigen receptor framework
antigens are the markers that have been most frequently studied in this regard. CD45RO
and CD43 are almost never expressed by Hodgkin’s cells. Finally, Reed-Sternberg cells
have only rarely been considered positive for antigens expressed on histiocytic or
myelomonocytic cells, except for CD15.
These observations can be exploited diagnostically to separate cases of classical
Hodgkin’s lymphoma from non-Hodgkin’s lymphomas (152). Paraffin section immuno-
histochemistry is preferred for this application. In general, if the large cells of a
lymphoma express CD15 and CD30 and are negative for CD45 (this is a critical negative
result), the diagnosis of Hodgkin’s lymphoma of one of the classical types is supported.
On the other hand, if the large cells of a lymphoma are positive for CD45 (95% of
non-Hodgkin’s lymphomas express CD45) and positive or negative for either CD15
or CD30, the diagnosis of non-Hodgkin’s lymphoma is supported. Expression of CD20
in cells of classical Hodgkin’s lymphoma, when observed, tends to be variable; some
cells are strongly positive, others are negative, and still others exhibit weak stain-
ing. Conversely, the cells of non-Hodgkin’s lymphomas of a B-cell lineage, including
T-cell-rich B-cell lymphomas, generally show strong and uniform staining for CD20.
T-cell lineage antigen expression by Hodgkin’s cells (CD3, CD45R0, or CD43) is
distinctly uncommon in paraffin sections and usually can be taken as evidence against
the diagnosis of Hodgkin’s lymphoma.
In cases of nodular lymphocyte predominance Hodgkin’s lymphoma, the L and H
variants of Reed-Sternberg cells have a distinctive and consistent phenotype, indicative
of B-cell lineage. They are positive for pan B-cell markers, such as CD20 and CD22 in
Diagnosis of Lymphoid Lesions 303
Fig. 14. Hodgkin’s lymphoma, nodular sclerosis type involving a lymph node. (A) Hematoxy-
lin and eosin-stained section demonstrating the large neoplastic cells admixed with small lympho-
cytes, eosinophils, plasma cells, and macrophages. Immunoperoxidase stains for (B) CD15, (C)
CD30, and (D) fascin performed on paraffin sections. The neoplastic cells are positive for all
three antigens. Note that the stains for CD15 and CD30 decorate the tumor cells in a membrane
and punctate perinuclear staining pattern, whereas fascin positivity is diffusely present in the
cytoplasm. (See Color Plate 8 following page 208.)
304 Kurtin
frozen sections and CD20 in paraffin sections (166–173) (Fig. 15 and Color Plate 8
following page 208). L and H variants of Reed-Sternberg cells consistently stain for J-
chain, a polypeptide synthesized exclusively by B-lymphocytes (174,175), and it has been
demonstrated (159,160), that L and H variants express CD79a, a component of the B-
cell-specific phosphoprotein heterodimer associated with the B-cell antigen receptor com-
plex. In contrast to other types of Hodgkin’s lymphoma, the L and H variants of Reed-
Sternberg cells are frequently positive for CD45; they lack staining for CD15, and they
have variable staining for CD30. In a subset of cases, they express epithelial membrane
antigen (166,169).
Diagnosis of Lymphoid Lesions 305
Fig. 15 (Color Plate 8). Hodgkin’s lymphoma, nodular lymphocyte predominance type,
involving a lymph node. Immunoperoxidase stain performed on paraffin sections for CD20.
The large L and H variants of Reed-Sternberg cells, characteristically observed in this type of
Hodgkin’s lymphoma, are positive for CD20. They are admixed with numerous CD20-positive
nonneoplastic small lymphocytes, but the cells in the immediate vicinity of the large cells are
negative for CD20. They are CD3- and CD57-positive T-cells (not illustrated).
Table 12
Paraffin Section Immunoperoxidase Characterization of Acute Myelogenous
Leukemia by FAB Typea
FAB type MPO Lysozyme CD68 FVIII/CD61 Hemoglobin CD34/TdT
M1/M0 + + − − − +
M2 + + − − − ±
M3 + + − − − −
M4 ± + ± − − ±
M5 − + + − − ±
M6 − − − − + −
M7 − − − + − −
a
FAB, French-American-British classification; MPO, myeloperoxidase; CD68, PG-M1; FVIII, factor
VIII-related antigen (von Willebrand factor); TdT, terminal deoxynucleotidyl transferase.
Fig. 17. Acute myelogenous leukemia, FAB subtype M1. Immunoperoxidase stain for myelo-
peroxidase. The blasts are variably positive for myeloperoxidase (see also Table 12). (See Color
Plate 8 following page 208.)
Fig. 18. Langerhans’ cell histiocytosis involving the parotid gland. (A) Hematoxylin and
eosin stain exhibits the typical cytologic features of this disorder. The architecture of the salivary
gland is effaced by a monomorphous population of large cells with grooved and irregular nuclei
and abundant pale eosinophilic cytoplasm. (B) Immunoperoxidase stain for CD1a performed
in paraffin sections. The tumor cells are positive. (See Color Plate 8 following page 208.)
Table 13
Phenotypic Characteristics of Antigen-Presenting Cell Tumors
Tumor type Fascin CD1a S-100 CD21 CD23 CD35
Langerhans’ cell histiocytosis + + + − − −
Follicular dendritic cell sarcoma + − − + + +
Interdigitating reticulum cell sarcoma + − + − − −
lineage specific, care must be taken to exclude other spindle cell sarcomas and sarcoma-
toid carcinomas when considering the diagnosis of a dendritic cell tumor.
Most cases formerly diagnosed as malignant histiocytosis have been shown to be
neoplasms of other cell lineages (e.g., CD30-positive anaplastic large cell lymphoma
and others) (206,207) or reactive processes characterized by prominent hemophago-
cytosis (198). The criteria for diagnosis of a true histiocytic malignancy (now termed
histiocytic sarcoma) include demonstrating expression of histiocytic lineage antigens
by the neoplastic cells and demonstrating absence of antigens expressed by T-cells, B-
cells, melanocytes, and epithelial cells (207–210). Because histiocytes do not rearrange
the T-cell antigen receptor genes or the immunoglobulin heavy and light chain genes,
sometimes this also requires application of molecular genetics techniques to demonstrate
absence of clonal rearrangements of these genes. In general, neoplastic histiocytes will
express CD45, CD45RO, CD45RA, CD43, lysozyme, CD68, and CD14 in varying
combinations (Fig. 19).
Diagnosis of Lymphoid Lesions 309
Fig. 19. Histiocytic sarcoma. Immunoperoxidase stains for (A) CD45 and (B) CD68 performed
in paraffin sections and for (C) CD13 performed on frozen sections of the tumor. The neoplastic
cells are positive for all three of these antigens and negative for a large number of T-cell and
B-cell lineage-associated antigens. In conjunction with the morphology (not illustrated), the
findings indicate that the neoplastic cells are of a histiocytic lineage. (See Color Plate 8 following
page 208.)
310 Kurtin
REFERENCES
1. Harris, N. L., Jaffe, E. S., Diebold, J., et al. (2000) The World Health Organization
classification of hematological malignancies report of the Clinical Advisory Committee
Meeting, Airlie House, Virginia, November 1997. Mod. Pathol. 13, 193–207.
2. Kurtin, P. J., Hobday, K. S., Ziesmer, S., and Caron, B. L. (1999) Demonstration of
Diagnosis of Lymphoid Lesions 311
37. Mokhtar, N., Hsu, S. M., Lad, R. P., Haynes, B. F., and Jaffe, E. S. (1984) Thymoma:
lymphoid and epithelial components mirror the phenotype of normal thymus. Hum. Pathol.
15, 378–384.
38. Sato, Y., Watanabe, S., Mukai, K., et al. (1986) An immunohistochemical study of thymic
epithelial tumors. II. Lymphoid component. Am. J. Surg. Pathol. 10, 862–870.
39. Zutter, M., Hockenbery, D., Silverman, G. A., and Korsmeyer, S. J. (1991) Immunolocaliza-
tion of the Bcl-2 protein within hematopoietic neoplasms. Blood 78, 1062–1068.
40. Pezzella, F., Tse, A. G., Cordell, J. L., Pulford, K. A., Gatter, K. C., and Mason, D. Y.
(1990) Expression of the bcl-2 oncogene protein is not specific for the 14;18 chromosomal
translocation. Am. J. Pathol. 137, 225–232.
41. Ngan, B. Y., Chen-Levy, Z., Weiss, L. M., Warnke, R. A., and Cleary, M. L. (1988)
Expression in non-Hodgkin’s lymphoma of the bcl-2 protein associated with the t(14;18)
chromosomal translocation. N. Engl. J. Med. 318, 1638–1644.
42. Krajewski, S., Bodrug, S., Gascoyne, R., Berean, K., Krajewska, M., and Reed, J. C.
(1994) Immunohistochemical analysis of Mcl-1 and Bcl-2 proteins in normal and neoplastic
lymph nodes. Am. J. Pathol. 145, 515–525.
43. Gaulard, P., d’Agay, M. F., Peuchmaur, M., et al. (1992) Expression of the bcl-2 gene
product in follicular lymphoma. Am. J. Pathol. 140, 1089–1095.
44. Lai, R., Arber, D. A., Chang, K. L., Wilson, C. S., and Weiss, L. M. (1998) Frequency
of bcl-2 expression in non-Hodgkin’s lymphoma: a study of 778 cases with comparison
of marginal zone lymphoma and monocytoid B-cell hyperplasia. Mod. Pathol. 11, 864–
869.
45. Zukerberg, L. R., Yang, W. I., Arnold, A., and Harris, N. L. (1995) Cyclin D1 expression
in non-Hodgkin’s lymphomas. Detection by immunohistochemistry. Am. J. Clin. Pathol.
103, 756–760.
46. Yang, W. I., Zukerberg, L. R., Motokura, T., Arnold, A., and Harris, N. L. (1994) Cyclin
D1 (Bcl-1, PRAD1) protein expression in low-grade B-cell lymphomas and reactive
hyperplasia. Am. J. Pathol. 145, 86–96.
47. Swerdlow, S. H., Zukerberg, L. R., Yang, W. I., Harris, N. L., and Williams, M. E. (1996)
The morphologic spectrum of non-Hodgkin’s lymphomas with BCL1/cyclin D1 gene
rearrangements. Am. J. Surg. Pathol. 20, 627–640.
48. Swerdlow, S. H., Yang, W. I., Zukerberg, L. R., Harris, N. L., Arnold, A., and Williams,
M. E. (1995) Expression of cyclin D1 protein in centrocytic/mantle cell lymphomas with
and without rearrangement of the BCL1/cyclin D1 gene. Hum. Pathol. 26, 999–1004.
49. Vasef, M. A., Medeiros, L. J., Koo, C., McCourty, A., and Brynes, R. K. (1997) Cyclin
D1 immunohistochemical staining is useful in distinguishing mantle cell lymphoma
from other low-grade B-cell neoplasms in bone marrow. Am. J. Clin. Pathol. 108, 302–307.
50. Hoyer, J. D., Hanson, C. A., Fonseca, R., Greipp, P. R., Dewald, G. W., and Kurtin,
P. J. (2000) The (11;14)(q13;q32) translocation in multiple myeloma. A morphologic and
immunohistochemical study. Am. J. Clin. Pathol. 113, 831–837.
51. Vasef, M. A., Medeiros, L. J., Yospur, L. S., Sun, N. C., McCourty, A., and Brynes, R. K.
(1997) Cyclin D1 protein in multiple myeloma and plasmacytoma: an immunohistochemi-
cal study using fixed, paraffin-embedded tissue sections. Mod. Pathol. 10, 927–932.
52. Benharroch, D., Meguerian-Bedoyan, Z., Lamant, L., et al. (1998) ALK-positive lymph-
oma: a single disease with a broad spectrum of morphology. Blood 91, 2076–2084.
53. Cataldo, K. A., Jalal, S. M., Law, M. E., et al. (1999) Detection of t(2;5) in anaplastic
large cell lymphoma: comparison of immunohistochemical studies, FISH, and RT-PCR
in paraffin-embedded tissue. Am. J. Surg. Pathol. 23, 1386–1392.
54. Falini, B., Bigerna, B., Fizzotti, M., et al. (1998) ALK expression defines a distinct group
of T/null lymphomas with a wide morphological spectrum. Am. J. Pathol. 153, 875–886.
314 Kurtin
55. Pittaluga, S., Wiodarska, I., Pulford, K., et al. (1997) The monoclonal antibody ALK1
identifies a distinct morphological subtype of anaplastic large cell lymphoma associated
with 2p23/ALK rearrangements. Am. J. Pathol. 151, 343–351.
56. Pulford, K., Lamant, L., Morris, S. W., et al. (1997) Detection of anaplastic lymphoma
kinase (ALK) and nucleolar protein nucleophosmin (NPM)-ALK proteins in normal and
neoplastic cells with the monoclonal antibody ALK1. Blood 89, 1394–1404.
57. Shiota, M., Fujimoto, J., Takenaga, M., et al. (1994) Diagnosis of t(2;5)(p23;q35)-associ-
ated Ki-1 lymphoma with immunohistochemistry. Blood 84, 3648–3652.
58. Simonitsch, I., Panzer-Gruemayer, E. R., Ghali, D. W., et al. (1996) NPM/ALK gene
fusion transcripts identify a distinct subgroup of null type Ki-1 positive anaplastic large
cell lymphomas. Br. J. Haematol. 92, 866–871.
59. Cartun, R. W., Coles, F. B., and Pastuszak, W. T. (1987) Utilization of monoclonal
antibody L26 in the identification and confirmation of B-cell lymphomas. A sensitive and
specific marker applicable to formalin-and B5-fixed, paraffin-embedded tissues. Am. J.
Pathol. 129, 415–421.
60. Mason, D. Y., Comans-Bitter, M. W., Cordell, J. L., Verhoeven, M. A., and van Dongen,
J. J. (1990) Antibody L26 recognizes an intracellular epitope on the B-cell-associated
CD20 antigen. Am. J. Pathol. 136, 1215–1222.
61. Norton, A. J. and Isaacson, P. G. (1987) Monoclonal antibody L26: an antibody that is
reactive with normal and neoplastic B lymphocytes in routinely fixed and paraffin wax
embedded tissues. J. Clin. Pathol. 40, 1405–1412.
62. Kurtin, P. J. and Roche, P. C. (1993) Immunoperoxidase staining of non-Hodgkin’s
lymphomas for T-cell lineage associated antigens in paraffin sections. Comparison of the
performance characteristics of four commercially available antibody preparations. Am. J.
Surg. Pathol. 17, 898–904.
63. Chadburn, A. and Knowles D. M. (1994) Paraffin-resistant antigens detectable by antibodies
L26 and polyclonal CD3 predict the B- or T-cell lineage of 95% of diffuse aggressive
non-Hodgkin’s lymphomas. Am. J. Clin. Pathol. 102, 284–291.
64. Mason, D. Y., Cordell, J. L., Brown, M. H., et al. (1995) CD79a: a novel marker for B-
cell neoplasms in routinely processed tissue samples. Blood 86, 1453–1459.
65. Cabecadas, J. M. and Isaacson, P. G. (1991) Phenotyping of T-cell lymphomas in paraffin
sections—which antibodies? Histopathology 19, 419–424.
66. Mason, D. Y., Cordell, J., Brown, M., et al. (1989) Detection of T cells in paraffin wax
embedded tissue using antibodies against a peptide sequence from the CD3 antigen. J. Clin.
Pathol. 42, 1194–1200.
67. Anderson, C., Rezuke, W. N., Kosciol, C. M., Pastuszak, W. T., and Cartun, R. W. (1991)
Methods in pathology. Identification of T-cell lymphomas in paraffin-embedded tissues
using polyclonal anti-CD3 antibody: comparison with frozen section immunophenotyping
and genotypic analysis. Mod. Pathol. 4, 358–362.
68. Norton, A. J., Ramsay, A. D., Smith, S. H., Beverley, P. C., and Isaacson, P. G. (1986)
Monoclonal antibody (UCHL1) that recognizes normal and neoplastic T cells in routinely
fixed tissues. J. Clin. Pathol. 39, 399–405.
69. Macon, W. R. and Salhany, K. E. (1998) T-cell subset analysis of peripheral T-cell
lymphomas by paraffin section immunohistology and correlation of CD4/CD8 results with
flow cytometry. Am. J. Clin. Pathol. 109, 610–617.
70. Said, J. W., Shintaku, I. P., Parekh, K., and Pinkus, G. S. (1990) Specific phenotyping of
T-cell proliferations in formalin-fixed paraffin-embedded tissues. Use of antibodies to the
T-cell receptor beta F1. Am. J. Clin. Pathol. 93, 382–386.
71. Chan, W. C., Borowitz, M. J., Hammami, A., Wu, Y. J., and Ip S. J. (1988) T-cell receptor
antibodies in the immunohistochemical studies of normal and malignant lymphoid cells.
Cancer 62, 2118–2124.
Diagnosis of Lymphoid Lesions 315
72. Ng, C. S., Chan, J. K., Hui, P. K., Chan, W. C., and Lo, S. T. (1988) Application of a T
cell receptor antibody beta F1 for immunophenotypic analysis of malignant lymphomas.
Am. J. Pathol. 132, 365–371.
73. Ohno, T., Yamaguchi, M., Oka, K., Miwa, H., Kita, K., and Shirakawa, S. (1995) Frequent
expression of CD3 epsilon in CD3 (Leu 4)-negative nasal T-cell lymphomas. Leukemia
9, 44–52.
74. Ohsawa, M., Nakatsuka, S., Kanno, H., et al. (1999) Immunophenotypic and genotypic
characterization of nasal lymphoma with polymorphic reticulosis morphology. Int. J.
Cancer 81, 865–870.
75. Suzumiya, J., Takeshita, M., Kimura, N., et al. (1994) Expression of adult and fetal natural
killer cell markers in sinonasal lymphomas. Blood 83, 2255–2260.
76. Robertson, M. J. and Ritz, J. (1990) Biology and clinical relevance of human natural killer
cells. Blood 76, 2421–2438.
77. Lanier, L. L., Phillips, J. H., Hackett, J., Jr., Tutt, M., and Kumar, V. (1986) Natural killer
cells: definition of a cell type rather than a function. J. Immunol. 137, 2735–2739.
78. Felgar, R. E., Macon, W. R., Kinney, M. C., Roberts, S., Pasha, T., and Salhany, K. E.
(1997) TIA-1 expression in lymphoid neoplasms. Identification of subsets with cytotoxic
T lymphocyte or natural killer cell differentiation. Am. J. Pathol. 150, 1893–1900.
79. de Bruin, P. C., Kummer, J. A., van der Valk, P., et al. (1994) Granzyme B-expressing
peripheral T-cell lymphomas: neoplastic equivalents of activated cytotoxic T cells with
preference for mucosa-associated lymphoid tissue localization. Blood 84, 3785–3791.
80. Jaffe, E. S., Chan, J. K., Su, I. J., Frizzera, G., Mori, S., Feller, A. C., and Ho, F. C.
(1996) Report of the Workshop on Nasal and Related Extranodal Angiocentric T/Natural
Killer Cell Lymphomas. Definitions, differential diagnosis, and epidemiology. Am. J. Surg.
Pathol. 20, 103–111.
81. Harris, N. L., Jaffe, E. S., Stein, H., et al. (1994) A revised European-American classification
of lymphoid neoplasms: a proposal from the International Lymphoma Study Group [See
comments]. Blood 84, 1361–1392.
82. Sallan, S. E., Ritz, J., Pesando, J., et al. (1980) Cell surface antigens: prognostic implications
in childhood acute lymphoblastic leukemia. Blood 55, 395–402.
83. Pui, C. H., Behm, F. G., and Crist, W. M. (1993) Clinical and biologic relevance of
immunologic marker studies in childhood acute lymphoblastic leukemia. Blood 82, 343–
362.
84. Khalidi, H. S., Chang, K. L., Medeiros, L. J., et al. (1999) Acute lymphoblastic leukemia.
Survey of immunophenotype, French-American-British classification, frequency of
myeloid antigen expression, and karyotypic abnormalities in 210 pediatric and adult cases.
Am. J. Clin. Pathol. 111, 467–476.
85. Borowitz, M. J. (1990) Immunologic markers in childhood acute lymphoblastic leukemia.
Hematol. Oncol. Clin. North Am. 4, 743–765.
86. Borowitz, M. J., Shuster, J. J., Civin, C. I., et al. (1990) Prognostic significance of CD34
expression in childhood B-precursor acute lymphocytic leukemia: a Pediatric Oncology
Group study. J. Clin. Oncol. 8, 1389–1398.
87. Vogler, L. B., Crist, W. M., Bockman, D. E., Pearl, E. R., Lawton, A. R., and Cooper,
M. D. (1978) Pre-B-cell leukemia. A new phenotype of childhood lymphoblastic leukemia.
N. Engl. J. Med. 298, 872–878.
88. Kilo, M. N. and Dorfman, D. M. (1996) The utility of flow cytometric immunophenotypic
analysis in the distinction of small lymphocytic lymphoma/chronic lymphocytic leukemia
from mantle cell lymphoma. Am. J. Clin. Pathol. 105, 451–457.
89. Harris, N. L. and Bhan, A. K. (1985) B-cell neoplasms of the lymphocytic, lymphoplasma-
cytoid, and plasma cell types: immunohistologic analysis and clinical correlation. Hum.
Pathol. 16, 829–837.
316 Kurtin
90. Matutes, E. and Catavosky, D. (1994) The value of scoring systems for the diagnosis of
biphenotypic leukemia and mature B-cell disorders. Leuk. Lymphoma 13, 11–14.
91. Weisenburger, D., Harrington, J., and Armitage, J. (1990) B-cell neoplasia: a conceptual
understanding based on the normal humoral immune system. Pathol. Annu. 25, 99–
115.
92. Dorfman, D. M. and Pinkus, G. S. (1994) Distinction between small lymphocytic and
mantle cell lymphoma by immunoreactivity for CD23. Mod. Pathol. 7, 326–331.
93. Argatoff, L. H., Connors, J. M., Klasa, R. J., Horsman, D. E., and Gascoyne, R. D. (1997)
Mantle cell lymphoma: a clinicopathologic study of 80 cases. Blood 89, 2067–2078.
94. Pittaluga, S., Wlodarska, I., Stul, M. S., et al. (1995) Mantle cell lymphoma: a clinicopatho-
logical study of 55 cases. Histopathology 26, 17–24.
95. Dierlamm, J., Pittaluga, S., Wlodarksa, I., et al. (1996) Marginal zone B-cell lymphomas
of different sites share similar cytogenetic and morphologic features. Blood 87, 299–307.
96. Hammer, R. D., Glick, A. D., Greer, J. P., Collins, R. D., and Cousar, J. B. (1996)
Splenic marginal zone lymphoma. A distinct B-cell neoplasm. Am. J. Surg. Pathol.
20, 613–626.
97. Mollejo, M., Menarguez, J., Lloret, E., et al. (1995) Splenic marginal zone lymphoma: a
distinctive type of low-grade B-cell lymphoma. A clinicopathological study of 13 cases.
Am. J. Surg. Pathol. 19, 1146–1157.
98. Schmid, C., Kirkham, N., Diss, T., and Isaacson, P. G. (1992) Splenic marginal zone cell
lymphoma. Am. J. Surg. Pathol. 16, 455–466.
99. Savilo, E., Campo, E., Mollejo, M., et al. (1998) Absence of cyclin D1 protein expression
in splenic marginal zone lymphoma. Mod. Pathol. 11, 601–606.
100. de Wolf-Peeters, C., Pittaluga, S., Dierlamm, J., Wlodarska, I., and Van Den Berghe, H.
(1997) Marginal zone B-cell lymphomas including musosa-associated lymphoid tissue
type lymphoma (MALT), monocytoid B-cell lymphoma and splenic marginal zone cell
lymphoma and their relation to the reactive marginal zone. Leuk. Lymphoma 26, 467–478.
101. Almasri, N. M., Iturraspe, J. A., and Braylan, R. C. (1998) CD10 expression in follicular
lymphoma and large cell lymphoma is different from that of reactive lymph node follicles.
Arch. Pathol. Lab. Med. 122, 539–544.
102. Carey, J. and Hanson, C. (1994) Flow cytometric analysis of leukemia and lymphoma, in
Flow Cytometry and Clinical Diagnosis (Keren, D., Hanson, C., and Hurtubise, P., eds.)
American Society of Clinical Pathologists, Chicago IL, pp. 197–308.
103. Kumar, S., Green, G. A., Teruya-Feldstein, J., Raffeld, M., and Jaffe, E. S. (1996) Use
of CD23 (BU38) on paraffin sections in the diagnosis of small lymphocytic lymphoma
and mantle cell lymphoma. Mod. Pathol. 9, 925–929.
104. Singh, N. and Wright, D. H. (1997) The value of immunohistochemistry on paraffin wax
embedded tissue sections in the differentiation of small lymphocytic and mantle cell
lymphomas. J. Clin. Pathol. 50, 16–21.
105. Murray, P. G., Janmohamed, R. M., and Crocker, J. (1991) CD23 expression in non-
Hodgkin lymphoma: immunohistochemical demonstration using the antibody BU38 on
paraffin sections. J. Pathol. 165, 125–128.
106. Doggett, R. S., Wood, G. S., and Horning, S. (1984) The immunologic characterization
of 95 nodal and extranodal diffuse large cell lymphomas in 89 patients. Am. J. Pathol.
115, 245–252.
107. Pezzella, F., Munson, P. J., Miller, K. D., Goldstone, A. H., and Gatter, K. C. (2000) The
diagnosis of low-grade peripheral B-cell neoplasms in bone marrow trephines. Br. J.
Haematol. 108, 369–376.
108. McIntosh, G. G., Lodge, A. J., Watson, P., et al. (1999) NCL-CD10-270: a new monoclonal
antibody recognizing CD10 in paraffin-embedded tissue. Am. J. Pathol. 154, 77–82.
Diagnosis of Lymphoid Lesions 317
109. Kaufmann, O., Flath, B., Spath-Schwalbe, E., Possinger, K., and Dietel, M. (1999) Immuno-
histochemical detection of CD10 with monoclonal antibody 56C6 on paraffin sections.
Am. J. Clin. Pathol. 111, 117–122.
110. Pangalis, G. A., Angelopoulou, M. K., Vassilakopoulos, T. P., Siakantaris, M. P., and
Kittas, C. (1999) B-chronic lymphocytic leukemia, small lymphocytic lymphoma, and
lymphoplasmacytic lymphoma, including Waldenstrom’s macroglobulinemia: a clinical,
morphologic, and biologic spectrum of similar disorders. Semin. Hematol. 36, 104–114.
111. Feiner, H. D., Rizk, C. C., Finfer, M. D., et al. (1990) IgM monoclonal gammopathy/
Waldenstrom’s macroglobulinemia: a morphological and immunophenotypic study of the
bone marrow. Mod. Pathol. 3, 348–356.
112. San Miguel, J. F., Caballero, M. D., Gonzalez, M., Zola, H., and Lopez, Borrasca A.
(1986) Immunological phenotype of neoplasms involving the B cell in the last step of
differentiation. Br. J. Haematol. 62, 75–83.
113. Martinsson, U., Sundstrom, C., and Glimelius, B. (1989) Immunophenotype analysis of
B-CLL lymphoma and immunocytoma. APMIS 97, 1025–1032.
114. Shapiro, J. L., Miller, M. L., Pohlman, B., Mascha, E., and Fishleder, A. J. (1999) CD5-
B-cell lymphoproliferative disorders presenting in blood and bone marrow. A clinicopatho-
logic study of 40 patients. Am. J. Clin. Pathol. 111, 477–487.
115. Nakamine, H., Bagin, R. G., Vose, J. M., et al. (1993) Prognostic significance of clinical
and pathologic features in diffuse large B-cell lymphoma. Cancer 71, 3130–3137.
116. Dogan, A., Bagdi, E., Munson, P., and Isaacson, P. G. (2000) CD10 and BCL-6 expression
in paraffin sections of normal lymphoid tissue and B-cell lymphomas. Am. J. Surg. Pathol.
24, 846–852.
117. Harada, S., Suzuki, R., Uehira, K., et al. (1999) Molecular and immunological dissection
of diffuse large B cell lymphoma: CD5+, and CD5− with CD10+ groups may constitute
clinically relevant subtypes. Leukemia 13, 1441–1447.
118. Fang, J. M., Finn, W. G., Hussong, J. W., Goolsby, C. L., Cubbon, A. R., and Variakojis,
D. (1999) CD10 antigen expression correlates with the t(14;18)(q32;q21) major breakpoint
region in diffuse large B-cell lymphoma. Mod. Pathol. 12, 295–300.
119. Gascoyne, R. D., Adomat, S. A., Krajewski, S., et al. (1997) Prognostic significance of Bcl-
2 protein expression and Bcl-2 gene rearrangement in diffuse aggressive non-Hodgkin’s
lymphoma. Blood 90, 244–251.
120. Matolcsy, A., Chadburn, A., and Knowles, D. M. (1995) De novo CD5-positive and
Richter’s syndrome-associated diffuse large B cell lymphomas are genotypically distinct.
Am. J. Pathol. 147, 207–216.
121. Moller, P., Moldenhauer, G., Momburg, F., et al. (1987) Mediastinal lymphoma of clear
cell type is a tumor corresponding to terminal steps of B cell differentiation. Blood
69, 1087–1095.
122. Garcia, C. F., Weiss, L. M., and Warnke, R. A. (1986) Small noncleaved cell lymphoma:
an immunophenotypic study of 18 cases and comparison with large cell lymphoma. Hum.
Pathol. 17, 454–461.
123. Strickler, J. G., Audeh, M. W., Copenhaver, C. M., and Warnke, R. A. (1988) Immunophe-
notypic differences between plasmacytoma/multiple myeloma and immunoblastic lympho-
ma. Cancer 61, 1782–1786.
124. Feller, A. C., Griesser, H., Schilling, C. V., et al. (1988) Clonal gene rearrangement
patterns correlate with immunophenotype and clinical parameters in patients with angio-
immunoblastic lymphadenopathy. Am. J. Pathol. 133, 549–556.
125. Namikawa, R., Suchi, T., Ueda, R., et al. (1987) Phenotyping of proliferating lymphocytes
in angioimmunoblastic lymphadenopathy and related lesions by the double immunoenzy-
matic staining technique. Am. J. Pathol. 127, 279–287.
318 Kurtin
126. Kaneko, Y., Maseki, N., Sakurai, M., et al. (1988) Characteristic karyotypic pattern in T-
cell lymphoproliferative disorders with reactive angioimmunoblastic lymphadenopathy
with dysproteinemia features. Blood 72, 413–421.
127. Watanabe, S., Sato, Y., Shimoyama, M., Minato, K., and Shimosato, Y. (1986) Immu-
noblastic lymphadenopathy, angioimmunoblastic lymphadenopathy, and IBL-like T-cell
lymphoma. A spectrum of T-cell neoplasia. Cancer 58, 2224–2232.
128. Pallesen, G., Hamilton-Dutoit, S. J., and Zhou, X. (1993) The association of Epstein-
Barr virus (EBV) with T cell lymphoproliferations and Hodgkin’s disease: two new
developments in the EBV field. Adv. Cancer Res. 62, 179–239.
129. Leung, C. Y., Ho, F. C., Srivastava, G., Loke, S. L., Liu, Y. T., and Chan, A. C.
(1993) Usefulness of follicular dendritic cell pattern in classification of peripheral T-cell
lymphomas. Histopathology 23, 433–437.
130. Patsouris, E., Noel, H., and Lennert, K. (1989) Angioimmunoblastic lymphadenopathy—
type of T-cell lymphoma with a high content of epitheloid cells. Histopathology and
comparison with lymphoepithelioid cell lymphoma. Am. J. Surg. Pathol. 13, 262–275.
131. Salhany, K. E., Macon, W. R., Choi, J. K., et al. (1998) Subcutaneous panniculitis-like-
T-cell lymphoma: clinicopathologic, immunophenotypic, and genotypic analysis of alpha/
beta and gamma/delta subtypes. Am. J. Surg. Pathol. 22, 881–893.
132. Kumar, S., Krenacs, L., Medeiros, J., et al. (1998) Subcutaneous panniculitic T-cell
lymphoma is a tumor of cytotoxic T lymphocytes. Hum. Pathol. 29, 397–403.
133. Cooke, C. B., Krenacs, L., Stetler-Stevenson, M., et al. (1996) Hepatosplenic T-cell
lymphoma: a distinct clinicopathologic entity of cytotoxic gamma delta T-cell origin.
Blood 88, 4265–4274.
134. Bagdi, E., Diss, T. C., Munson, P., and Isaacson, P. G. (1999) Mucosal intra-epithelial
lymphocytes in enteropathy-associated T-cell lymphoma, ulcerative jejunitis, and refractory
celiac disease constitute a neoplastic population. Blood 94, 260–264.
135. Feller, A. C., Parwaresch, M. R., Stein, H., Ziegler, A., Herbst, H., and Lennert, K. (1986)
Immunophenotyping of T-lymphoblastic lymphoma/leukemia: correlation with normal T-
cell maturation. Leuk. Res. 10, 1025–1031.
136. Sheibani, K., Nathwani, B. N., Winberg, C. D., et al. (1987) Antigenically defined sub-
groups of lymphoblastic lymphoma. Relationship to clinical presentation and biologic
behavior. Cancer 60, 183–190.
137. Weiss, L. M., Bindl, J. M., Picozzi, V. J., Link, M. P., and Warnke, R. A. (1986)
Lymphoblastic lymphoma: an immunophenotype study of 26 cases with comparison to T
cell acute lymphoblastic leukemia. Blood 67, 474–478.
138. Delsol, G., Al Saati, T., Gatter, K. C., et al. (1988) Coexpression of epithelial membrane
antigen (EMA), Ki-1, and interleukin-2 receptor by anaplastic large cell lymphomas.
Diagnostic value in so-called malignant histiocytosis. Am. J. Pathol. 130, 59–70.
139. Chott, A., Kaserer, K., Augustin, I., et al. (1990) Ki-1-positive large cell lymphoma. A
clinicopathologic study of 41 cases. Am. J. Surg. Pathol. 14, 439–448.
140. Filippa, D. A., Ladanyi, M., Wollner, N., et al. (1996) CD30 (Ki-1)-positive malignant
lymphomas: clinical, immunophenotypic, histologic, and genetic characteristics and differ-
ences with Hodgkin’s disease. Blood 87, 2905–2917.
141. Krenacs, L., Wellmann, A., Sorbara, L., et al. (1997) Cytotoxic cell antigen expression
in anaplastic large cell lymphomas of T- and null-cell type and Hodgkin’s disease: evidence
for distinct cellular origin. Blood 89, 980–989.
142. Pileri, S., Bocchia, M., Baroni, C. D., et al. (1994) Anaplastic large cell lymphoma (CD30
+/Ki-1+): results of a prospective clinico-pathological study of 69 cases. Br. J. Haematol.
86, 513–523.
143. Pileri, S. A., Piccaluga, A., Poggi, S., et al. (1995) Anaplastic large cell lymphoma: update
of findings. Leuk. Lymphoma 18, 17–25.
Diagnosis of Lymphoid Lesions 319
144. Tilly, H., Gaulard, P., Lepage, E., et al. (1997) Primary anaplastic large-cell lymphoma
in adults: clinical presentation, immunophenotype, and outcome. Blood 90, 3727–3734.
145. Hall, P. A., D’Ardenne, A. J., and Stansfield, A. G. (1988) Paraffin section immunohisto-
chemistry. II. Hodgkin’s disease and large cell anaplastic (Ki-1) lymphoma. Histopathology
13, 161–169.
146. Hall, P. A. and D’Ardenne, A. J. (1987) Value of CD15 immunostaining in diagnosing
Hodgkin’s disease: a review of published literature. J. Clin. Pathol. 40, 1298–1304.
147. Hsu, S. M. and Jaffe, E. S. (1984) Leu M1 and peanut agglutinin stain the neoplastic cells
of Hodgkin’s disease. Am. J. Clin. Pathol. 82, 29–32.
148. Hsu, S. M., Yang, K., and Jaffe, E. S. (1985) Phenotypic expression of Hodgkin’s and
Reed-Sternberg cells in Hodgkin’s disease. Am. J. Pathol. 118, 209–217.
149. Schwarting, R., Gerdes, J., Dürkop, H., Falini, B., Pileri, S., and Stein, H. (1989) Ber-
H2: a new anti-Ki-1 (CD30) monoclonal antibody directed at a formol-resistant epitope.
Blood 74, 1678–1689.
150. Stein, H., Gerdes, J., Schwab, U., et al. (1982) Identification of Hodgkin and Sternberg-
Reed cells as a unique cell type derived from a newly-detected small-cell population. Int.
J. Cancer 30, 445–459.
151. Arber, D. A. and Weiss, L. M. (1993) CD15: a review. Appl. Immunohistochem. 1, 17–30.
152. Chittal, S. M., Caveriviere, P., Schwarting, R., et al. (1988) Monoclonal antibodies in the
diagnosis of Hodgkin’s disease. The search for a rational panel. Am. J. Surg. Pathol.
12, 9–21.
153. Pinkus, G. S., Pinkus, J. L., Langhoff, E., et al. (1997) Fascin, a sensitive new marker
for Reed-Sternberg cells of Hodgkin’s disease. Evidence for a dendritic or B cell derivation?
Am. J. Pathol. 150, 543–562.
154. Stein, H., Mason, D. Y., Gerdes, J., et al. (1985) The expression of the Hodgkin’s disease
associated antigen Ki-1 in reactive and neoplastic lymphoid tissue: evidence that Reed-
Sternberg cells and histiocytic malignancies are derived from activated lymphoid cells.
Blood 66, 848–858.
155. Nguyen, P. L., Harris, N. L., Ritz, J., and Robertson, M. J. (1996) Expression of CD95
antigen and bcl-2 protein in non-Hodgkin’s lymphomas and Hodgkin’s disease. Am. J.
Pathol. 148, 847–853.
156. Casey, T., Olson, S. J., Cousar, J. B., and Collins, R. D. (1989) Immunophenotypes of
Reed-Sternberg cells: a study of 19 cases of Hodgkin’s disease in plastic-embedded
sections. Blood 74, 2624–2628.
157. Falini, B., Stein, H., Pileri, S., et al. (1987) Expression of lymphoid-associated antigens on
Hodgkin’s and Reed-Sternberg cells of Hodgkin’s disease. Histopathology 11, 1129–1242.
158. Kadin, M. E., Muramoto, L., and Said, J. (1988) Expression of T-cell antigens on Reed-
Sternberg cells in a subset of patients with nodular sclerosing and mixed cellularity
Hodgkin’s disease. Am. J. Pathol. 130, 345–353.
159. Korkolopoulou, P., Cordell, J., Jones, M., et al. (1994) The expression of the B-cell marker
Mb-1 (CD79a) in Hodgkin’s disease. Histopathology 24, 511–515.
160. Kuzu, I., Delsol, G., Jones, M., Gatter, K. C., and Mason, D. Y. (1993) Expression of
the Ig associated heterodimer (Mb-1 and B29) in Hodgkin’s disease. Histopathology
22, 141–144.
161. Schmid, C., Pan, L., Diss, T., and Isaacson, P. G. (1991) Expression of B-cell antigens
by Hodgkin’s and Reed-Sternberg cells. Am. J. Pathol. 139, 701–707.
162. Abdulaziz, Z., Mason, D. Y., Stein, H., Gatter, K. C., and Nash, J. R. G. (1984) An
immunohistochemical study of the cellular constituents of Hodgkin’s disease using an
monoclonal antibody panel. Histopathology 8, 1–25.
163. Agnarsson, B. A. and Kadin, M. E. (1989) The immunophenotype of Reed-Sternberg cells.
A study of 50 cases of Hodgkin’s disease using fixed frozen tissues. Cancer 63, 2083–2087.
320 Kurtin
164. Angel, C. A., Warford, A., Campbell, A. C., Pringle, J. H., and Lauder, I. (1987) The
immunohistology of Hodgkin’s disease—Reed-Sternberg cells and their variants. J. Pathol.
153, 21–30.
165. Cibull, M. L., Stein, H., Gatter, K. C., and Mason, D. Y. (1989) The expression of the
CD3 antigen in Hodgkin’s disease. Histopathology 15, 597–605.
166. Bishop, P. W., Harris, M., Smith, A. P., and Elsam, K. J. (1991) Immunophenotypic study
of lymphocyte predominance Hodgkin’s disease. Histopathology 18, 19–24.
167. Coles, R. B., Cartun, R. W., and Pastuszak, W. T. (1988) Hodgkin’s disease, lymphocyte-
predominant type: immunoreactivity with B-cell antibodies. Mod. Pathol. 1, 274–278.
168. Hansmann, M. L., Wacker, H. H., and Radzun, H. J. (1986) Paragranuloma is a variant
of Hodgkin’s disease with predominance of B-cells. Virchows Arch. Pathol. Anat. 409, 171–
181.
169. Nicholas, D. S., Harris, S., and Wright, D. H. (1990) Lymphocyte predominance Hodgkin’s
disease—an immunohistochemical study. Histopathology 16, 157–165.
170. Pinkus, G. S. and Said, J. W. (1985) Hodgkin’s disease, lymphocyte predominance type,
nodular—a distinct entity? Unique staining profile for L & H variants of Reed-Sternberg
cells defined by monoclonal antibodies to leukocyte common antigen, granulocyte-specific
antigen, and B-cell specific antigen. Am. J. Pathol. 133, 211–217.
171. Pinkus, G. S. and Said, J. W. (1988) Hodgkin’s disease, lymphocyte predominance type,
nodular—further evidence for a B cell derivation. L & H variants of Reed-Sternberg cells
express L26, a pan B cell marker. Am. J. Pathol. 133, 211–217.
172. Poppema, S. (1992) Lymphocyte-predominance Hodgkin’s disease. Int. Rev. Exp. Pathol.
33, 53–79.
173. Poppema, S., Timens, W., and Visser, L. (1985) Nodular lymphocyte predominance type
of Hodgkin’s disease is a B cell lymphoma. Adv. Exp. Med. Biol. 186, 963–969.
174. Poppema, S. (1980) The diversity of the immunohistological staining pattern of Sternberg-
Reed cells. J. Histochem. Cytochem. 28, 788–791.
175. Stein, H., Hansmann, M. L., Lennert, K., Brandtzaeg, P., Gatter, K. C., and Mason, D. Y.
(1986) Reed-Sternberg and Hodgkin cells in lymphocyte-predominant Hodgkin’s disease
of nodular subtype contain J chain. Am. J. Clin. Pathol. 86, 292–297.
176. Falini, B., Bigerna, B., Pasqualucci, L., et al. (1996) Distinctive expression pattern of the
BCL-6 protein in nodular lymphocyte predominance Hodgkin’s disease. Blood 87, 465–
471.
177. Hansmann, M. L., Fellbaum, C., Hui, P. K., and Zwingers, T. (1988) Correlation of content
of B cells and Leu7-positive cells with subtype and stage in lymphocyte predominance
type Hodgkin’s disease. J. Cancer Res. Clin. Oncol. 114, 405–410.
178. Poppema, S. (1989) The nature of the lymphocytes surrounding Reed-Sternberg cells in
nodular lymphocyte predominance and in other types of Hodgkin’s disease. Am. J. Pathol.
135, 351–357.
179. Bene, M. C., Castoldi, G., Knapp, W., et al. (1995) Proposals for the immunological
classification of acute leukemias. European Group for the Immunological Characterization
of Leukemias (EGIL). Leukemia 9, 1783–1786.
180. Bradstock, K. F. (1993) The diagnostic and prognostic value of immunophenotyping in
acute leukemia. Pathology 25, 367–374.
181. Brunning, R. and McKenna, R. (1994) Tumors of the bone marrow, in Atlas of Tumor
Pathology, vol. 9 (Rosai, J., ed.) Armed Forces Institute of Pathology, Washington, DC.
182. Hanson, C. A., Gajl-Peczalska, K. J., Parkin, J. L., and Brunning, R. D. (1987) Immunophe-
notyping of acute myeloid leukemia using monoclonal antibodies and the alkaline phospha-
tase-antialkaline phosphatase technique. Blood 70, 83–89.
183. Khalidi, H. S., Medeiros, L. J., Chang, K. L., Brynes, R. K., Slovak, M. L., and Arber,
D. A. (1998) The immunophenotype of adult acute myeloid leukemia: high frequency of
Diagnosis of Lymphoid Lesions 321
202. Schriever, F., Freedman, A. S., Freeman, G., et al. (1989) Isolated human follicular
dendritic cells display a unique antigenic phenotype. J. Exp. Med. 169, 2043–2058.
203. Weiss, L. M., Berry, G. J., Dorfman, R. F., et al. (1990) Spindle cell neoplasms of lymph
nodes of probable reticulum cell lineage. True reticulum cell sarcoma? Am. J. Surg. Pathol.
14, 405–414.
204. Feltkamp, C., van Heerde, P., Feltkamp-Vroom, T., and Koudstaal, J. (1981) A malignant
tumor arising from interdigitating cells; light microscopical, ultrastructural, immuno- and
enzyme-histochemical characteristics. Virchows Arch. Pathol. Anat. 393, 183–192.
205. Nakamura, S., Hara, K., Suchi, T., et al. (1988) Interdigitating cell sarcoma. A morphologic,
immunohistologic, and enzyme-histochemical study. Cancer 61, 562–568.
206. Wilson, M. S., Weiss, L. M., Gatter, K. C., Mason, D. Y., Dorfman, R. F., and Warnke,
R. A. (1990) Malignant histiocytosis. A reassessment of cases previously reported in 1975
based on paraffin section immunophenotyping studies. Cancer 66, 530–536.
207. Arai, E., Su, W. P., Roche, P. C., and Li, C. Y. (1993) Cutaneous histiocytic malignancy.
Immunohistochemical re-examination of cases previously diagnosed as cutaneous histio-
cytic lymphoma; malignant histiocytosis. J. Cutan. Pathol. 20, 115–120.
208. Kamel, O. W., Gocke, C. D., Kell, D. L., Cleary, M. L., and Warnke, R. A. (1995) True
histiocytic lymphoma: a study of 12 cases based on current definition. Leuk. Lymphoma
18, 81–86.
209. Lauritzen, A. F. and Ralfkaier, E. (1995) Histiocytic sarcomas. Leuk. Lymphoma 18, 73–80.
210. Hsu, S. M., Ho, Y. S., and Hsu, P. L. (1991) Lymphomas of true histiocytic origin.
Expression of different phenotypes in so-called true histiocytic lymphoma and malignant
histiocytosis. Am. J. Pathol. 138, 1389–1404.
17
Applications of Immunohistochemistry in the
Diagnosis of Undifferentiated Tumors
INTRODUCTION
Not uncommonly, pathologists are confronted with neoplasms that have few distin-
guishing microscopic characteristics and, therefore, appear to be “undifferentiated.”
The differential diagnostic considerations in such cases are many, and special studies
are almost always necessary to reach a definite conclusion. Determining the probable
site of origin for a metastatic carcinoma and making a diagnostic separation between
two differentiated but histologically similar neoplasms represent additional challenges
in surgical pathology that require ancillary laboratory analyses. This presentation will
address approaches to these and other dilemmas that are particularly suited to resolution
by immunodiagnosis. The following discussion is not intended to be exhaustive or all-
inclusive, rather, the authors’ aim is to provide a framework for the contemporary use
of immunohistochemistry in the interpretation of solid tumors.
323
324 Wick and Cerilli
Fig. 1. Spindle and pleomorphic cells showing strong keratin labeling that facilitates recogni-
tion of sarcomatoid carcinomas. Evaluation of this case disclosed a corresponding renal cell
carcinoma with sarcomatoid features.
availability of “epitope retrieval” procedures has done much to lessen the impact that
formalin has on the innumoreactivity of clinical specimens (3).
These new techniques, and currently available antibody-enzyme bridge methods,
have been reviewed extensively in other reference sources and will not be recounted here
(3–6). Nonetheless, it is worthwhile reiterating that select procedures lend themselves to
the detection of some antigens and not others (7,8). Moreover, tissues with a high
endogenous level of peroxidase (e.g., bone marrow sections) may require a different
chromogenic enzyme (alkaline phosphatase, glucose oxidase) to obtain high-quality
results (9,10). Each laboratory is encouraged to make such determinations individually,
since nuances in techniques at different institutions are well known to affect the outcome.
Immunohistochemical Reagents
Intermediate Filaments
Cytokeratins are constituents of the intermediate filaments (IFs) of epithelial cells
expressed in various combinations depending on the epithelial type and the degree of
differentiation. This class of IFs remains among the most commonly studied determi-
nants in immunohistochemistry. Cytokeratin positivity helps corroborate a diagnosis
of sarcomatoid carcinoma (Fig. 1) and usually excludes the possibility of sarcoma,
malignant lymphoma, or melanoma (16,17). Notable exceptions include synovial sar-
coma (Fig. 2), chordoma (Fig. 3), Ewing’s sarcoma, and epithelioid sarcoma; some
smooth muscle cells may also react.
Undifferentiated and Morphologically Ambiguous Neoplasms 325
Fig. 2. Monophasic spindle cell synovial sarcoma shows unequivocal staining for keratin.
Monoclonal antibodies are now available to a wide range of keratin proteins (40–67
kDa) (11). To maximize cytokeratin detection, proteolysis (0.4% ficin in phosphate-
buffered saline [pH 7.4] for 20 min at room temperature [12]) or microwave-mediated
epitope retrieval in citrate buffer (7,8) is mandatory before application of primary
antibodies to rehydrated paraffin sections. Since, in most cases, the question is whether
or not any cytokeratin is present in a given neoplasm, combinations or “cocktails” of
monoclonal antibodies may be prepared to evaluate the widest range of kilodalton
326 Wick and Cerilli
weights. Keratin “cocktails” are the most useful in the diagnosis of poorly differentiated
epithelial tumors (13), but monospecific keratin antibodies also have distinct advantages
in selected circumstances. For example, Merkel cell carcinoma of the skin—an example
of a small round cell undifferentiated malignancy—regularly expresses keratin 20 (Fig.
4), whereas its differential diagnostic simulators generally do not (14). To prepare
keratin antibody “cocktails,” each of several monoclonal anticytokeratins is added to
the same tube of diluent, in a quantity yielding the optimal titer of that particular
antibody. For example, if two antibodies are selected to prepare 4 mL of “cocktail,”
one with an optimal titer of 1:200 and the other 1:400, 20 µL of the first antibody and
10 µL of the second would be added to the same 4 mL of solution.
Of course, one must be assured that the anticytokeratin preparations being utilized
are specific. This point is particularly important with respect to the immunodetection
of IFs as a group (cytokeratin, vimentin, desmin, glial fibrillary acidic protein [GFAP],
and neurofilament protein), since all share a common peptide sequence (15).
Vimentin is an IF that is present in most mesenchymal neoplasms and a variety of
other classes of neoplasms (e.g., endometrial carcinoma) (16–18). It is highly valuable
as a marker of adequate tissue fixation and processing, in which positive and appropriate
vimentin staining provides certainty of the ability of the tissue to react with antibodies
in general (19). Desmin is another IF present in mesenchymal lesions, found principally
in myogenous tumors such as rhabdomyosarcoma and leiomyosarcoma (20) (Fig. 5).
The neural IFs include neurofilament protein and GFAP. The first of these is restricted
to neuronal and neuroendocrine cellular proliferations (21); however, it is suboptimally
Undifferentiated and Morphologically Ambiguous Neoplasms 327
Fig. 6. Soft tissue rhabdoid tumor shows common coexpression of vimentin and epithelial
markers. This tumor shows membranous straining for epithelial membrane antigen (EMA).
of keratin, with some caveats; in particular, not all epithelia are positive—hepatocellular
carcinomas, adrenocortical carcinomas, and malignant germ cell neoplasms are EMA
negative (26,29). Also, EMA may be seen outside of nonepithelial lesions, including
large cell anaplastic lymphoma, plasmacytoma (Fig. 7), some T-cell lymphomas (30,31),
epithelioid sarcoma, and synovial sarcoma; meningioma may also show EMA reactivity
(32). The use of supplementary antibody panels that include other epithelial markers
obviates potential misclassification.
MOC-31
MOC-31 is a 41-kDa glycoprotein that is cell membrane based; it is widely distributed
in epithelial cells and tumors in many tissue sites (33,34). Monoclonal antibodies to
this determinant have most often been used in the diagnostic distinction between serosal
adenocarcinoma and mesothelioma (which typically lacks MOC-31) (35), but they also
fail to label a selected group of carcinomas that includes hepatocellular carcinoma,
germ cell malignancies, and renal cell carcinoma (36).
Tumor-Associated Glycoprotein (B72.3)
The monoclonal antibody known as B72.3 labels a plasmalemmal glycoprotein
designated tumor-associated glycoprotein-72 (TAG-72) (37). It was isolated from a
human breast carcinoma cell line and appears to be virtually pan-carcinomatous in
distribution (37,38). In likeness to MOC-31, however, TAG-72 is characteristically
absent in mesotheliomas, adrenocortical carcinomas, hepatocellular carcinomas, renal
Undifferentiated and Morphologically Ambiguous Neoplasms 329
Fig. 7. Plasmacytomas frequently show epithelial membrane antigen (EMA; shown) and
vimentin positivity, which may lead to misclassification as a carcinoma or sarcoma.
Fig. 9. CD30 is frequently positive in embryonal carcinoma (shown) and absent in seminoma,
a helpful distinction in keratin-positive germ cell tumors lacking specific features on routine his-
tology.
Undifferentiated and Morphologically Ambiguous Neoplasms 331
α-Fetoprotein
AFP is another oncofetal antigen that is relatively restricted in its expression to
nonseminomatous germ cell tumors (Fig. 10), hepatocellular carcinomas, and rare
examples of somatic malignancies (42–44). Hence, it has its greatest use in immunohisto-
chemistry in the characterization of polygonal large cell undifferentiated neoplasms.
In an AFP-positive polygonal cell tumor, PLAP positivity is desirable before favoring
the diagnosis of germ cell tumor over hepatocellular carcinoma, which is devoid of
alkaline phosphatase immunoreactivity (29).
Carcinoembryonic Antigen
Carcinoembryonic antigen (CEA) has enjoyed its greatest recognition in clinical
medicine as a serologic indicator of the growth of colorectal cancer. Immunohistochemi-
cally, CEA is strongly expressed in colorectal adenocarcinoma (Fig. 11), but it may
be found in other epithelial tumors (45–47). Monoclonal antibodies to CEA represent
prototypic epitope-specific probes, which recognize a small portion of a large antigen
(45,47). Different carcinomas express a common portion of the CEA molecule but may
also produce mutually exclusive epitopes that are tissue restricted. If the latter are
“mapped” by immunohistochemical surveys of primary carcinomas, the corresponding
antibody reagents can be employed to yield diagnostic information on the possible
sources of lesions that present with metastasis. For example, adenocarcinomas of the
lung, breast, and gastrointestinal (GI) tract show uniform reactivity with certain epitope-
specific anti-CEAs and thereby may narrow the potential anatomic sites for metastatic
glandular neoplasms (45–47). In general, monoclonal anti-CEA preparations are prefera-
ble to polyclonal heteroantisera, because most of the latter are inferior in specificity
and require adsorption with crossreacting antigens.
332 Wick and Cerilli
Fig. 11. Adenocarcinoma of the colon shows uniform strong carcinoembryonic antigen
(CEA) expression.
Prostate-Specific Antigen
As its name suggests, prostate-specific antigen (PSA) is largely restricted in its tissue
distribution to epithelial cells of the prostate and prostatic adenocarcinomas (Fig. 12)
(48–50). This cytoplasmic determinant appears to be expressed at an early stage of
embryonic development, inasmuch as even poorly differentiated prostatic malignancies
Undifferentiated and Morphologically Ambiguous Neoplasms 333
Fig. 13. The presence of thyroglobulin reactivity in a metastatic tumor showing a nested
pattern helps establish the diagnosis of poorly differentiated thyroid carcinoma.
display its presence. Hence, PSA represents one of exceedingly few tissue-specific
markers available in diagnostic immunohistochemistry and has been widely utilized in
the recognition of metastatic prostate cancer.
Some reports have impugned the exclusivity of this association and have documented
apparent PSA reactivity in selected adenocarcinomas of the bladder (51) or in neuroendo-
crine tumors (52). In our experience, however, such problems are likely to be related
to the heteroantiserum reagents that were used in these studies. We have employed a
monoclonal anti-PSA in analyses of over 3000 carcinomas from diverse locations; to
date, it has demonstrated absolute specificity for prostatic neoplasms in these cases.
However, extraprostatic expression of PSA has also been noted in periurethral gland
adenocarcinoma in women, rectal carcinoid, and extramammary Paget’s disease (52a).
These entities can generally be easily addressed using a panel of immunostains that
includes PSA, as well as neuroendocrine markers and CEA.
Thyroglobulin
Thyroglobulin is likewise restricted in its expression to the follicular thyroid
epithelium and related neoplasms (53,54). The greatest utility of this marker is to
confirm a thyroid origin for metastatic carcinomas in lymph nodes, lung, or bone.
Thyroglobulin reactivity is also helpful in the diagnosis of struma ovarii. Unfortu-
nately, “anaplastic” thyroid cancer seldom demonstrates this determinant immunohis-
tochemically (55), limiting its usefulness in the evaluation of undifferentiated tumors
to “solid” follicular carcinomas, including “insular” thyroid carcinoma and insular
carcinoma (Fig. 13) (56).
334 Wick and Cerilli
Fig. 14. The cells of metastatic lobular carcinoma are readily detected in a lymph node using
an antibody to GCDFP-15.
Fig. 15. This metastatic adenocarcinoma shows strong GCDFP-15 reactivity. Although typi-
cally regarded as a marker for breast cancer, salivary duct carcinoma (SDC) also shows strong
reactivity. SDC lacks estrogen receptor and progesterone receptor staining, which may be useful
in differential diagnosis.
336 Wick and Cerilli
Fig. 16. Adenocarcinoma of the ovary frequently shows strong reactivity for CA-125.
Today it is recognized as the marker most strongly associated with epithelial gynecologic
tumors (Fig. 16). CA-125 has been detected immunohistochemically, most often in
neoplasms of the müllerian tract, but roughly one-half of tumors of the biliary tree and
pancreas may also react (77,78). Carcinomas in other sites are uncommonly CA-125
positive. In diagnostic pathology CA-125 plays a role in identifying the primary locations
of metastatic carcinoma of unknown origin. The use of CA-125 antibodies is recom-
mended not in a solitary setting but in combination with CEA, GCDF5-15, and vimentin
to discriminate among the sites of origin of metastatic carcinoma.
CA19-9
CA19-9 is a glycoprotein that is related to the Lea blood group antigen, and it is
labeled by the monoclonal antibody known as 1116NS19-9 (79). The latter reagent
was raised against a human colon carcinoma cell line. Most carcinomas of the GI tract,
pancreas, biliary tree, urinary bladder, ovaries, and endometrium manifest CA19-9
reactivity (80) (Fig. 17), whereas tumors of other anatomic locations are only sporadi-
cally positive. In particular, hepatocellular carcinoma and renal cell carcinoma are
consistently negative for this marker (71,80).
CD15
CD15 is a hematopoietic differentiation antigen shared by granulocytes, monocytes,
Reed-Stemberg cells, and subsets of neoplastic B-cells and T-lymphocytes (81,82).
Undifferentiated and Morphologically Ambiguous Neoplasms 337
Fig. 17. CA19-9 highlights the malignant cells of pancreatic adenocarcinoma. This marker
may also be positive in tumors of the gastrointestinal tract, biliary tree, urinary bladder, ovaries,
and endometrium.
Several epithelial cell lines (most of which are glandular or neuroendocrine) also
expresses CD15 (81), and it may be detected immunohistochemically in their malignant
counterparts (Fig. 18). Reactivity for CD15 is applicable to the separation of metastatic
adenocarcinomas from histologically similar lesions such as malignant mesotheliomas,
since it is lacking in the latter tumors (83). When used in combination with leukocyte
common antigen (CD45), CD15 also provides information on the cell lineage of lympho-
reticular neoplasms; CD15-reactive tumors in this class predominantly include
Hodgkin’s disease (Fig. 19) and T-cell non-Hodgkin’s lymphomas (82).
Calretinin
Calretinin is a calcium-binding protein that is virtually universally expressed by
mesothelial cells and malignant mesotheliomas (84). Calretinin is a recognized useful
marker for discriminating mesothelioma of the epithelial type from adenocarcinoma
in both paraffin sections and serous effusions (85,85a–c). A limited number of cases
of sarcomatoid mesothelioma may also show calretinin expression, separating these
lesions from a spectrum of other spindle cell neoplasms of the pleura (85d).
S-100 Protein
S-100 protein is a calcium flux determinant, expressed by normal melanocytes,
Langerhans’ histiocytes, cartilaginous cells, adipocytes, Schwann cells, astrocytes, oli-
godendroglia, ependyma, eccrine sweat glands, reticulum cells, salivary glands, and
myoepithelial cells (86–89). It has its greatest use in the diagnosis of solid tumors in
338 Wick and Cerilli
Fig. 18. Positive CD15 (Leu-M1) staining is diagnostically useful to distinguish carcinoma
from epithelial mesothelioma. It is not useful, however, in sarcomatoid tumors.
Fig. 19. CD15 (Leu-M1) highlights the Reed-Sternberg cells in Hodgkin’s disease.
Undifferentiated and Morphologically Ambiguous Neoplasms 339
Fig. 20. The presence of S-100 protein reactivity is useful to corroborate a diagnosis of
metastatic adenocarcinoma of the breast.
Fig. 21. The presence of HMB-45 reactivity in the spindle cell variant of malignant melanoma
is rare but helps distinguish this tumor from other spindle cell tumors.
other hand, MART-1 reactivity is seen in most adrenocortical carcinomas, and KBA62
may decorate occasional poorly differentiated squamous and renal cell carcinomas as
well as rare rhabdomyosarcomas (92). However, the overall utility of those two antibod-
ies as melanocytic markers is still high.
Chromogranin A
Chromogranin A (CgA) is a protein that is indigenous to the matrices of neurosecre-
tory granules. It has a relatively ubiquitous distribution in neuroendocrine tissues,
e.g., those of the anterior pituitary gland, thyroid C-cell system, parathyroid glands,
paraganglion system, adrenal medulla, and pancreatic islets (93,94). Accordingly, anti-
bodies to CgA are exceedingly specific in the delineation of neuroendocrine differentia-
tion in epithelial tumors. However, because CgA reactivity is directly related to the
relative number of cytoplasmic endocrine granules in any given neoplasm, this marker
is somewhat insensitive. For example, roughly 30% of small cell neuroendocrine carci-
nomas and neuroblastomas will exhibit labeling with anti-CgA, and it is distinctly
unusual for primitive neuroectodermal tumors to do so (95).
Synaptophysin
Synaptophysin is an integral membrane glycoprotein of presynaptic vesicles that is
detectable in a wide range of normal and neoplastic neuroendocrine cells (Fig. 22)
(95–99). Monoclonal antibodies to this marker are widely used in diagnostic surgical
pathology and cytopathology with good success. In light of its subcellular associations,
Undifferentiated and Morphologically Ambiguous Neoplasms 341
Fig. 22. Desmoplastic small cell tumors show a diverse array of differentiation, which often
includes strong synaptophysin reactivity.
one might assume that synaptophysin would have a tissue distribution synonymous to
that of the chromogranins; however, in practicality, that is not true (99). The vesicles
that contain synaptophysin can colocalize with neurofilaments or epithelial filaments,
showing expression independent of other neuronal differentiation markers. A sizable
proportion of neuroendocrine neoplasms will label for CGA, but not synaptophysin,
and the converse of that relationship also applies. Thus, anti-synaptophysin and anti-
chromogranin should be conceptualized as complementary reagents.
CD57
CD57 is recognized by the monoclonal antibody HNK-1 and was initially described
as a membrane antigen of hematopoietic cells, namely, natural killer lymphocytes. (100).
Subsequent analyses have demonstrated a relative rarity of CD57-positive malignant
lymphomas (101), relegating this marker to a minor role in hematopathology; it is most
often seen in “large granular cell” lymphoproliferative disorders, and in the Reed-
Sternberg-like cells of lymphocyte-predominant Hodgkin’s disease (which is probably,
in actuality, a B-cell lymphoma) (101). Unexpectedly, HNK-1 was shown to recognize
an epitope of myelin-associated glycoprotein and a glycoproteinaceous determinant in
neurosecretory granule matrices (100,102,103). Hence, it is most often used today
in solid tumor pathology, to label schwannian, neuroectodermal, and neuroendocrine
proliferations. Because of the variety of cell lineages involved in this group, it should
be apparent that results of other immunostains are required to put the significance of
CD57 reactivity into proper perspective, in any given case.
342 Wick and Cerilli
Fig. 23. Among small round cell tumors, CD99 shows a characteristic membranous pattern
in acute lymphoblastic lymphoma (shown) as well as Ewing’s sarcoma.
CD34
CD34, or the human hematopoietic progenitor cell antigen, is recognized by several
monoclonal antibodies including My10, QBEND10, and BI-3C5 (108). It is a 110 kDa
protein that, as its name suggests, is expressed by embryonic cells of the hematopoietic
system. These include lymphoid and myelogenous elements and also endothelial cells
(109). Correspondingly, again in the setting of soft tissue tumors, CD34 is a potential
indicator of vascular differentiation. It is highly sensitive for endothelial differentiation,
regardless of tumor grade, and recognizes >85% of angiosarcomas and Kaposi’s sarco-
mas (108,109). Nevertheless, the specificity of CD34 is a problem, inasmuchas it has
been reported in some leiomyosarcomas, peripheral nerve sheath tumors, and epithelioid
sarcomas, which could potentially simulate variants of angiosarcoma or hemangioendo-
thelioma (109). In addition, CD34 is so commonly present in dermatofibrosarcoma
protuberans (and its variants), spindle cell lipoma, and solitary fibrous tumor that it is
regularly used as an adjunct for the diagnosis of those tumors (110,111). Thus, as
endothelial markers, antibodies to CD34 are best used in a panel of reagents that is
designed to account for these other diagnostic possibilities.
CD31
The platelet-endothelial cell adhesion molecule-1 (PECAM-1) is also known as
CD31. It is a 130-kDa transmembrane glycoprotein shared by vascular lining cells,
megakaryocytes, platelets, and selected other hematopoietic elements and is recognized
by the monoclonal antibody JC/70A (108,111). This marker is highly restricted to
endothelial neoplasms among all tumors of the soft tissue, and its sensitivity is also
excellent. In our hands, virtually 100% of angiosarcomas are CD31 positive, regardless
of grade or histotype (112) (Fig. 17), and the same statement applies to hemangioma
and hemangioendothelioma variants. It must be acknowledged, however, that Kaposi’s
sarcoma is labeled more consistently for CD34 than CD31, for unknown reasons.
Thrombomodulin
Thrombomodulin (TMN) is a 75-kDa membrano-cytoplasmic glycoprotein that is dis-
tributed among endothelial cells, mesothelial cells, osteoblasts, mononuclear phagocytic
cells, and selected epithelia (113). Its physiologic role is to convert thrombin from a
coagulant protein to an anticoagulant. TMN may be present in some metastatic carcinomas
and most mesotheliomas (114), both of which may be confused with epithelioid angio-
sarcomas; therefore, it should be interpreted with this caveat in mind and in the context
of the results of other vascular markers. Nevertheless, TMN has proved to be a highly
sensitive indicator of endothelial differentiation, particularly in poorly differentiated vas-
cular malignancies. Kaposi’s sarcoma is likewise consistently immunoreactive for this
determinant (115). Thus, its inclusion in antibody panels is certainly worthwhile.
Ulex europaeus I Agglutinin
Ulex europaeus I (UEAI) agglutinin is not an antibody reagent but instead represents a
lectin produced by the gorse plant. It recognizes the Fuc-α-1-2-Gal linkage in fucosylated
oligosaccharides, which comprise portions of various glycoproteins (116). In particular,
the H blood group antigen and carcinoembryonic antigen regularly bind to UEAI, as
does a separate fucosylated protein that is expressed by endothelial cells. Biotinylated
344 Wick and Cerilli
Fig. 24. Algorithmic approach to the immunohistochemical diagnosis of small cell tumors.
in a uniform manner, whereas other neoplasms in this category are devoid of reactivity
(23), with extremely rare exceptions (117a). Neuroblastomas and neuroepitheliomas
express synaptophysin and CD57 (118). Rhabdomyosarcomas also share potential reac-
tivity for CD57, but, excluding rare examples of PNET with divergent differentiation
(119), primitive striated muscle sarcomas are distinguished from small cell neurogenic
neoplasms by their desmin or actin positivity, as mentioned above.
A particularly diagnostically challenging member of this tumor group is, in fact,
Ewing’s sarcoma/PNET. Our approach is to make this unqualified diagnosis when a
small cell neoplasm of soft tissue is reactive for CD99, with or without vimentin,
synaptophysin, and CD57, and in the absence of desmin, actin, S100 protein, and CD45.
Some PNETs may also synthesize keratin in an “aberrant” fashion (118).
Small Cell Carcinomas
Small cell carcinomas may exhibit neuroendocrine, squamous, or glandular differenti-
ation. However, a determination of cell lineage is often difficult on conventional micros-
copy of these lesions, prompting the use of discriminating immunostains. CD57, synap-
tophysin, and CGA are restricted to neuroendocrine neoplasms in this morphologic
class of lesions; nevertheless, it should be reemphasized that they are seen in only
30–50% of cases (120). Thus, most of the time, immunohistologic studies will not
successfully corroborate a neuroendocrine lineage for small cell carcinoma, even though
it is really present, and one could seriously question whether or not such analyses are
cost effective. On the other hand, small cell adenocarcinomas usually dislay MOC-31
or TAG-72, whereas small cell squamous tumors do not. All three forms of small cell
carcinoma express cytokeratin reactivity, separating them from most small cell sarcomas
of soft tissue. This fact is more important than one would assume at first glance, because
Eusebi et al. (120a) have described small cell carcinomas that were aberrantly labeled
for myogenic markers as well as keratin. One should also keep in mind the potential
overlap between PNET and small cell neuroendocrine carcinoma, in regard to the
possibility of conjoint staining for CD99 and keratin in those neoplasms.
Spindle Cell Neoplasms
Spindle cell (“sarcomatoid”) carcinomas have been described in a diversity of organs,
including the skin, aerodigestive tract, kidney, bladder, female genital tract, lung, and
other sites, and must be distinguished from melanomas and true sarcomas in such
locations (121) (Fig. 25). In addition, spindle cell mesenchymal malignancies of soft
tissue can be difficult to separate from one another diagnostically.
Spindle Cell Carcinomas
In mucosal or organ-based locations, any spindle cell tumor that displays reactivity
for cytokeratin or EMA can be defined as carcinomatous (121). Most sarcomatoid
carcinomas also coexpress vimentin, but the latter protein should be regarded as nonspe-
cific when seen in conjunction with other intermediate filaments.
Spindle Cell Amelanotic Malignant Melanomas
Melanomas that are composed exclusively of fusiform cells retain the general immu-
nohistochemical attributes of other malignant melanocytic neoplasms, with one excep-
tion. For unknown reasons, spindle cell melanomas react poorly or not at all with
346 Wick and Cerilli
HMB-45, KBA62, MART-1, and antityrosinase, making the conjoint use of antibodies
to S100, vimentin, EMA, and cytokeratin paramount in this diagnostic context (92,122).
Positivity for either of the last two of these determinants excludes the possibility of
malignant melanoma.
Spindle Cell Sarcomas of Soft Tissue
The malignant spindle cell neoplasms of soft tissue include fibrosarcoma, leiomyo-
sarcoma, malignant peripheral nerve sheath tumor, monophasic synovial sarcoma,
malignant fibrous histiocytoma (MFH), and Kaposi’s sarcoma (18,111). Among these,
fibrosarcoma is characterized by its reactivity for vimentin, to the exclusion of desmin,
actin, S-100 protein, CD57, cytokeratin, and EMA. In addition, it lacks endothelial
markers. Leiomyosarcomas are unique in their diffuse reactivity for desmin and muscle-
specific actin but may also express S-100 protein and CD57. Malignant peripheral
nerve sheath tumors generally demonstrate only the last two of these four determinants.
Synovial sarcoma represents the sole lesion in this category that is capable of cytokeratin
and EMA synthesis; it may also express CD57 and CD99 (18). The immunohistologic
features of Kaposi’s sarcomas are controversial. Some investigators have obtained
consistent labeling of such neoplasms with UEAI, CD34, or CD31 (108), whereas
others (117) (including ourselves) have observed this finding to be less than universal.
Fortunately, conventional histologic analysis and clinical correlation usually allow for
confident identification of this lesion.
Undifferentiated and Morphologically Ambiguous Neoplasms 347
It is distinctive in its immunoreactivity for vimentin and CD117 (C-kit protein). with
or without CD34, desmin, muscle-specific actin, or S-100 protein (130).
Pleomorphic Neoplasms
Tumors that are capable of assuming extremely bizarre, pleomorphic cellular shapes
include MFH, pleomorphic liposarcoma, pleomorphic rhabdomyosarcoma, “dedifferen-
tiated” leiomyosarcoma, malignant peripheral nerve sheath tumor, and metastatic carci-
noma or melanoma (Fig. 26).
Metastatic Carcinomas and Melanomas
Metastases of carcinomas and melanomas with a pleomorphic microscopic appear-
ance are delineated by their reactivity with anticytokeratin or with melanocyte-selective
markers, respectively. These immunophenotypes are not shared by any other pleomor-
phic malignancy.
Pleomorphic Soft Tissue Tumors
Malignant peripheral nerve sheath tumors are unique among pleomorphic soft tissue
sarcomas, because they are the only lesions in this class that are capable of diffuse
expression of S100 protein, with or without CD57 (18). Pleomorphic rhabdomyosarco-
mas are globally positive for desmin and muscle-specific actin; in addition, they express
myoglobin, Myo-D1, or myogenin, all of which are proteins that are restricted to striated
muscle (18,111,118). Of the latter, Myo-D1 and myogenin are strictly limited to the
nuclei. Although it is a relatively insensitive marker in other forms of rhabdomyo-
sarcoma, myoglobin assumes some importance among pleomorphic neoplasms, because
“dedifferentiated” leiomyosarcomas also exhibit desmin and actin reactivity but are
devoid of myoglobin (18). Pleomorphic MFH lacks all antigens except vimentin;
Undifferentiated and Morphologically Ambiguous Neoplasms 349
pleomorphic liposarcoma differs from this profile in showing focal S100 protein positiv-
ity in signet ring or multivacuolated tumor lipoblasts.
several topographic origins. The proffered listings attempt to account for most, if not
all, of the combinations of immunoreactivity that may be encountered. In addition,
they cite the primary sites that are most likely for each combination of results, in order
of frequency. For example, the combination of CEA+/CA19-9+ (lower midportion of
tier 3) would suggest a tumor origin in the GI tract, pancreas, bladder, endocervix, and
lung, in relative order of likelihood. Although such lists may seem unsatisfying at first
glance (because they do not provide only one anatomic choice), integration of clinical
information and morphologic details will usually allow for a single diagnostic choice
to be made with confidence.
Because the information presented here has been accumulated over a long period, the
authors have had the opportunity to undertake validation studies of the algorithm for
metastatic carcinoma of unknown origin (MCUO). In those cases in which primary sites
were found after initial biopsy of the tumor, either by clinical evaluations or by autopsy,
overall correspondence with the pathologic prediction for the site of origin has been 67%.
Others have had similar rates of success with immunohistochemical evaluations. This
figure may not seem overly impressive, but it should be compared with the accuracy
and efficiency of radiographic and other nonmorphologic methods, as outlined in the
introduction of this presentation. Moreover, when mistaken predictions occurred in our
experience, they most often involved tumors in the same general immunophenotypic class;
for example, a tumor felt to be most consistent with a pancreatic carcinoma actually proved
to be a lung cancer with a largely superimposable set of immunoreactivity patterns.
Another practical aspect of our algorithmic approach is still early in its evolution
but is mentioned her for interest. Oncologists are well aware that only a minority of
MCUO cases have definable primary tumors that can be found during life. Accordingly,
most patients in this group have been treated with a standard consensus panel of
chemotherapeutic agents designed to have activity against a broad spectrum of malignan-
cies. Nevertheless, that approach is probably not an optimal one compared with specific
available treatments aimed at specific organ sites, judging by the dismal survival
results that are realized. Thus, some oncologist colleagues have elected to administer
chemotherapy regimens that are dictated by the predictive result of the MCUO algorithm.
In other words, if lung carcinoma is the favored interpretation by the pathologist, a set
of drugs tailored to that diagnosis would be given to the patient even if no pulmonary
mass were evident by clinical evaluation. Very preliminary results of such a process
have been promising, showing outcomes that are comparable to those of patients with
high-stage tumors of known anatomic derivation and somewhat better than those treated
with generic MCUO therapy. Obviously, considerable work will be required to decide
whether this is ultimately the optimal means of treating MCUO cases.
The tenets that have just been discussed are most applicable to the diagnosis of
poorly differentiated carcinomas that lack neuroendocrine differentiation. Neuroendo-
crine tumors are more homogeneous, regardless of their anatomic origins; all of them
share the potential to express CD57, CGA, synaptophysin, and various neuropeptides
and amines. Therefore, immunohistology is not useful in predicting where such tumors
arose. (See also Chapter 18.)
352 Wick and Cerilli
PERSPECTIVES
Throughout this chapter, an integrated multiparametric approach to the immunohisto-
logic diagnosis of poorly differentiated and morphologically ambiguous malignant
tumors has been emphasized. This is indeed an effective process, as reflected by its
beneficial results in many medical centers. Nonetheless, one should not lose sight of
the fact that surgical pathology is still based on skill in morphologic interpretation;
one cannot achieve success as a diagnostic immunohistochemist without having a sound
foundation in histopathologic analysis. Also, we do not mean to imply that another
method of detailed pathologic assessment—electron microscopy—has been relegated
to a historical perspective. It is still invaluable in the evaluation of selected neoplasms
and often supplies information that is “synergistic” with that of immunohistology.
Also, immunohistochemistry is not itself being outmoded by more “molecular”
techniques, which are being increasingly integrated currently into surgical pathology.
It is rather anticipated that all these methods will be employed selectively in the future,
at the discretion of the pathologist, to provide rapid and comprehensive solutions for
problem cases in anatomic pathology.
REFERENCES
1. Prento, P. and Lyon, H. (1997) Commercial formalin substitutes for histopathology. Bio-
tech. Histochem. 72, 273–282.
2. Arnold, M. M., Srivastava, S., Fredenburgh, J., Stockard, C. R., Myers, R. B., and Grizzle,
W. E. (1996) Effects of fixation and tissue processing on immunohistochemical demonstra-
tion of specific antigens. Biotech. Histochem. 71, 224–230.
3. Miller, R. T., Swanson, P. E., and Wick, M. R. Fixation and epitope retrieval in diagnostic
immunohistochemistry: a concise review with practical considerations. Appl. Immunohisto-
chem. Mol. Morphol. 8, 228–235.
4. Falini, B., and Taylor, C. R. (1983) New developments in immunoperoxidase techniques
and their application. Arch. Pathol. Lab. Med. 107, 105–117.
5. Giorno, R. (1984) A comparison of two immunoperoxidase staining methods based on
the avidin-biotin interaction. Diagn. Immunol. 2, 161–169.
6. Swanson, P. E. (1988) Foundations of immunohistochemistry: a practical review. Am. J.
Clin. Pathol. 90, 333–339.
7. Gown, A. M., DeWever, N., and Battifora, H. (1993) Microwave-based antigenic unmask-
ing: a revolutionary new technique for routine immunohistochemistry. Appl. Immunohisto-
chem. 1, 1256–1266.
8. Cattoretti, G., and Suurmeijer, A. (1995) Antigen unmasking for formalin-fixed paraffin-
embedded tissues: a review. Adv. Anat. Pathol. 2, 2–9.
9. DeJong, A. S. H., Van Kessel-Van Mark, M., and Raap, A. K. (1985) Sensitivity of various
visualization methods for peroxidase and alkaline phosphatase activity in immunoenzyme
histochemistry. Histochem. J. 17, 119–130.
354 Wick and Cerilli
10. Suffin, S. C., Muck, K. B., and Young, J. C. (1979) Improvement of the glucose oxidase
immunoenzyme technique: use of a tetrazolium whose formazan is stable without heavy
metal chelation. Am. J. Clin. Pathol. 71, 492–496.
11. Miettinen, M. (1993) Keratin immunohistochemistry: update on applications and pitfalls.
Pathol. Annu. 28, 113–143.
12. Andrade, R. E., Hagen, K. A., Swanson, P. E., and Wick, M. R. (1988) The use of ficin
for proteolysis in immunostaining of paraffin sections. Am. J. Clin. Pathol. 90, 33–39.
13. Battifora, H. (1988) Diagnostic uses of antibodies to keratins. Prog. Surg. Pathol. 8, 1–16.
14. Chan, J. K. C., Suster, S., Wenig, B. M., Tsang, W. Y., Chan, J. B., and Lau, A. L.
(1997) Cytokeratin 20 immunoreactivity distinguishes Merkel cell (primary cutaneous
neuroendocrine) carcinomas and salivary gland small cell carcinomas from small cell
carcinomas of various sites. Am. J. Surg. Pathol. 21, 226–234.
15. Pruss, R. M., Mirsky, R., Raff, M. C., Thrope, R., Dowling, A. J., and Anderton, B. H.
(1981) All classes of intermediate filaments share a common antigenic determinant defined
by a monoclonal antibody. Cell 27, 419–428.
16. Caselitz, J., Janner, M., Breitbart, E., Weber, K., and Osborn, M. (1983) Malignant
melanomas contain only the vimentin type of intermediate filaments. Virchows Arch. A.
400, 43–51.
17. Gustmann, C., Altmannsberger, M., Osborn, M., Griesser, H., and Feller, A. C. (1991)
Cytokeratin expression and vimentin content in large cell anaplastic lymphomas and other
non-Hodgkin’s lymphomas. Am. J. Pathol. 138, 1413–1422.
18. Swanson, P. E., and Wick, M. R. (1995) Soft tissue tumors, in Diagnostic Immunopathol-
ogy, 2nd ed. (Colvin, R., Bhan, A., and McCluskey, R., eds.), Lippincott-Raven, Philadel-
phia, pp. 599–632.
19. Battifora, H. (1991) Assessment of antigen damage in immunohistochemistry: the vimentin
internal control. Am. J. Clin. Pathol. 96, 669–671.
20. Truong, L. D., Rangdaeng, S., Cagle, P. T., Ro, J. Y., Hawkins, H., and Font, R. L. (1990)
The diagnostic utility of desmin: a study of 584 cases and review of the literature. Am.
J. Clin. Pathol. 93, 305–314.
21. Trojanowski, J. Q., Lee, V. M. Y., and Schlaepfer, W. W. (1984) An immunohistochemical
study of human central and peripheral nervous system tumors with monoclonal antibodies
against neurofilaments and glial filaments, Hum. Pathol. 15, 248–257.
22. Morrison, C. D. and Prayson, R. A. (2000) Immunohistochemistry in the diagnosis of
neoplasms of the central nervous system Semin. Diagn. Pathol. 17, 204–215.
23. Kurtin, P. J. and Pinkus, G. S. (1985) Leukocyte common antigen—a diagnostic discrimi-
nant between hematopoietic and nonhematopoietic neoplasms in paraffin sections using
monoclonal antibodies: correlation with immunologic studies and ultrastructural localiza-
tion. Hum. Pathol. 16, 353–365.
24. Wick, M. R. (1988) Monoclonal antibodies to leukocyte common antigen, in Monoclonal
Antibodies in Diagnostic Immunohistochemistry (Wick, M. R. and Siegal, G. P., eds.),
Marcel Dekker, New York, pp. 285–307.
25. Heyderman, E., Steele, K., and Ormerod, M. G. (1979) A new antigen on the epithelial
membrane: its immunoperoxidase localization in normal and neoplastic tissue. J. Clin.
Pathol. 32, 35–44.
26. Sloane, J. P. and Ormerod, M. G. (1981) Distribution of epithelial membrane antigen in
normal and neoplastic tissues and its value in diagnostic pathology, Cancer 47, 178–185.
27. Pinkus, G. S. and Kurtin, P. J. (1985) Epithelial membrane antigen—a diagnostic discrimi-
nant in surgical pathology. Hum. Pathol. 16, 929–938.
28. Swanson, P. E., Manivel, J. C., Scheithauer, B. W., and Wick, M. R. (1989) Epithelial mem-
brane antigen in human sarcomas: an immunohistochemical study. Surg. Pathol. 2, 313–322.
Undifferentiated and Morphologically Ambiguous Neoplasms 355
29. Wick, M. R., Swanson, P. E., and Manivel, J. C. (1987) Placental-like alkaline phosphatase
reactivity in human tumors: an immunohistochemical study of 520 cases. Hum. Pathol.
18, 946–954.
30. Elgin, J., Phillips, J. G., Reddy, V. V., Gibbs, P. O., and Listinsky, C. M. (1999) Hodgkin’s
and non-Hodgkin’s lymphoma: spectrum of morphologic and immunophenotypic overlap.
Ann. Diagn. Pathol. 3, 263–275.
31. Salter, D. M., Krajewski, A. S., Miller, E. P., and Dewar, A. E. (1985) Expression of
leukocyte common antigen and epithelial membrane antigen in plasmacytic malignancies.
J. Clin. Pathol. 38, 843–844.
32. Schnitt, S. J. and Vogel, H. (1986) Meningiomas: diagnostic value of immunoperoxidase
staining for epithelial membrane antigen. Am. J. Surg. Pathol. 10, 640–649.
33. Ruitenbeek, C. T., Gouw, A. S. H., and Poppema, S. (1994) Immunocytology of body
cavity fluids: MOC-31, a monoclonal antibody discriminating between mesothelial and
epithelial cells. Arch. Pathol. Lab. Med. 118, 265–269.
34. Sosolik, R. C., McGaughy, V. R., and DeYoung, B. R. (1997) Anti-MOC31: a potential
addition to the pulmonary adenocarcinoma versus mesothelioma immunohistochemistry
panel. Mod. Pathol. 10, 716–719.
35. Chenard-Neu, M. P., Kabou, A., Mechine, A., Brolly, F., Orion, B., and Bellocq, J. P. (1998)
Immunohistochemistry in the differential diagnosis of mesothelioma and adenocarcinoma:
evaluation of 5 new antibodies and 6 traditional antibodies. Ann. Pathol. 18, 460–465.
36. DeYoung, B. R. and Wick, M. R. (2000) Immunohistologic evaluation of metastatic
carcinomas of unknown origin: an algorithmic approach. Semin. Diagn. Pathol. 17, 184–
193.
37. Thor, A., Ohuchi, N., Szpak, C. A., Johnston, W. W., and Schlom, J. (1986) Distribution
of oncofetal antigen tumor-associated glycoprotein-72, defined by monoclonal antibody
B72.3. Cancer Res. 46, 3118–3124.
38. Loy, T. S. and Nashelsky, M. B. (1993) Reactivity of B72.3 with adenocarcinomas: an
immunohistochemical study of 476 cases. Cancer 72, 2495–2498.
39. Manivel, J. C., Jessurun, J., Wick, M. R., et al. (1987) Placental alkaline phosphatase
immunoreactivity in testicular germ cell neoplasms. Am. J. Surg. Pathol. 11, 21–30.
40. Niehans, G. A., Manivel, J. C., Copland, G. T., et al. (1988) Immunohistochemistry of
germ cell and trophoblastic neoplasms. Cancer 62, 1113–1123.
41. Suster, S., Moran, C. A., Dominguez-Malagon, H., and Quevedo-Blanco, P. (1998) Germ
cell tumors of the mediastinum and testis: a comparative immunohistochemical study of
120 cases. Hum. Pathol. 29, 737–742.
41a.Cheville, J. C., Rao, S., Iczkowshi, K. A., Lohse, C. M., and Pankratz, V. S. (2000)
Cytokeratin expression in seminoma of the human testis. Am. J. Clin. Pathol. 113, 583–588.
42. Thung, S. N., Gerber, M. A., Sarno, R., and Popper, H. (1979) Distribution of five antigens
in hepatocellular carcinoma. Lab. Invest. 43, 101–105.
43. Stiller, D., Bahn, H., and Pressler, H. (1986) Immunohistochemical demonstration of
alpha-fetoprotein in testicular germ cell tumors. Acta. Histochem. Suppl. 33, 225–231.
44. Okamoto, T., Hirabayashi, K., and Ishiguro, T. (1993) Immunohistochemical type distinc-
tion of alpha-fetoprotein in various alpha-fetoprotein-secreting tumors. Jpn. J. Cancer Res.
84, 360–364.
45. Colcher, D., Hand, P. H., Nuti, M., and Schlom, J. (1983) Differential binding to human
mammary and nonmammary tumors of monoclonal antibodies reactive with carcinoembry-
onic antigen. Cancer Invest. 1, 127–138.
46. Wick, M. R. (1988) Monoclonal antibodies to carcinoembryonic antigen, in Monoclonal
Antibodies in Diagnostic Immunohistochemistry (Wick, M. R. and Siegal, G. P., eds.),
Marcel Dekker, New York, pp. 539–567.
356 Wick and Cerilli
47. Sheahan, K., O’Brien, M. J., Burke, B., et al. (1990) Differential reactivities of carcinoem-
bronic antigen (CEA) and CEA-related monoclonal and polyclonal antibodies in common
epithelial malignancies. Am. J. Clin. Pathol. 94, 157–164.
48. Nadji, M., Tabei, S. Z., Castro, A., et al. (1981) Prostatic-specific antigen: an immunohisto-
logic marker for prostatic neoplasms. Cancer 48, 1229–1232.
49. Miller, G. J. (1982) The use of histochemistry and immunohistochemistry in evaluating
prostatic neoplasia. Prog. Surg. Pathol. 5, 115–134.
50. Papsidero, L. D., Croghan, G. A., Asirwatham, J., et al. (1985) Immunohistochemical
demonstration of prostate-specific antigen in metastases with the use of monoclonal anti-
body F5. Am. J. Pathol. 121, 451–454.
51. Torenbeek, R., Lagendijk, J. H., Van Diest, P. J., et al. (1998) Value of a panel of antibodies
to identify the primary origin of adenocarcinomas presenting as bladder carcinoma. Histo-
pathology 32, 20–27.
52. Bostwick, D. G. (1994) Prostate-specific antigen: current role in diagnostic pathology of
prostate cancer. Am. J. Clin. Pathol. 102 (4 suppl. 1), S31–S37.
52a.Bostwick, D. G. (1994) Prostate specific antigen. Current role in diagnostic pathology of
prostate cancer. Am. J. Clin. Pathol. 102, 831–837.
53. Permanetter, W., Nathrath, W. B. J., and Lohrs, U. (1982) Immunohistochemical analysis
of thyroglobulin and keratin in benign and malignant thyroid tumors. Virchows Arch.
398, 221–228.
54. DeMicco, C., Ruf, J., Carayon, P., et al. (1987) Immunohistochemical study of thyroglobu-
lin in thyroid carcinomas with monoclonal antibodies. Cancer 59, 471–476.
55. Carcangiu, M. L., Steeper, T., Zampi, G., and Rosai, J. (1985) Anaplastic thyroid carcinoma:
a study of 70 cases. Am. J. Clin. Pathol. 83, 135–158.
56. Carcangiu, M. L., Zampi, G., and Rosai, J. (1984) Poorly differentiated (“insular”) thyroid
carcinoma: a reinterpretation of Langhans’ “wuchernde Struma.” Am. J. Surg. Pathol.
8, 655–668.
57. Guazzi, S., Price, M., DeFelice, M., et al. (1990) Thyroid nuclear factor-1 (TTF-1) contains
a homeodomain and displays a novel DNA binding specificity. EMBO J. 9, 3631–3639.
58. Lazzaro, D., Price, M., DeFelice, D., and DiLauro, R. (1991) The transcription factor
TTF-1 is expressed at the onset of thyroid and lung morphogenesis and in restricted regions
of the fetal brain. Development 113, 1093–1104.
59. Kelly, S. E., Bachurski, C. J., Burhans, M. S., and Glassor, J. W. (1996) Transcription
of the lung-specific surfactant protein C gene is mediated by TTF-1. J. Biol. Chem.
27, 6881–6888.
60. Ordonez, N. G. (2000) Thyroid transcription factor-1 is a marker for lung and thyroid
carcinomas. Adv. Anat. Pathol. 7, 123–127.
61. Ordonez, N. G. (2000) Value of thyroid transcription factor-1, E-cadherin, BG8, WT1,
and CD44S immunostaining in distinguishing epithelial pleural mesothelioma from pulmo-
nary and nonpulmonary adenocarcinoma. Am. J. Surg. Pathol. 24, 598–606.
62. Katoh, R., Kawaoi, A., Miyagi, E., et al. (2000) Thyroid transcription factor-1 in normal,
hyperplastic, and neoplastic follicular thyroid cells examined by immunohistochemistry
and nonradioactive in-situ hybridization. Mod. Pathol. 13, 570–576.
63. Kaufmann, O. and Dietel, M. (2000) Expression of thyroid transcription factor-1 in pulmo-
nary and extrapulmonary small cell carcinomas and other neuroendocrine carcinomas of
various primary sites. Histopathology 36, 415–420.
64. Agoff, S. N., Lamps, L. W., Philip, A. T., et al. (2000) Thyroid transcription factor-1
is expressed in extrapulmonary small cell carcinomas but not in other extrapulmonary
neuroendocrine tumors. Mod. Pathol. 13, 238–242.
65. Hanly, A. J., Elgart, G. W., Jorda, M., Smith, J., and Nadji, M. (2000) Analysis of thyroid
Undifferentiated and Morphologically Ambiguous Neoplasms 357
transcription factor-1 and cytokeratin 20 separates Merkel cell carcinoma from small cell
carcinoma of lung. J. Cutan. Pathol. 27, 118–120.
66. Wick, M. R., Lillemoe, T. J., Copland, G. T., et al. (1989) Gross cystic disease fluid
protein-15 as a marker for breast carcinoma. Hum. Pathol. 20, 281–287.
67. Mazoujian, G., and Haagensen, D. E., Jr. (1990) The immunopathology of gross cystic
disease fluid proteins. Ann. NY Acad. Sci. 586, 188–197.
68. Lee, A. K., DeLellis, R. A., Rosen, P. P., et al. (1984) Alpha-lactalbumin as an immunohis-
tochemical marker for metastatic breast carcinomas. Am. J. Surg. Pathol. 8, 93–100.
69. Mesa-Tejada, R., Palakodety, R. B., Leon, J. A., Khatcherian, A. O., and Greaton, C. J.
(1988) Immunocytochemical distribution of a breast carcinoma-associated glycoprotein
identified by monoclonal antibodies. Am. J. Pathol. 130, 305–314.
70. Loy, T. S., Chapman, R. K., Diaz-Arias, A. A., Bulatao, I. S., and Bickel, J. T. (1991)
Distribution of BCA-225 in adenocarcinomas: an immunohistochemical study of 446
cases. Am. J. Clin. Pathol. 96, 326–329.
71. Gatalica, Z. and Miettinen, M. (1994) Distribution of carcinoma antigens CA19-9 and CA15-
3; an immunohistochemical study of 400 tumors. Appl. Immunohistochem. 2, 205–211.
71a.McCluggage, W. G., Patternson, A., and Maxweill, P. (2000) Aggressive angiomyxoma
of pelvic parts exhibits oestrogen and progesterone receptor positivity. J. Clin. Pathol.
53, 603–605.
72. Yokazaki, H., Takekura, N., Takanashi, A., Tabuchi, J., Haruta, R., and Tahara, E. (1988)
Estrogen receptors in gastric adenocarcinoma: a retrospective immunohistochemical analy-
sis. Virchows Arch. A. (Pathol. Anat.) 413, 297–302.
73. Diaz, N. M., Mazoujian, G., and Wick, M. R. (1991) Estrogen-receptor protein in thyroid
neoplasms: an immunohistochemical analysis of papillary carcinoma, follicular carcinoma,
and follicular adenoma. Arch. Pathol. Lab. Med. 115, 1203–1207.
74. Kaufmann, O., Baume, H., and Dietel, M. (1998) Detection of estrogen receptors in
noninvasive and invasive transitional cell carcinomas of the urinary bladder using both
conventional immunohistochemistry and the tyramide staining amplification technique.
J. Pathol. 186, 165–168.
75. Bast, R. C., Feeney, M., Lazarus, H., et al. (1981) Reactivity of a monoclonal antibody
with human ovarian carcinoma. J. Clin. Invest. 68, 1331–1337.
76. Kabawat, S. E., Bast, R. C., Welch, W. R., et al. (1983) Immunopathologic characterization
of a monoclonal antibody that recognizes common surface antigens of human ovarian
tumors of serous, endometrioid, and clear cell types. Am. J. Clin. Pathol. 79, 98–104.
77. Koelma, I. A., Nap, M., Rodenburg, C. J., and Fleuren, G. J. (1987) The value of tumor
marker CA125 in surgical pathology. Histopathology 11, 287–294.
78. Loy, T. S., Quesenberry, J. T., and Sharp, S. C. (1992) Distribution of CA125 in adenocarci-
nomas: an immunohistochemical study of 481 cases. Am. J. Clin. Pathol. 98, 175–179.
79. Itzkowitz, S. H., Yuan, M., Fukushi, Y., et al. (1988) Immunohistochemical comparison
of Lea, monosialosyl Lea (CA19-9), and disialosyl Lea antigens in human colorectal and
pancreatic tissues. Cancer Res. 48, 3834–3842.
80. Loy, T. S., Sharp, S. C., Andershock, C. J., and Craig, S. B. (1993) Distribution of CA19-
9 in adenocarcinomas and transitional cell carcinomas: an immunohistochemical study of
527 cases. Am. J. Clin. Pathol. 99, 726–728.
81. Sheibani, K., Battifora, H., Burke, J. S., and Rappaport, H. (1986) Leu-M1 antigen in human
neoplasms: an immunohistologic study of 400 cases. Am. J. Surg. Pathol. 10, 227–236.
82. Arber, D. A. and Weiss, L. (1993) CD15: a review. Appl. Immunohistochem. 1, 17–30.
83. Wick, M. R., Mills, S. E., and Swanson, P. E. (1990) Expression of “myelomonocytic”
antigens in mesotheliomas and adenocarcinomas involving the serosal surfaces. Am. J.
Clin. Pathol. 94, 18–26.
358 Wick and Cerilli
84. Tos, A. P. and Doglioni, C. (1998) Calretinin: a novel tool for diagnostic immunohistochem-
istry. Adv. Anat. Pathol. 5, 61–66.
85. Gotzos, V., Wintergerst, E. S., Musy, J. P., Spichtin, H. P., and Genton, C. Y. (1999)
Selective distribution of calretinin in adenocarcinomas of the human colon and adjacent
tissues. Am. J. Surg. Pathol. 23, 701–711.
85a.Chhieng, D. C., et al. (2000) Calretinin staining pattern aids in the differentiation of
mesothelioma from adenocarcinoma in serous effusions. Cancer 90, 194–200.
85b.Kitazume, H., et al. (2000) Cytologic differential diagnosis among reactive mesothelial
cells, malignant mesothelioma, and adenocarcinoma: utility of combined E-cadherin and
calretinin immunostaining. Cancer 90, 55–60.
85c.Ordonez, N. G. (1998) Value of calretinin immunostaining in differentiating epithelial
mesothelioma from lung adenocarcinoma. Mod. Pathol. 11, 929–933.
85d.Atanoos, R. L., Dojcinov, S. D., Webb, R., and Gibbs, A. R. (2000) Antimesothelial
markers in sarcomatoid mesothelioma and other spindle cell neoplasms. Histopathology
37, 224–231.
86. Drier, J., Swanson, P. E., Cherwitz, D. L., and Wick, M. R. (1987) S100 protein immuno-
reactivity in poorly-differentiated carcinomas: immunohistochemical comparison with
malignant melanoma. Arch. Pathol. Lab. Med. 111, 447–452.
87. Wick, M. R., Ockner, D. M., Mills, S. E., Ritter, J. H., and Swanson, P. E. (1998)
Homologous carcinomas of the breasts, salivary glands, and skin. Am. J. Clin. Pathol.
109, 75–84.
88. Matsushima, S., Mori, M., Adachi, Y., Matsukuma, A., and Sugimachi, K. (1994) S100
protein-positive breast carcinomas: an immunohistochemical study. J. Surg. Oncol. 55,
108–113.
89. Loeffel, S. C., Gillespie, G. Y., Mirmiran, A., Sawhney, D., Askin, F. B., and Siegal, G.
P. (1985) Cellular immunolocalization of S100 protein within fixed tissue sections by
monoclonal antibodies. Arch. Pathol. Lab. Med. 109, 117–122.
90. Gown, A. M., Vogel, A. M., Hoak, D., Gough, F., and McNutt, M. A. (1986) Monoclonal
antibodies specific for melanocytic tumors distinguish subpopulations of melanocytes. Am.
J. Pathol. 123, 195–203.
91. Wick, M. R., Swanson, P. E., and Rocamora, A. (1988) Recognition of malignant melanoma
by monoclonal antibody HMB-45. J. Cutan. Pathol. 15, 201–207.
92. Kaufmann, O., Koch, S., Burghardt, J., Andring, H., and Dietel, M. (1998) Tyrosinase,
Melan-A, and KBA62 as markers for the immunohistochemical identification of metastatic
amelanotic melanomas on paraffin sections. Mod. Pathol. 11, 740–746.
93. Lloyd, R. V. and Wilson, B. S. (1983) Specific endocrine tissue marker defined by a
monoclonal antibody. Science 222, 628–630.
94. Wilson, B. S. and Lloyd, R. V. (1984) Detection of chromogranin in neuroendocrine cells
with a monoclonal antibody. Am. J. Pathol. 115, 458–468.
95. Wick, M. R. (2000) Immunohistology of neuroendocrine and neuroectodermal tumors.
Semin. Diagn. Pathol. 17, 194–203.
96. Miettinen, M. (1987) Synaptophysin and neurofilament protein as markers for neuroendo-
crine tumors. Arch. Pathol. Lab. Med. 111, 813–818.
97. Gould, V. E., Lee, I., Wiedenmann, B., et al. (1986) Synaptophysin: a novel marker for
neurons, certain neuroendocrine cells, and their neoplasms. Hum. Pathol. 17, 979–983.
98. Buffa, R., Rindi, G., Sessa, F., et al. (1987) Synaptophysin immunoreactivity and
small clear vesicles in neuroendocrine cells and related tumors. Mol. Cell. Probes
1, 367–381.
99. Stirdsberg, M. (1995) The use of chromogranin, synaptophysin, and islet amyloid polypep-
tide as markers for neuroendocrine tumors. Ups. J. Med. Sci. 100, 169–199.
Undifferentiated and Morphologically Ambiguous Neoplasms 359
100. Lipinski, M., Braham, K., Caillaud, J. M., et al. (1983) HNK-1 antibody detects an antigen
expressed on neuroectodermal cells. J. Exp. Med. 158, 1775–1780.
101. Chu, P. G., Chang, K. L., Arber, D. A., and Weiss, L. M. (2000) Immunophenotyping of
hematopoietic neoplasms. Semin. Diagn. Pathol. 17, 236–256.
102. Perentes, E. and Rubinstein, L. J. (1985) Immunohistochemical recognition of human
nerve sheath tumors by anti-Leu 7 (HNK-1) monoclonal antibody. Acta. Neuropathol.
68, 319–324.
103. Tsutsumi, Y. (1984) Leu-7 immunoreactivity as a histochemical marker for paraffin-
embedded neuroendocrine tumors. Acta Histochem. Cytochem. 17, 15–21.
104. Dehner, L. P. (1993) Primitive neuroectodermal tumor and Ewing’s sarcoma. Am. J. Surg.
Pathol. 17, 1–13.
105. Lumadue, J. A., Askin, F. B., and Perlman, E. J. (1994) MIC2 analysis of small cell
carcinoma. Am. J. Clin. Pathol. 102, 692–694.
106. Tsukada, T., McNutt, M. A., Ross, R., and Gown, A. M. (1987) HHF35, a muscle actin-
specific monoclonal antibody. II. Reactivity in normal, reactive, and neoplastic human
tissues. Am. J. Pathol. 127, 389–402.
107. Rangdaeng, S. and Truong, L. D. (1991) Comparative immunohistochemical staining
for desmin and muscle-specific actin: a study of 576 cases. Am. J. Clin. Pathol. 96, 32–
45.
108. Miettinen, M., Lindenmayer, A. E., and Chanbal, A. (1994) Endothelial cell markers
CD31, CD34, and BNH9 antibody to H- and Y- antigens: evaluation of their specificity
and sensitivity in the diagnosis of vascular tumors and comparison with von Willebrand
factor. Mod. Pathol. 7, 82–90.
109. Traweek, S. T., Kandalaft, P. L., Mehta, P., and Battifora, H. (1991) The human
hematopoietic progenitor cell antigen (CD34) in vascular neoplasms. Am. J. Clin.
Pathol. 96, 25–31.
110. Cohen, P. R., Rapini, R. P., and Farhood, A. I. (1993) Expression of the human hematopoi-
etic progenitor cell antigen CD34 in vascular and spindle cell tumors. J. Cutan. Pathol.
20, 15–20.
111. Suster, S. (2000) Recent advances in the application of immunohistochemical markers for
the diagnosis of soft tissue tumors. Semin. Diagn. Pathol. 17, 225–235.
112. DeYoung, B. R., Wick, M. R., Fitzgibbon, J. F., et al. (1993) CD31: an immunospecific
marker for endothelial differentiation in human neoplasms. Appl. Immunohistochem. 1, 97–
100.
113. Karmochkine, M. and Boffa, M. C. (1997) Thrombomodulin: physiology and clinical
applications. Rev. Med. Interne 18, 119–125.
114. Ordonez, N. G. (1997) Value of thrombomodulin immunostaining in the diagnosis of
mesothelioma. Histopathology 31, 25–30.
115. Appleton, M. A., Attanoos, R. L., and Jasani, B. (1996) Thrombomodulin as a marker of
vascular and lymphatic tumors. Histopathology 29, 1531–157.
116. Holthofer, H., Virtanen, I., Karineimi, A. L., et al. (1982) Ulex europaeus 1 lectin as a
marker for vascular endothelium in human tissues. Lab. Invest. 47, 60–66.
117. Leader, M., Collins, M., Patel, J., and Henry, K. (1986) Staining of factor VIII-related
antigen and Ulex europaeus 1 (UEA-1) in 230 tumors: an assessment of their specificity
for angiosarcoma and Kaposi’s sarcoma. Histopathology 18, 1153–1162.
117a.Nandedkar, M. A., Palazzo, J., Abbondanzo, S. L., Lasota, J., and Miettinen, M. (1998)
CD45 (leukocyte common antigen) immunoreactivity in metastatic undifferentiated and
neuroendocrine carcinoma: a potential diagnostic pitfall. Mod. Pathol. 11, 1204–1210.
118. Devoe, R. and Weidner, N. (2000) Immunohistochemistry of small round-cell tumors.
Semin. Diagn. Pathol. 17, 216–224.
360 Wick and Cerilli
119. Parham, D. M., Dias, P., Kelly, D. R., Rutledge, J. C., and Houghton, P. (1992) Desmin-
positivity in primitive neuroectodermal tumors of childhood. Am. J. Surg. Pathol. 16, 483–
492.
120. Guinee, D. G., Jr., Fishback, N. F., Koss, M. N., et al. (1994) The spectrum of immunohisto-
chemical staining of small cell lung carcinoma in specimens from transbronchial and open-
lung biopsies. Am. J. Clin. Pathol. 102, 406–414.
120a.Eusebi, V., Damiani, S., Pasquinelli, G., et al. (2000) Small cell neuroendocrine carcinoma
with skeletal muscle differentiation. Am. J. Surg. Pathol. 24, 223–230.
121. Wick, M. R. and Swanson, P. E. (1993) “Carcinosarcomas”—current perspectives and an
historical review of nosological concepts. Semin. Diagn. Pathol. 10, 118–127.
122. Wick, M. R., Fitzgibbon, J. F., and Swanson, P. E. (1993) Cutaneous sarcomas and
sarcomatoid neoplasms of the skin. Semin. Diagn. Pathol. 10, 148–158.
123. Elgin, J., Phillips, J. G., Reddy, V. V., Gibbs, P. O., and Listinsky, C. M. (1999) Hodgkin’s
and non-Hodgkin’s lymphoma: spectrum of morphologic and immunophenotypic overlap.
Ann. Diagn. Pathol. 3, 263–275.
124. Manivel, J. C., Wick, M. R., Dehner, L. P., and Sibley, R. K. (1987) Epithelioid sarcoma:
an immunohistochemical study. Am. J. Clin. Pathol. 87, 319–326.
125. Schmidt, D. and Harms, D. (1987) Epithelioid sarcoma in children and adolescents: an
immunohistochemical study. Virchows Arch. A 410, 423–431.
126. Papas-Corden, P. C., Zarbo, R. J., Gown, A. M., and Crissman, J. D. (1989) Immunohisto-
chemical characterization of synovial, epithelioid, and clear-cell sarcomas. Surg. Pathol.
2, 43–58.
127. Swanson, P. E., Dehner, L. P., Sirgi, K. E., and Wick, M. R. (1994) Cytokeratin-immuno-
reactivity in malignant tumors of bone and soft tissue. Appl. Immunohistochem. 2, 103–112.
128. Sirgi, K. E., Wick, M. R., and Swanson, P. E. (1993) B72.3 and CD34 immunoreactivity
in malignant epithelioid soft tissue tumors: adjuncts in the recognition of endothelial
neoplasms. Am. J. Surg. Pathol. 17, 179–185.
129. Wick, M. R., Swanson, P. E., Scheithauer, B. W., and Manivel, J. C. (1987) Malignant
peripheral nerve sheath tumor: an immunohistochemical study of 62 cases. Am. J. Clin.
Pathol. 87, 425–433.
130. Sarlomo-Rikala, M., Kovatich, A. J., Barusevicius, A., et al. (1998) CD117: a sensitive
marker for gastrointestinal stromal tumors that is more specific than CD34. Mod. Pathol.
11, 728–734.
131. Brown, R. W., Campagna, L. B., Dunn, J. K., et al. (1997) Immunohistochemical identifica-
tion of tumor markers in metastatic adenocarcinoma: a diagnostic adjunct in the determina-
tion of primary site. Am. J. Clin. Pathol. 107, 12–19.
132. Moran, C. A., Wick, M. R., and Suster, S. (2000) The role of immunohistochemistry in
the diagnosis of malignant mesothelioma. Semin. Diagn. Pathol. 17, 178–183.
133. Wick, M. R., Loy, T., Mills, S. E., et al. (1990) Malignant epithelioid pleural mesothelioma
versus peripheral pulmonary adenocarcinoma: a histochemical, ultrastructural, and immu-
nohistochemical study of 103 cases. Hum. Pathol. 21, 759–766.
18
Applications of Immunohistochemistry in the
Diagnosis of Endocrine Lesions
INTRODUCTION
Endocrine cells and tumors can be divided into neuroendocrine and nonneuroendo-
crine depending on their immunophenotypic and ultrastructural features. Neuroendo-
crine cells and tumors are members of the diffuse or dispersed neuroendocrine system
(DNES), all of which have secretory granules demonstrated by ultrastructural studies.
They can also be recognized with broad-spectrum neuroendocrine markers such as
chromogranin and synaptophysin. Members of the DNES also produce specific peptides
that can serve as target markers for cell and tumor identification.
Nonneuroendocrine cells and tumors include adrenal cortical cells, thyroid follicular
cells, and steroid-producing cells of the ovary and testes. Markers such as inhibins are
useful for detecting adrenocortical cells and tumors and gonadal tumors, whereas
thyroglobulin is relatively specific for thyroid follicular cells.
The use of specific antibodies, also referred to as markers, has provided a great deal
of specificity for the characterization of endocrine cells and tumors.
Endocrine System
The endocrine system can be divided into neuroendocrine and nonneuroendocrine
cells and tumors. Most endocrine cells belong to the neuroendocrine category (Table 1).
The neuroendocrine system is also referred to as the diffuse or dispersed neuroendocrine
system (DNES), which consists of endocrine cells that are present in most areas of the
body. Although DNES cells are not all derived from the neural crest, as originally
proposed, they all share common features, including expression of chromogranin and
synaptophysin, and have dense-core granules detected by ultrastructural examination
(1–4). The nonneuroendocrine cells consist of the steroid-producing cells of the adrenal
cortex, ovary, and testes and the thyroglobulin-producing follicular cells of the thyroid.
There are many broad-spectrum and specific immunohistochemical markers that can
be used to characterize cells and tumors of the DNES (Tables 1 and 2). Many of these,
such as the chromogranin PGP 9.5 and synaptophysin, have been in use for over a
decade. Others, such as synaptic proteins and neuroendocrine-specific protein reticulons,
have been recently described in the past few years. To determine whether a tumor is
361
362 Lloyd
Table 1
Specific Peptide and Other Markers for Neuroendocrine Cells and Tumors
Cell type Neoplasm(s) Markersa
Adrenal medulla Pheochromocytoma, Catecholamines, erkephalin, VIP,
and neuroblastoma, somatostatin, S-100 protein
paraganglia ganglioneuroblastoma,
ganglioma
Gastrointestinal Benign and malignant Gastrin, somatostatin, VIP,
tract neuroendocrine tumors secretin, insulin, glucagons,
pancreatic polypeptide,
cholecystokinin
Lung Benign and malignant ACTH, endorphin, calcitonin,
neuroendocrine tumors, secretin, gastrin-releasing
including carcinoids peptide
Pancreas Benign and malignant pancreatic Insulin, glucagon pancreatic
endocrine tumors polypeptide, somatostatin, VIP,
gastrin, serotonin
Parathyroid Hyperplasia, adenoma, carcinoma Parathyroid hormone, parathyroid
hormone, related peptide
Pituitary Adenoma and carcinoma ACTH, growth hormone,
prolactin, follicle-stimulating
hormone, luteinizing hormone,
thyroid-stimulating hormone,
calcitonin
Thyroid C-cell Medullary thyroid carcinoma Calcitonin, somatostatin, ACTH
a
VIP, vasoactive intestinal polypeptide; ACTH, adrenocorticotropic hormone.
Table 2
Broad-Spectrum Neuroendocrine Markers
Chromogranin/secretogranin
Synaptophysin
PGP9.5
Leu7
Synaptic proteins
SNAP-25
Rab3A
Neural cell adhesion molecules (NCAM) CD57
Peptidylglycine α-amidating monooxygenase
Neuroendocrine-specific proteins (NSP)-reticulons
Diagnosis of Endocrine Lesions 363
Fig. 1. Chromogranin A expression in the adrenal gland. The medullary cells are positive,
and the cortical cells are negative. Immunoperoxidase-diaminobenzidine (DAB), original magni-
fication ×200.
Fig. 2. The neuroendocrine cells in the colonic crypts are strongly positive for chromogranin
A. Immunoperoxidase-DAB, original magnification ×200.
and -B usually increase the sensitivity for detecting neuroendocrine cells with Cg/
Sg antibodies.
Synaptophysin
Synaptophysin, a 38-kDa protein molecule, is a component of the membrane of
presynaptic vesicles. Gould et al. (11) first reported that it was widely distributed in
neurons and neuroendocrine cells and their neoplasms. It is another good broad-spectrum
neuroendocrine marker. Unlike the Cg/Sg proteins, which are well preserved in formalin-
fixed tissues, ethanol fixation provides optimal preservation of the synaptophysin anti-
gen. However, most of the anti-synaptophysin monoclonal antibodies work well in
formalin-fixed sections. Synaptophysin is present in vesicles in neuroendocrine cells
of the tumors, but immunostaining is present diffusely in the cytoplasm (Figs. 3 and 4).
Proconvertases
The proconvertases (PCs) are enzymes that process propeptides into active peptides
within cells (12,13). Some of these, including PC1/PC3 and PC2, are highly specific
for neuroendocrine cells and tumors and can be used as specific neuroendocrine markers.
Others, such as PC4, are present in the testes; PC5/6 is more prevalent in the gastrointesti-
nal tract and adrenal.
Neuron-Specific Enolase
Neuron-specific enolase (NSE), an enzyme also known as γ-enolase, is a highly
sensitive but not very specific marker for neuroendocrine cells and tumors. It is
Diagnosis of Endocrine Lesions 365
Fig. 3. Pancreatic islet showing diffuse staining for the broad-spectrum neuroendocrine
marker synaptophysin. Immunoperoxidase-DAB, original magnification ×200.
Fig. 4. Laryngeal neuroendocrine tumor staining positively for the broad-spectrum neuroendo-
crine marker synaptophysin. Immunoperoxidase-DAB, original magnification ×300.
366 Lloyd
commonly found in neurons, peripheral nerves, and neuroendocrine cells (14,15). Some
nonneuroendocrine cells and neoplasms also react with antisera against NSE. NSE
should be used only with other broad-spectrum markers of neuroendocrine cells in the
diagnosis of neuroendocrine tumors because of its relative lack of specificity. It is used
only infrequently in our laboratory.
PGP9.5
PGP9.5 is a soluble protein that was originally isolated from brain. It is a good
general marker for neuronal and neuroendocrine tissues (20,21). Interestingly, about
half of melanomas stain for PGP9.5, whereas these melanocytic tumors are usually
negative for Cg/Sg and for synaptophysin. PGP is a cytoplasmic soluble protein that
frequently colocalizes in normal and neoplastic neuroendocrine tissues.
Synaptic Proteins
A series of proteins involved in neurotransmitter secretion have also been associated
with neuroendocrine cells and tumors. These synaptic proteins include SNAP-25 nd
Rab3A (35–42).
SNAP-25 was identified as a neuron-specific protein associated with the plasma
membrane of the presynaptic nerve terminal in early studies. It was shown to be part
of the putative docking complex that is implicated in membrane fusion. A homologous
protein, SNAP-23, is expressed ubiquitously in human nonneuronal tissues, including
endocrine organs.
Rab3A, a small GTP-binding protein of the rab family, is expressed mainly in
neurons and neuroendocrine cells (39–42). Rab3A is thought to be an important control
system for exocytosis in neuron and neuroendocrine cells. A recent study showed
increased SNAP-25 immunoreactivity in most pituitary prolactin-cell adenomas and in
growth hormone-cell adenomas, suggesting that this protein is involved in the mecha-
nism of exocytosis in neoplasms derived from these cell types (42).
SPECIFIC MARKERS
Neuroendocrine Cells and Tumors
Once a neoplasm has been characterized as neuroendocrine by use of the broad-
spectrum endocrine markers, antibodies to specific peptide hormones can be used to
characterize the lesion better. A summary of the most common peptides and other
markers is shown in Table 2 (Figs. 5 and 6). General approaches in the use of these
markers should include the following considerations.
1. Very few peptides are restricted to one neuroendocrine cell type or organ. For
example, although calcitonin is more commonly produced by the thyroid C-cells, other
normal neuroendocrine tissues, including the anterior pituitary gland, also produce calci-
tonin.
2. Ectopic production of hormones by neoplasms is the rule rather than the exception.
Although medullary thyroid carcinomas usually produce calcitonin, hypercalcitonemia
may be associated with other tumors such as atypical laryngeal carcinoids, lung carci-
noids, or small cell lung carcinomas and others.
368 Lloyd
Fig. 5. Pancreatic islet staining positively for insulin with anti-insulin antibody. Immunoperox-
idase-DAB, original magnification ×300.
Fig. 6. Medullary thyroid carcinoma of the thyroid staining positively for calcitonin. This
marker helps to classify this tumor as a thyroid carcinoma derived from the C-cells. Immunoper-
oxidase-DAB, original magnification ×300.
Diagnosis of Endocrine Lesions 369
Table 3
Peptides and Other Markers for Nonneuroendocrine Cells and Tumors
Cell type Neoplasm(s) Marker
Thyroid Adenomas and carcinomas Thyroglobulin, keratin
follicular cell Thyroid transcription factor-1
Adrenal cortex Adenomas and carcinomas Inhibin, synaptophysin, keratin,
Ad4BP/SF1, vimentin, steroid-
metabolizing enzymes
Steroid- Benign and malignant neoplasms Inhibin, vimentin, steroid-
producing metabolizing enzymes
cells of gonads
3. Although the production of a hormone may be associated with specific signs and
symptoms, some other substances such as growth factors or related peptides may lead
to similar symptoms. For example, although hypoglycemia is commonly associated
with excess insulin production, insulin-like growth factor-II (IGF-II) may also be the
cause of the hypoglycemia. IGF-II is produced by some neuroendocrine tumors as well
as by mesenchymal and other neoplasms (47,48).
4. Failure to detect specific peptides in neuroendocrine tumors does not necessarily
mean that the substance is not being produced by the cells. It is possible that the peptide
may be rapidly secreted and not stored in the cells. The use of in situ hybridization to
detect mRNA usually helps to answer this question (see Chapter 3).
Additional markers such as S-100 acidic protein are helpful in characterizing normal
adrenal medulla and pheochromocytomas, whereas cytokeratin 20 is useful in separating
Merkel’s cell carcinoma from other neuroendocrine carcinomas.
CONCLUSIONS
Immunohistochemistry has contributed greatly to the characterization and diagnosis
of endocrine cells and tumors. Many more advances can be anticipated in the near
future with the development of new antibodies for the study of gene expression.
370 Lloyd
Fig. 7. The thyroid follicular cells are nonneuroendocrine cells. These cells express thyroglob-
ulin, which is a relatively specific marker of thyroid follicular cell differentiation. Immunoperoxi-
dase-DAB, original magnification ×300.
Fig. 8. The transcription factor Ad4BP/SF1 is relatively specific for adrenal cortical cells.
Positive staining is present in the nucleus of an adrenal cortical carcinoma. Nuclear localization is
typically seen with transcription factors. Immunoperoxidase-DAB, original magnification ×200.
Diagnosis of Endocrine Lesions 371
REFERENCES
1. Pearse, A. G. (1974) The APUD cell concept and its implications in pathology. Pathol.
Annu. 9, 27–41.
2. Pearse, A. G. and Takor, T. (1979) Embryology of the diffuse neuroendocrine system and
its relationship to the common peptides. Fed. Proc. 38, 2288–2294.
3. DeLellis, R. A. and Wolfe, H. J. (1981) The polypeptide hormone-producing neuroendocrine
cells and their tumors: an immunohistochemical analysis. Methods Achiev. Exp. Pathol.
10, 190–220.
4. Lloyd, R. V. (1996) Overview of neuroendocrine cells and tumors. Endocr. Pathol. 7, 323–
328.
5. Lloyd, R. V., Mervak, T., Schmidt, K., Warner, T. F., and Wilson, B. S. (1984) Immunohisto-
chemical detection of chromogranin and neuron-specific enolase in pancreatic endocrine
neoplasms. Am. J. Surg. Pathol. 8, 607–614.
6. Lloyd, R. V., Sisson, J. C., Shapiro, B., and Verhofstad, A. A. (1986) Immunohistochemical
localization of epinephrine, norephinephrine, catecholamine-synthesizing enzymes, and
chromogranin neuroendocrine cells and tumors. Am. J. Pathol. 125, 45–54.
7. Wilson, B. S. and Lloyd, R. V. (1984) Detection of chromogranin in neuroendocrine cells
with a monoclonal antibody. Am. J. Pathol. 115, 458–468.
8. Wiedenmann, B. and Huttner, W. B. (1989) Synaptophysin and chromogranins/secretogran-
ins—widespread constituents of distinct types of neuroendocrine vesicles and new tools in
tumor diagnosis. Virchows Arch. (B) 58, 95–121.
9. Schmid, K. W., Kroll, M., Hittmair, A., et al. (1991) Chromogranin A and B in adenomas
of the pituitary. An immunohistochemical study of 42 cases. Am. J. Surg. Pathol. 15, 1072–
1077.
372 Lloyd
10. Al-Khafaji, B., Noffsinger, A. E., Miller, M. A., DeVoe, G., Stemmermann, G. N., and
Fenoglio-Preiser, C. (1999) Immunohistologic analysis of gastrointestinal and pulmonary
carcinoid tumors. Hum. Pathol. 29, 992–998.
11. Gould, V. E., Lee, I., Wiedemann, B., Moll, R., Chejfec, G., and Franke, W. W. (1986)
Synaptophysin: a novel marker for neurons, certain neuroendocrine cells, and their neo-
plasms. Hum. Pathol. 17, 979–983.
12. Scopsi, L., Gullo, M., Rilke, F., Martin, S., and Steiner, D. F. (1995) Proprotein convertases
(PC1/PC3 and PC2) in normal and neoplastic human tissues: their use as markers of
neuroendocrine differentiation. J. Clin. Endocrinol. Metab. 80, 294–301.
13. Lloyd, R. V., Jin, L., Qian, X., Scheithauer, B. W., Young, W. F., Jr., and Davis, D. H.
(1995) Analysis of the chromogranin A post-translational cleavage product pancreastatin
and the prohormone convertases PC2 and PC3 in normal and neoplastic human pituitaries.
Am. J. Pathol. 146, 1188–1198.
14. Lloyd, R. V. and Warner, T. F. (1984) Immunohistochemistry of neuron-specific enolase,
in Advances in Immunochemistry (DeLellis, R. A., ed.), Masson, New York, pp. 127–140.
15. Schmechel, D., Marangos, P. J., and Brightman, M. (1978) Neurone-specific enolase is a
molecular marker for peripheral and central neuroendocrine cells. Nature 276, 834–836.
16. Wharton, J., Polak, J. M., Bloom, S. R., et al. (1978) Bombesin-like immunoreactivity in
the lung. Nature 273, 769–770.
17. Bostwick, D. G., Roth, K. A., Evans, C. J., Barchas, J. D., and Bensch, K. G. (1984)
Gastrin-releasing peptide, a mammalian analog of bombesin, is present in human neuroendo-
crine lung tumors. Am. J. Pathol. 117, 195–200.
18. Bunn, P. A., Jr., Linnoila, I., Minna, J. D., Carney, D., and Gazdar, A. F. (1985) Small
cell lung cancer, endocrine cells of the fetal bronchus, and other neuroendocrine cells
express the Leu-7 antigenic determinant present on natural killer-cells. Blood 65, 764–768.
19. Tischler, A. S., Mobtaker, H., Mann, K., et al. (1986) Anti-lymphocyte antibody Leu-7
(HNK-1) recognizes a constituent of neuroendocrine granule matrix. J. Histochem. Cyto-
chem. 34, 1213–1216.
20. Thompson, R. J., Doran, J. F., Jackson, P., Dhillon, A. P., and Rode, J. (1983) PGP 9.5—
a new marker for vertebrate neurons and neuroendocrine cells. Brain Res. 278, 224–228.
21. Rode, J., Dhillon, A. P., Doran, J. F., Jackson, P., and Thompson, R. J. (1985) PGP 9.5,
a new marker for human neuroendocrine tumours. Histopathology 9, 147–158.
22. Cunningham, B. A., Hemperly, J. J., Murray, B. A., Prediger, E. A., Brackenbury, R., and
Edelman, G. M. (1987) Neural cell adhesion molecule: structure, immunoglobulin-like
domains, cell surface modulation, and alternative RNA splicing. Science 236, 799–806.
23. Jin, L., Hemperly, J. J., and Lloyd, R. V. (1991) Expression of neural cell adhesion molecule
in normal and neoplastic human neuroendocrine tissues. Am. J. Pathol. 138, 961–969.
24. Patel, K., Moore, S. E., Dickson, G., et al. (1989) Neural cell adhesion molecule (NCAM)
is the antigen recognized by monoclonal antibodies of similar specificity in small-cell lung
carcinoma of neuroblastoma. Int. J. Cancer 44, 573–558.
25. Rygaard, K., Moller, C., Bock, E., and Spang-Thomsen, M. (1992) Expression of cadherin
and NCAM in human small cell lung cancer cell liens and xenografts. Br. J. Cancer
65, 573–577.
26. Kibbelaar, R. E., Moolenaar, K. E., Michalides, R. J., et al. (1991) Neural cell adhesion
molecule expression, neuroendocrine differentiation and prognosis in lung carcinoma. Eur.
J. Cancer 27, 431–435.
27. Eipper, B. A., Stoffers, D. A., and Mains, R. E. (1992) The biosynthesis of neuropeptides:
peptide alpha-amidation. Annu. Rev. Neurosci. 15, 57–85.
28. Grimmelikhuijzen, C. J., Leviev, I., and Carstensen, K. (1996) Peptides in the nervous
systems of cnidarians: structure, function, and biosynthesis. Int. Rev. Cytol. 167, 37–89.
29. Gether, U., Aakerlund, L., and Schwartz, T. W. (1991) Comparison of peptidyl-glycine
Diagnosis of Endocrine Lesions 373
INTRODUCTION
He would need practice if he was going to see the things above. First he would most
easily see shadows and then the images of men and other things reflected in water, and
then those things themselves, and the things in the heavens and the heavens themselves.
(Plato 428–348 BC: The Republic)
Plato could not imagine the important role that transmission electron microscopy, “the
world of shadows,” would play in the progress of science. Its crucial role in diagnostic
pathology and cell type identification is undisputed. By allowing recognition of subcellu-
lar structural features not discernible by light microscopy, electron microscopy has
contributed significantly to the better understanding of derivation and function of various
cell types in health and disease.
Although transmission electron microscopy has been largely replaced lately by
immunocytochemical and nucleic acid hybridization techniques at the light microscopic
level, the combination of ultrastructural investigation and immunocytochemistry (immu-
noelectron microscopy) plays a key role in the accurate localization of immunoreactive
substances. This method, by allowing colocalization of two different antigenic epitopes,
combines the resolution of ultrastructural features with immunocytochemical specificity.
Although immunoelectron microscopy is not the first method of choice in routine study
of tissues, it is very useful when ultrastructural and immunocytochemical findings are
not pathognomonic and are subject to diverse interpretations.
From their inception in the 1960s, light microscopic immunocytochemical techniques
underwent several modifications, and the variants are widely and successfully used.
The stringent criterion of optimal preservation of subcellular structures makes electron
microscopy much less adaptable and renders immunoelectron microscopy a somewhat
tedious technique used chiefly in research.
Preembedding immunoelectron microscopy on vibratome or cryostat sections is a
technique not much different from light microscopic immunocytochemistry and can
be applied easily. However, the inaccessibility of antigenic sites to antibodies creates
serious problems. Postembedding immunolabeling and ultrathin cryosection techniques
have solved the problem of inaccessibility. However, these methods are sometimes
375
376 Vidal, Horvath, and Kovacs
Preembedding Methods
In preembedding methods, the antibody is applied to relatively thick tissue slices
before or after any fixation procedure. Following the immunocytochemical procedure,
the tissue is fixed, postosmicated, and routinely processed for ultrastructural examina-
tion. Preembedding techniques are difficult and time-consuming procedures requiring
immediate tissue handling. For diagnostic purposes, utilization of such techniques
in laboratories involved primarily in diagnostic work is extremely cumbersome and
unrealistic. The use of preembedding immunocytochemistry remains limited to demon-
stration of antigens that cannot withstand exposure to fixatives and/or processing.
Alternatively, tissue cryofixation can be applied before postembedding labeling, thereby
maintaining antigenic properties.
The major disadvantage of preembedding immunoelectron microscopy is the poor
morphologic preservation of ultrastructural details. Another significant problem is poor
antibody penetration. In addition, a large amount of tissue is needed, since each antigenic
determinant has to be tested on individual tissue slices. The need for fresh unfixed
tissue precludes the use of preembeding techniques in retrospective studies utilizing
archival material.
These disadvantages notwithstanding, preembedding techniques are of subtantial
value, allowing rapid scan of large tissue portions on thick sections (vibratome or
cryosections) and selection of specific areas for electron microscopy. This technique
is particularly useful to find cells/tissue structures that are rare and difficult to locate.
Another notable advantage of the preembedding method is the visualization of mem-
brane-bound antigens, such as synaptophysin (1). In postembedding methods, immuno-
reactivity is limited to the availability of epitopes exposed on the surface of the ultrathin
sections. Membranes usually appear as “lines” in ultrathin sections, and the chance of
exposing membrane-bound epitopes is understandably low. In the preembedding proto-
col, antibodies can reach the antigenic sites in the tissue before embedding, and the
immunocytochemical chromogen product formed can be visualized on the ultrathin
Ultrastructural Immunohistochemistry 377
sections even if the antigenic sites themselves are not exposed on the surface of
the section.
The success of preembedding immunocytochemistry, expressed as efficient labeling
as well as good tissue preservation, depends on a number of variables such as fixation
and tissue permeability.
Fixation Procedures
Accurate identification of antigens is profoundly dependent on tissue preparation.
Rapid fixation is essential to render tissue components insoluble, in order to preserve
tissue morphology and prevent diffusion or loss of tissue antigens from their original
site. Thus the rate of penetration of the fixative will play an important role in the
precise localization of soluble antigens (2,3). If the fixative fails to reach and stabilize
the antigens before they diffuse from their natural sites, diffused antigens may subse-
quently be fixed at some other sites, resulting in erroneous localization. This is particu-
larly significant when the antigens are small and/or highly soluble such as some peptide
hormones. Thus diffuse immunocytochemical staining was observed in adrenocortico-
tropic (ACTH) cells of the anterior pituitary and in C-cells of thyroid (3,4).
The ideal fixative should completely preserve all spatial relationships of tissue constit-
uents without affecting the specific binding of antibodies to tissue antigens. However,
fixatives that achieve satisfactory preservation of cytologic details typically reduce the
number of immunoreactive sites on the antigens, in turn decreasing the number of sites
available for antibody binding. This may not prevent the visualization of antigens that are
present in high concentration, but antigens present only in trace amounts may be rendered
undetectable. The fixation process will be influenced not only by the type of fixative but
also by the type of antigen and by its state and/or concentration. For instance, hormones
may be more easily denatured while being synthesized than during the process of secretion.
Such differences could explain the frequent absence of immunoreactivity within the sac-
culi of the Golgi complex. However, it is also possible that the lack of immunopositivity
is because the antigens have relatively low concentration in the synthetic machinery and
are more vulnerable to the effects of fixatives. It has been demonstrated that in pituitary
lactotrophs, prolactin is 200 times more concentrated in the secretory granules than within
the endoplasmic reticulum (5).
Different strategies have been used to prevent the loss of antigenicity during the
immunocytochemical procedures. Excluding fixation would not solve the problem.
Antigen immunoreactivity is theoretically maximal in unfixed tissue, but so are the
loss of cytologic details and diffusion of antigens. The strategy most successfully used
is to match the fixative to the nature of the antigen. If the antigen is a polypeptide,
aldehyde fixative should be used. If the antigen is a glycoprotein, a fixative directed toward
its carbohydrate moiety would be superior to an aldehyde fixative since the antigenicity
of the peptide portion would not be significantly altered. Furthermore, the fixative should
preferably react with the antigen molecule at a site distant from the antigenic determinant.
For example, if an aldehyde fixative was used to stabilize ACTH, the anti-ACTH should
be directed against the C-terminal of the hormone rather than the midportion, since the
aldehyde is expected to attach to the lysine residues on the 11th, 15th, 16th, and 21st
portions of the molecule (6).
378 Vidal, Horvath, and Kovacs
also exert its effect by rapidly removing free aldehyde from the tissue. Thus, when the
tissue is sufficiently crosslinked after fixation to retain its ultrastructure, treatment with
50% ethanol may wash out free aldehydes. If they are left in the tissue, additional
crosslinks could be produced, hindering either antibody penetration or antibody binding.
In contrast to freezing procedures, ethanol treatment has the advantage that the severity
of tissue damage can easily be reduced by varying the length of exposure and ethanol
concentration to adapt to the characteristics of different tissues. Thus membranes appear
fragmented if the tissue is washed with too high concentrations of ethanol or with 50%
ethanol for too long.
Although it is possible to vary the degree of cryoprotection, freezing procedures are
less flexible in terms of controlling the severity of cell damage. However, freezing
procedures do have the advantage that the tissue can be sliced at any desired thickness
on a sliding microtome while still frozen. In contrast to this advantage, ethanol treatment
has been successfully used for locating structures to be examined with the electron
microscope in whole-mount preparations. This method has been applied to study periph-
eral nerves for more than a century and can be adapted to other tissues of small size
that can be stained immunocytochemically in whole mounts.
Microwave irradiation of free-floating sections during the incubation with primary
antibodies can also be considered for enhancing the penetration of antisera in vibratome
and cryostat sections (15). Using microwave technology, incubation time with the
primary antibody can be drastically shortened; at the same time the ultrastructural
details remain well preserved. This is in contrast to the moderate to severe damage of
the ultrastructure observed in tissues fixed by means of microwave irradiation. The
question of whether incubation under microwave conditions results in enhanced penetra-
tion of the antisera into the sections is still unanswered. The direct and indirect effects
of microwave irradiation could account for the acceleration of the immunocytochemical
procedure. The direct effect may be the vigorous motion of electrically charged mole-
cules present in the incubating medium and tissue components. The indirect effect, the
stimulation of the movement of the bipolar water molecules by the electromagnetic
waves, results in an increase of the temperature.
Permeabilization methods permitted increased application of immunocytochemical
preembedding procedures. Visualization of antigen-antibody binding in situ requires a
reporter system. The most frequently applied reporters in preembedding immunocyto-
chemistry for electron microscopy are enzymes, mainly horseradish peroxidase. In this
technique, peroxidase is conjugated to a secondary antibody or other immunoreagent
(e.g., protein A) that is used to detect the primary antibody directed against the antigen
of interest. Variations include labeling the primary antibody directly or labeling a
tertiary antibody. The chromogen chosen to visualize the immunocytochemical reaction
depends on the requirements of a given procedure or experiment. In this regard, for
preembedding immunoelectron the peroxidase-diaminobenzidine (DAB) procedure is
favored by many authors, mainly because this technique provides consistent results
and allows good visualization of the labeled structures (16). A significant disadvantage
of the peroxidase-DAB procedure is that the label is too large for precise antigen
localization and that the reaction product tends to diffuse and may bind nonspecific-
ally to other cell structures. Furthermore, it does not allow quantification of the
antigen.
Ultrastructural Immunohistochemistry 381
procedures, in preembedding methods the embedding medium does not influence label-
ing efficiency. However, the use of appropriate media is fundamental to obtain optimal
preservation of ultrastructure. The ideal embedding medium permits the necessary
dehydration steps, infiltration, and embedding under the mildest possible conditions
within a reasonably short time (20,21). The consistency of the embedded material
should allow semithin sectioning and possibly ultrathin sectioning for studies to be
performed at both the light and electron microscopic levels. It is sometimes also
convenient to use water-miscible polymers for conventional enzyme histochemistry.
Finally, it is desirable to use polymers of sufficient hardness without britleness for
proper sectioning of a variety of tissues. Presently both hydrophobic epoxy (Epon,
Araldite) and hydrophilic acrylic (LR white, LR gold) resins are successfully employed
for preembedding immunocytochemistry. Because of their physicochemical properties,
acrylic resins have an advantage: they can be used following both full and partial
dehydration protocols at room temperature or low temperature. Their hydrophilic proper-
ties also permit a combination of preembedding immunocytochemistry with postembed-
ding immunostaining or performance of enzyme histochemical methods using semi-
thin sections.
The detailed protocol for preembedding immunocytochemistry given at the end of
the chapter provides information on the detection of many different antigens, including
some at relatively inaccessible locations.
Postembedding Methods
Postembedding electron microscopic immunocytochemical techniques require the
fixation and embedding appropriate for the cells/tissues under study as well as the
preparation of semithin or ultrathin sections thereby exposing the antigens to reagents
and antibodies. Some advantages of this approach are immediately obvious. Penetration
of antibodies into tissue sections is not necessary. Also, localization of two different
antigens can be accomplished by staining adjacent sections from the same blocks.
Multiple-labeling immunohistochemistry may be used for demonstration of different
antigens in the same cell at the electron microscopic level. Despite the advantages,
postembedding electron microscopic immunocytochemistry carries its own limitations
in terms of labeling efficiency. Attempts have been made to amplify the end result
by using low-temperature tissue embedding in hydrophilic resins, antigen retrieval
techniques, introduction of ultrasmall gold markers, and other techniques.
Currently postembedding methods are the most frequently used procedures to study
localization of hormones, peptides, and proteins in different cells and tissues at the
electron microscopic level. In postembedding immunocytochemistry, the careful plan-
ning of a protocol is pivotal to accomplish reliable and reproducible results. The strategic
approach should consider the variables that influence labeling efficiency such as fixation,
methods available for processing and embedding, or applicability of the preferred
immunocytochemical procedure to localize antigenic sites in the tissue. In the last
decade there has been an explosion in the number of publications that have employed
the various techniques, reviewed the theoretical aspects of their applications, and exam-
ined the role of variables. Here we summarize recent advances in these fields and
present the current methodology available to yield the highest specificity and signal to
noise-ratio.
Ultrastructural Immunohistochemistry 383
When selecting a method for investigation, the aspects to consider are tissue type,
appropriate fixation, method of dehydration, type of resin, mode of resin polymerization,
and immunocytochemical detection method.
Fixation Strategies
Optimal fixation is crucial for ultrastructural immunocytochemistry. At present there
are no firm rules, and one must explore the options best for the tissues/antigens to be
studied. Tissue fixation for postembedding immunocytochemistry can follow one of two
paths: the tissue can either be stabilized by chemical crosslinking and/or precipitation,
involving the use of chemical fixatives (discussed in the previous section), or it can
be physically immobilized by using cryotechniques (discussed in the next section).
Chemical fixation is used most frequently in tissue preparation for postembedding
immunocytochemistry. What was previously mentioned for preembedding protocols is
also valid for postembedding methods. However, it is important to note that the use
of glutaraldehyde is less limited, since inaccessibility of antigenic sites to antibodies
is not an issue when using semithin and ultrathin sections. For this reason retrospective
immunocytochemical studies of many tissue antigens can be conducted on routinely
processed material. Earlier there was little enthusiasm for glutaraldehyde fixation
because of its adverse effects on antigen immunoreactivity. At present glutaraldehyde
is an integral part of tissue preparation for postembedding immunocytochemistry at the
electron microscopic level. However, one has to consider that tissue antigen sensitivity is
inversely proportional to the concentration of the glutaraldehyde and the time period
for which it is used (22). Minimizing duration of glutaraldehyde fixation without
compromising ultrastructural preservation is a constant challenge. Using a high concen-
tration of glutaraldehyde (>1%) for periods of time exceeding 15 min has been shown
to be deleterious to antigene preservation. Antigenicity is often lost as a result of
conventional postfixation with osmium tetroxide used in diagnostic electron microscopy.
Osmium tetroxide is a very strong fixative reacting with and crosslinking unsaturated
fatty acids as well as stabilizing and contrasting cellular membranes. It also cross links
protein and polypeptide chains, as well as proteins and unsaturated lipids. However,
by reacting with membranes, it slows down the fixation process, impeding its own
penetration into the cell. For these reasons it should only be used as a postfixation
step. Since omission of osmic acid results in poor membrane contrast and thereby
compromises fine structural details, several treatments of the semithin and ultrathin
sections have been developed to localize different cellular antigens in tissues postfixed
in osmium tetroxide.
Most of these treatments use oxidizing agents such as H2O2 or sodium metaperiodate,
the effects of which on immunocytochemical staining are unclear at present (23,24).
Their role is probably related to the removal of osmium by reoxidizing and solubilizing
reduced osmium molecules, thereby unmasking antigens. Alternatively, various reagents
can be added to the fixative to replace osmium postfixation and prevent its negative
effect. These compounds, such as acrolein, en bloc uranyl acetate, dimethylsulfoxide
(DMSO), or carboiimides, have been advocated to preserve membrane structure without
seriously damaging antigenicity.
Cell suspensions and bacterial and cell cultures are much easier to handle when the
appropriate method has been selected. They can be fixed briefly in low concentrations
384 Vidal, Horvath, and Kovacs
of aldehyde, for example, 0.1% neutral buffered glutaraldehyde for 10 min. Subsequently
cell suspensions can be harvested or concentrated by centrifugation or filtration. The
cells are then resuspended in agar or in a low-melting-point agarose or gelatin mixture.
After cooling, the solid mixture can be cut into small blocks, which can be processed
similarly to tissue pieces (25).
Dehydration and Resin Embedding
The type of tissue and the method of fixation are important factors in deciding which
processing method and resin system should be employed. The application of resin
embedding, introduced by Newman and Borysko in 1949 (26), led to the most spectacular
advance in electron microscopic studies.
Since embedding media are usually not soluble in water, tissue samples have to be
dehydrated in series of solutions that are miscible with the embedding medium. The
use of organic solvents can be deleterious to both ultrastructure and antigenicity of
tissues. Alcohols, in particular ethanol, are the most popular dehydrating agents, but
sometimes an intermediary (to facilitate resin infiltration) is necessary. The process of
solvent exchange should be carried out at low temperatures to minimize disruption of
protein molecular structure. Processing duration should be as short as possible to
minimize extraction of cellular components. To avoid osmotic shock, dehydration has
to be carried out in a series of solutions with ascending concentrations of the dehydrating
agent. Acetone can also be applied to dehydrate small tissue pieces. It is less used than
ethanol since it causes more protein extraction. Acetone may be advantageous in
retaining lipid components and mixing with epoxy resins.
Dehydration protocols should be compatible with the embedding procedures. Resins
for electron microscopy are subdivided into two major groups, the epoxys and the
acrylics. Most of the resins also provide excellent semithin sections for light microscopy.
The opposite is not true, since the resins used for light microscopy (methacrylates) lack
stability under the electron beam and cannot provide sufficiently thin sections (20,21).
As alternatives to methacrylate, epoxy resins have been successfully employed for
ultrastructural investigation for many years. The most frequently used types for electron
microscopic immunocytochemistry are Araldite, Epon, and Spurr. The composition of
these epoxides is the same; they consist of a mixture of base resin, hardener, and
accelerator. Flexibilizers and plasticizers may be added to facilitate sectioning. All
these epoxy resins give uniform curing, and when cured they are resistant to the heat
and static electricity generated by the electron beam. They are also hydrophobic; thus
full tissue dehydration should be carried out before resin infiltration. Alcohols and
epoxy resin do not mix; thus an intermediary compound, to assist resin infiltration, is
necessary. Propylene oxide is satisfactory for this purpose. It is of note that propylene
oxide can bind to the resin during polymerization; if too much of it remains in the
tissue, it will change the characteristics of the resin, often making it too soft for
sectioning. Although dehydration and infiltration procedures can be carried out at room
temperature, the progressive lowering of temperature proposed by Kellenberger, et al.
(27) can improve preservation of antigenicity. The high viscosity of most epoxy resins
would necessitate prolonged room-temperature infiltration; thus the only practical way
of curing epoxides to sectionable blocks is heat (60°C).
Ultrastructural Immunohistochemistry 385
Although epoxy resins are the embedding medium of choice in conventional electron
microscopy, some of their attributes (tough, hydrophobic character, need for heat
polymerization, and their stabilizing effect on three-dimensional crosslinking between
antigenic determinants and the epoxy groups) constitute major drawbacks in their
application in postembedding immunocytochemistry. The resulting loss of sensitivity
led to progressive substitution of epoxy with acrylic resins.
The only hydrophilic epoxy resin available presently is Durcupan. Because of its
poor sectioning properties, this resin is used only as a dehydration medium, obviating
the use of solvents. Unfortunately, this advantage is offset by a fixation effect of the
resin, rendering the tissue unsuitable for immunocytochemistry.
Many problems associated with the three-dimensional bonding and hydrophobia of
epoxides can be avoided by using acrylic resins. Their favorable properties led to the
development of several acrylic resins such as LR white, LR gold, Bioacryl, and the
Lowicryls K4M, HM20, K11M, and HM 23. Acrylic resins can be used following both
full and partial dehydration protocols, since they are versatile, and some are more
miscible with water or 70% organic solvent than others. Also, dehydration can be
performed at room temperature, or at low temperature following progressive decrease
of temperature or cryoprocedures. LR white is more convenient to handle at room
temperature, whereas Lowicryls are more suitable for low-temperature protocols. In
addition to their hydrophilic properties and versatility, these resins have several advan-
tages. They are also hydrophilic when polymerized; thus the sections do not require
etching to remove the resin before immunolabeling or other staining procedures. Most
of them show no nonspecific binding to immunoreagents. Some have very low viscosity,
allowing rapid infiltration into tissues; they remain fluid even at −80°C, making them
invaluable for low- and very-low-temperature embedding methods.
All acrylics can be polymerized in a variety of ways: heat, chemical catalysis, UV
light, or a cool energy source. Acrylic resins preserve antigenicity much better than
the conventional epoxy resins. For instance, conventional embedding usually preserves
antigenicity only in secretory granules, whereas acrylic resins can demonstrate a variety
of substances in various cell organelles such as the rough endoplasmic reticulum, Golgi
saccules, and secretory granules and on secretory vesicles.
The most important drawback of acrylic resins is the storage of blocks once polymer-
ization is finished. In contrast to the stability of epoxy resin blocks, acrylic resin blocks
can lose immunocytochemical reactivity with time. It seems likely that polymerization
continues at a slow rate in blocks if left at ambient temperature, particularly in hot
climates. The loss of immunoreactivity can be prevented by storing the blocks at
temperatures below −20°C or in a deep freeze. If it is not possible to store blocks at
low temperature, storage at room temperature can be conveniently accomplished by
the removal of the gelatin capsules. Exposure of the resin blocks to the inhibitory
action of atmospheric oxygen prevents further polymerization. Another storage problem
is the hygroscopic character of Lowicryl blocks, requiring storage in a dessicator.
Sectioning
Sectioning of the tissue blocks should be carried out with care, avoiding any artifact
that may interfere with the labeling process. Semithin sections, cut with glass or diamond
386 Vidal, Horvath, and Kovacs
peroxidase activity (33). However, this method can not be used on acrylic sections,
which are soluble in methanol. For such sections, gentle methods also suitable for
frozen sections or whole-cells preparations can be applied. Several chemicals can be
used, such as phenylhydrazine, sodium-azide, or cyclopropanone (34).
Resin Section Immunolabeling
Immunoperoxidase and immunogold postembedding methods are the procedures
most frequently used to localize peptides, proteins, and glycoproteins on fixed semithin
or ultrathin plastic sections. The successful application of both protocols depends in
part on correct specific blocking. The incorrect or gratuitous use of this step can actually
contribute to high background, causing serious reduction or even complete abolition
of immunolabeling. The problem of background is especially important in immunogold
methods since it occurs in every case, but its level has to be kept as low as possible.
The nonspecific absorption of the gold probe by the tissue sections could be due to
surface charges related to fixation and/or stickiness of the resin or nonspecific adhesion
of the antibody molecules.
Specific blockings are of two types. In inhibitive blocking (preblocking), the agent
is applied on the section prior to incubation with the primary or secondary antibodies.
Competitive blocking takes place when reagents containing the primary or secondary
reagents are diluted with the blocking substance. Different blocking solutions have
been used such as ovoalbumin, bovine serum albumin (when anti-goat, anti-sheep, or
anti-bovine secondary antibodies or gold conjugate are not used), fish cold water gelatin,
and skim milk. Because of their wetting capacity, in some cases Triton and Tween-20
may also improve results. Normal serum (goat, rabbit, horse, and so on) is the most
common blocking agent in immunoperoxidase protocols. However, it should be avoided
in immunogold techniques, particularly when protein A, G-gold complexes are used,
since IgG molecules present in normal serum will adhere to the section and be visualized
by the probe, causing increased background levels.
Because of their high sensitivity, the immunoperoxidase techniques, the PAP tech-
nique, and the ABC technique are widely applied in light microscopic immunocytochem-
istry using DAB as the chromogen (35,36). These multistep methods are also used for
resin section immunolabeling, mainly on semithin sections. These procedures do not sig-
nificantly differ from those used on paraffin or frozen sections at the light microscopic level
and preembedding immunocytochemistry protocols (previously described). According to
the principles of the indirect immunocytochemical approach, the labeling protocol in both
techniques involves three main incubation steps. The primary antibody is localized using
a secondary antibody linked to peroxidase or biotin. Subsequently, in the third step, com-
plexes of antiperoxidase antibody with peroxidase enzyme (PAP) or peroxidase-conju-
gated avidin (ABC) are used to localize the secondary antibody linked to peroxidase or
biotin, respectively. After DAB detection of the reaction product, resin semithin sections
are usually treated with osmium tetroxide to enhance the color of the precipitated DAB
and obtain better localization. Sodium gold chloride can also be used for the same purpose,
providing even more accurate localization. Although both immunoperoxidase techniques
have been successfully applied to etched epoxy resin semithin sections, the ABC technique
is not recommended for acrylic sections because the very large complex does not penetrate
well, leading to less sensitive, uneven surface staining.
Ultrastructural Immunohistochemistry 389
Optimal immunogold labeling depends in part on the choice of colloidal gold probe
used. There is no general agreement on whether the direct or indirect immunogold
methods give more reliable results. Direct labeling requires large amounts of highly
purified specific immunoglobulins, which are usually difficult to obtain, but they provide
a high level of resolution. On the other hand, application of the indirect technique with
an IgG-gold or protein-gold complex is easier and more versatile, since these probes
can be used with a large variety of primary antibodies. The indirect approach also
provides greater intensity of labeling.
In this regard, different procedures have been used for amplification of the signal
when the labeling is limited to a few gold particles. This may be due to low levels of
antigen in the tissue or to loss of antigenicity during tissue processing. Most of the
amplification techniques are based on multistep reactions using different proteins or
antibody-gold complex. One of the first approaches, introduced by Bendayan and Duhr
(43), is based on multistep reactions with an anti-protein A antibody and protein A
complexes. Subsequently, other amplification protocols such as the biotin-antibiotin
gold technique or the biotin-streptavidin gold technique were successfully introduced.
The amplification factor in these techniques varies according to the dilution of the
primary antibody, but it is not less than five times the original signal. Recently Mayer
and Bendayan (44) adapted the catalyzed reporter deposition approach (CARD) for
electron microscopy. This technique, originally applied in cell biology for membrane
assays, Western blotting, in situ hybridization, and light microscopic immunocytochem-
istry, achieves in the majority of studies greater sensitivity, with reduced background
and enhancement of antigen detection of 8–10,000-fold (45). The procedure is based
on the ability of horseradish peroxidase to catalyze the deposition of substrate reporter
molecules by a short half-life free radical mechanism. It permitted the direct accumula-
tion of colloidal gold or Nanogold particles on target sites by enzymatic activity of
horseradish peroxidase, thus improving the efficiency of the immunogold labeling
compared with other amplification protocols. Moreover, the introduction of Nanogold
particles allowed a combination of the amplification potential of CARD with the
ultrasmall gold probe and the silver intensification method. Bendayan and co-workers
(46) reported that the immunolabeling obtained by the latter amplification strategy
appeared specific and highly sensitive, with enhanced intensity.
The size of the gold label has to be carefully selected (47). Smaller sizes provide
higher resolution and more intense labeling, but the small gold particles are difficult
to visualize at low magnification, and their use requires examination of small areas at
very high magnifications. The use of small gold particles also yields higher background
labeling because the exposed surface areas of gold particles initiate negative charges,
which interact with cationic structures on the tissue section.
A few specific gold complexes have been developed for studies of the extracellular
matrix: collagenase-gold complexes, elastase-gold complexes, and amylase–gold com-
plexes have been applied to localize collagen, elastin, and glycogen, respectively.
Immunogold protocols have been widely used for double-labeling immunocytochem-
istry for the demonstration of different antigens at the electron microscopic level.
Several techniques and protocols are available. The immunogold protocol at the end
of the chapter was designed to localize mamosomatotrophs (bihormonal pituitary cells
Ultrastructural Immunohistochemistry 391
containing GH and PRL in their cytoplasm) in a patient with acromegaly and hyperpro-
lactinemia.
Quantitative Evaluation
Quantitative evaluation became an important part of modern immunocytochemical
studies, allowing better assessment of the results as well as comparing levels of labeling
in different cellular compartments or between control and experimental groups. In
comparative studies, all conditions, such as steps of tissue preparation and labeling
protocols, should be identical. It is of note that no correlation can be made between
levels of labeling for two different antigens since they may react differently to conditions
of tissue processing. In addition, the two different antibodies utilized may display
different affinities and different titers.
Compared with other electron microscopic methods, postembedding immunogold
protocols have a great advantage in quantitative evaluation of the immunolabeling
(47,49,50). When compared with preembedding methods, accessibility of binding sites,
diffusion of marker, or thickness of sections are of no concern in postembedding gold
techniques. In the latter, only the binding sites at the surface of the section will be
revealed, and when gold particles of uniform size are used for labeling, the binding
capacity of each particle is assumed to be the same. Although surface labeling may
represent a limitation of the technique, it becomes a major advantage for quantitative
evaluation. Indeed, regardless of their localization within the tissue, cell, or subcellular
compartment, all antigenic sites exposed by sectioning have the same probability to
be labeled, because neither the thickness of the section or the diffusion of markers
interferes with the labeling.
Since the exact number of molecules per gold particles cannot be determined, absolute
amounts of antigen cannot be measured by immunoelectron microscopy. However,
information regarding the antigen can still be obtained by relating the quantity of gold
particles to structural parameters. Densities of labeling can easily and accurately be
related to surface areas of structures and expressed in number of gold particles per
unit surface. Thus, to evaluate labeling densities in tissues, cells, or a particular cellular
compartment, the surface (Sa) of the structure studied should be estimated using an
adequate morphometric approach. Upon determination of the area, the number of gold
particles (Ni) present over this area should be counted and the labeling density (Ns)
calculated accordingly: Ns = Ni/Sa (number of gold particles per surface unit). Labeling
densities can also be evaluated per length of membrane or any linear structure and
even in relation to volume using complex stereologic equation.
Quantitative evaluation of labeling densities also has to take into account the nonspe-
cific adherence of gold label, commonly referred to as background. It should be kept
as low and even as possible. In the case of high and variable background, controls for
the specificity of gold labeling have to be performed before any quantitative study.
Another factor to be considered is that gold particles are not simply points, but that
they have a certain size (5–30 nm). If only gold particles lying exclusively within the
confines of the subject of study were counted, the labeling density could be underesti-
mated if particles were lying over the boundaries of the structure. Increasing the gold
size increases the error because large profiles have less chance of lying completely
392 Vidal, Horvath, and Kovacs
within the compartment studied. To minimize this problem, different unbiased counting
rules have been introduced.
Cryotechniques
Fixatives affect tissues by chemical crosslinking of cellular components, which can
reduce or destroy antigenic reactivity as well as by altering membrane permeability
and inducing ionic leakage. These changes can led to rearrangements in cellular structure
and subcellular organization. Rapid freezing techniques and cryofixation avoid chemical
prefixation and offer the opportunity to preserve both structures and antigenic sites that
are sensitive even to low concentrations of fixative (20,51).
Cryofixation
In cryofixation, tissue is not fixed by chemical crosslinking but by physical immuno-
bilization in ice. The advantage of rapid freezing for preserving both structure and
antigenic reactivity is offset by the size of the samples that can be frozen successfully.
Tissue specimens have to be much smaller and thinner than those required for chemical
fixation. This limitation, as well as the difficulties found in cutting ultrathin frozen
sections and mounting them on grids, made this technique impractical for use in
diagnostic pathology. Currently it is used only for research.
Fernandez-Moran (52) was the first to apply cryofixation techniques for electron
microscopy. The principle of the technique is rapid freezing in order to immunobilize
the cellular structures physically. The temperatures as well as the speed of the procedure
are major factors. The objective is to freeze the water present in the specimen in an
amorphous state, avoiding crystallization, which would disrupt the structure and intro-
duce major artifacts. The formation of vitreous ice depends on the rapidity of the
thermal exchange between the cryogen and the tissue. Several methods developed to
perform an optimal cryofixation are listed below.
Plunging
There are many descriptions of methods for cryofixation by plunging. Liquid nitrogen
itself is a very good quenchant liquid, although it is used chiefly to cool and maintain
the temperature of more efficient coolant liquids such as the hydrocarbons ethane and
propane and the halocarbons, of which freon 22 is the most efficient. Plunging methods
are most appropriate for small blocks of tissue (<1 mm thick) in which the cells of
interest are not on an accessible surface.
Spraying
This technique is applicable only to specimens in the form of emulsions, suspensions,
or solutions. For these types of specimens the method is advantageous since micro-
droplets are formed and cooled extremely rapidly when sprayed into propane or ethane.
Jetting
This technique is a variant of the plunging technique, although the thermodynamic
efficiency of the method can exceed that of plunging. The technique is only applicable
to small specimens.
Hyperbaric Cryofixation
This technique depends on the high degree of subcooling that water undergoes when
subjected to a pressure of 2000–2100 bar. The advantage of this variant is that relatively
Ultrastructural Immunohistochemistry 393
large tissue samples can be cryofixed free of visible ice crystals. Its drawback is the
extremely high pressure to which tissues are subjected, inducing changes in the structure
and composition of tissues.
Metal Mirror Freezing
This is the technique most frequently used for tissue samples. The mirror metal
block is usually made of pure copper, which is precooled by liquid nitrogen (−195°C)
or liquid helium (−260°C) under vacuum. By breaking the vaccum, the specimen
is projected onto the copper block, inducing an immediate freezing of the sample.
However, in this procedure only the face of the specimen making contact with the
metal block is frozen rapidly enough to permit the vitrification of tissue water. In
this 5–40-µm layer of the sample, ultrastructural preservation is good. Deeper in
the specimen, ice crystals are formed, increasing progressively with the thickness
of the tissue, leading to artifacts such as breakage of membranes and alterations
of the nuclear chromatin pattern.
Following cryofixation, the tissue can be processed using various routes as listed
below.
Cryosubstitution
In this process the ice in the tissue is replaced at very low temperature by an organic
solvent miscible with the embedding medium. Infiltration and embedding follows at
any temperature between 60°C and −80°C. Although Lowicryls are the most frequently
used resins, embedding in epoxides is also feasible. Cryosubstitution is favored when
dealing with fixation/dehydration sensitive structures.
Freeze-Drying
In this process the ice sublimates, leaving the tissue completely dry. Subsequently
it can be infiltrated with resin; embedding may take place again at any temperature
from 60°C to −80°C. This technique may be applied to tissue that is subsequently
embedded in epoxide resins. Lowicryls can also be used; however, they are not as
favorable for immunocytochemistry as chemical fixation and dehydration in the progres-
sive lowering of temperature technique (27).
Freeze-drying avoids the use of organic solvents, although collapse or aggregation
of tissue substances can occur during drying procedures.
Alternatively, tissues following cryofixation can be cryosectioned using cryo-
ultramicrotomy.
Cryoultramicrotomy
Cryoultramicrotomy is a rapid technique providing blocks for sectioning at 30–60
min following fixation. Cryoultramicrotomy was first introduced for immunocytochem-
istry by Tokuyasu (53). His method utilizes and relies on prefixed tissue infiltrated
with sucrose, frozen, and sectioned at low temperatures. The sections are allowed to thaw
out before immunolabeling. The advantage of the technique is that neither dehydration in
organic solvents nor plastic embedding is required. Thus both antigens and membranes
are optimally preserved. In addition, the epitopes that are otherwise masked by embed-
ding remain accessible to the probes. A disadvantage of the method is the fragility of
394 Vidal, Horvath, and Kovacs
ACKNOWLEDGMENTS
This work was supported in part by the generous donation of Mr. and Mrs. Jarislowski
and the Lloyd Carr-Harris Foundation. Dr. Sergio Vidal was supported by a research
grant from the Conselleria de Educacion (Xunta de Galicia), Spain. The authors are
indebted to Ms. Sandy Briggs, Ms. Sandy Cohen, and Mr. Mark Moreland for technical
assistance and the staff of St. Michael’s Health Sciences Library for their contribution
to the study.
REFERENCES
1. Tanner, V. A., Ploug, T., and Tao-Cheng, J. H. (1996) Subcellular localization of SV2 and
other secretory vesicle components in PC12 cells by an efficient method of preembedding
EM immunocytochemistry for cell cultures. J. Histochem. Cytochem. 44, 1481–1488.
2. Hayat, M. A. (ed.) Principles and Techniques of Electron Microscopy. Biological applica-
tions, 4th ed. Cambridge University Press, Cambridge, United Kingdom, 2000.
396 Vidal, Horvath, and Kovacs
22. Hayat, M. A. (1986) Glutaraldehyde: role in electron microscopy. Micron Microsc. Acta
17, 115–135.
23. Baskin, D. G., Erlandsen, S. L., and Parsons, J. A. (1979) Influence of hydrogen peroxidase
or alcoholic sodium hydroxide on the immunocytochemical detection of growth hormone
and prolactin after osmium fixation. J. Histochem. Cytochem. 27, 1290–1292.
24. Bendayan, M. and Zollinger, M. (1983) Ultrastructural localization of antigenic sites on
osmium-fixed tissues applying the protein A-gold technique. J. Histochem. Cytochem.
31, 101–109.
25. Vidal, S., Oliveira, M. C., Kovacs, K., Scheithauer, B. W., and Lloyd, R. V. (2000)
Immunolocalization of vascular endothelial growth factor (VEGF) in the GH3 cell line.
Cell Tissue Res. 300, 83–88.
26. Newman, J. B., Borysko, E., and Swerdlow, M. (1949) New sectioning techniques for light
and electron microscopy. Science 110, 66–68.
27. Kellenberger, E., Carlemalm, E., Villiger, W., Roth, J., and Garavito, R. M. (eds.) (1980)
Low Denaturation Embedding for Electron Microscopy of Thin Sections. Chemische Werke
Löwi Gmbh, Waldkraiburg, Germany.
28. Beauvillain, J. C., Tramu, G., and Dubois, M. P. (1975) Characterization by different
techniques of adrenocorticotropin and gonadotropin producing cells in lerot pituitary (Elio-
mys quercinus). A superimposition technique and an immunocytochemical technique. Cell
Tissue Res. 158, 301–317.
29. Lane, B. P., and Europa, D. L. (1965) Differential staining of ultrathin sections of epon-
embedded tissues for light microscopy. J. Histochem. Cytochem. 13, 579–582.
30. Erlandsen, S. L., Parsons, J. A., and Rodning, C. B. (1979) Technical parameters of
immunostaining of osmicated tissue in epoxy sections. J. Histochem. Cytochem. 27, 1286–
1289.
31. Romijn, H. J., Janszen, A. W., Pool, C. W., and Buijs, R. M. (1993) An improved immunocy-
tochemical staining method for large semi-thin plastic epon sections: application to GABA
in rat cerebral cortex. J. Histochem. Cytochem. 41, 1259–1265.
32. Vidal, S., Lombardero, M., Sanchez, P., Roman, A., and Moya, L. (1995) An easy method
for the removal of epon resin from semi-thin sections. Application of the avidin-biotin
technique. Histochem. J. 27, 204–209.
33. Hittmair, A. and Schmid, K. W. (1989) Inhibition of endogenous peroxidase for the immuno-
cytochemical demonstration of intermediate filament proteins (IFP). J. Immunol. Methods
116, 199–205.
34. Andrew, S. M. and Jasani, B. (1987) An improved method for the inhibition of endogenous
peroxidase non-deleterious to lymphocyte surface markers. Application to immunoperoxi-
dase studies on eosinophil-rich tissue preparations. Histochem. J. 19, 426–430.
35. Sternberger, L. A., Hardy, P. H., Cuculis, J. J., and Meyer, H. G. (1970) The unlabeled
antibody method of immunocytochemistry. Preparation and properties of soluble antigen-
antibody complex (horseradish peroxidase-anti-horseradish peroxidase) and its use in the
identification of spirochetes. J. Histochem. Cytochem. 18, 315–333.
36. Hsu, S. M., Raine, L., and Fanger, H. (1981) Use of avidin-biotin-peroxidase complex (ABC)
in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP)
procedures. J. Histochem. Cytochem. 29, 577–580.
37. Roth, J. (1996) The silver anniversary of gold: 25 years of the colloidal gold marker system
for immunocytochemistry and histochemistry. Histochem. Cell Biol. 106, 1–8.
38. Faulk, W. P., and Taylor, G. M. (1971) An immunocolloid method for the electron micro-
scope. Immunochemistry 8, 1081–1083.
39. Roth, J. (1983) The colloidal gold marker system for light and electron microscopic cyto-
chemistry, in Techniques in Immunocytochemistry, vol. 2 (Bullock, G. R. and Petrusz, P.,
eds.), Academic, San Diego, pp. 217–284.
398 Vidal, Horvath, and Kovacs
40. Vila-Porcile, E. and Corvol, P. (1998) Angiotensinogen, prorenin, and renin are colocalized
in the secretory granules of all glandular cells of the rat anterior pituitary: an immunoultra-
structural study. J. Histochem. Cytochem. 46, 301–311.
41. Vidal, S., Syro, L., Horvath, E., Uribe, H., and Kovacs, K. (1999) Ultrastructural and
immunoelectron microscopic study of three unusual plurihormonal pituitary adenomas.
Ultrastruct. Pathol. 23, 141–148.
42. Vidal, S., Cohen, S. M., Horvath, E., et al. (2000) Subcellular localization of leptin in non-
tumorous and adenomatous human pituitaries: an immuno-ultrastructural study.
J. Histochem. Cytochem. 48, 1147–1152.
43. Bendayan, M. and Duhr, M. A. (1986) Modification of the protein A-gold immunocytochem-
ical technique for the enhancement of its efficiency. J. Histochem. Cytochem. 34, 569–575.
44. Mayer, G. and Bendayan, M. (1999) Immunogold signal amplification: application of the
CARD approach to electron microscopy. J. Histochem. Cytochem. 47, 421–430.
45. Bobrow, M. N., Harris, T. D., Shaughnessy, K. J., and Litt, G. J. (1989) Catalyzed reporter
deposition, a novel method of signal amplification. Application to immunoassays. J. Immu-
nol. Methods 125, 279–285.
46. Mayer, G., Leone, R. D., Hainfeld, J. F., and Bendayan, M. (2000) Introduction of a
novel HRP substrate-Nanogold probe for signal amplification in immunocytochemistry.
J. Histochem. Cytochem. 48, 461–470.
47. Bendayan, M. (1995) Colloidal gold post-embedding immunocytochemistry. Prog. Histo-
chem. Cytochem. 29, 1–163.
48. Bendayan, M. (1982) Double immunocytochemical labeling applying the protein A-gold
technique. J. Histochem. Cytochem. 30, 81–85.
49. Slot, J. W., Posthuma, G., Chang, L. Y., Crapo, J. D., and Geuze, H. J. (1989) Quantitative
aspects of immunogold labeling in embedded and in nonembedded sections. Am. J. Anat.
185, 271–281.
50. Lucocq, J. (1994) Quantitation of gold labelling and antigens in immunolabelled ultrathin
sections. J. Anat. 184, 1–13.
51. Elder, H. Y. (1989) Cryofixation, in Techniques in Immunocytochemistry, vol. 4 (Bullock,
G. R. and Petrusz, P., eds.), Academic, San Diego, pp. 217–284.
52. Fernandez-Moran, H. (1960) Low-temperature preparation techniques for electron micros-
copy of biological specimens based on rapid freezing with liquid helium II. Ann. NY Acad.
Sci. 85, 689–713.
53. Tokuyasu, K. T. (1973) A technique for ultracryotomy of cell suspensions and tissues.
J. Cell Biol. 57, 551–565.
54. Fernandez-Moran, H. (1952) Application of the ultrathin freezing-sectioning technique to
the study of cell structures with the electron microscope. Ark. Fys. 4, 471–491.
55. Liou, W., Geuze, H. J., and Slot, J. W. (1996) Improving structural integrity of cryosections
for immunogold labeling. Histochem. Cell Biol. 106, 41–58.
56. Stierhof, Y. D., Schwarz, H., Durrenberg, M., Villiger, W., and Kellenberger, E. (1991)
Yield of immunolabel compared to resin sections and thawed cryosections, in Colloidal
Gold: Principles, Methods and Applications, vol. 3 (Hayat, M. A., ed.), Academic, San
Diego, pp. 87–115.
57. Takizawa, T., and Robinson, J. M. (2000) FluoroNanogold is a bifunctional immunoprobe
for correlative fluorescence and electron microscopy. J. Histochem. Cytochem. 48, 481–485.
58. Hainfeld, J. F., and Powell, R. D. (2000) New frontiers in gold labeling. J. Histochem.
Cytochem. 48, 471–480.
59. Stirling, J. W. (1993) Controls for immunogold labeling. J. Histochem. Cytochem. 41, 1869–
1870.
PHT 399
PROTOCOLS
Ultrastructural Immunohistochemistry
PREEMBEDDING IMMUNOCYTOCHEMISTRY
The following protocol has been extensively used in the central nervous system to
localize neurons expressing specific antigens exemplified here by demonstration of
tyrosine hydroxylase within dopaminergic neurons in the caudate nucleus of the medulla
oblongata of the rat.
Penetration Enhancement
After several washes in 0.1 M PB, the medulla oblongata of the rat brain is serially
sliced with a Vibratome, and free-floating sections of 100 µm are brought up to room
temperature in PB. Subsequently the sections are washed with 50% ethanol in distilled
water for 30 min to improve antibody penetration.
8. Neurons positive for tyrosine hydroxylase staining in the caudate nucleus are identified
and photographed.
9. Subsequently, a dissecting needle is used to etch a rectangle around the areas of interest
to be popped off.
10. To perform the pop-off method, a plastic capsule is filled with liquid resin until the surface
is slightly concave. The capsule is inverted, placed over the etched, marked area, and
polymerized according to the plastic used.
11. Polymerized capsules are popped off the glass slides by dipping them quickly in and out
of a Dewar containing liquid nitrogen.
12. Semithin and ultrathin sections are then cut with a ultramicrotome.
13. Ultrathin sections are mounted on grids stained with lead citrate and examined in a transmis-
sion electron microscope.
8. Sections were then washed in TBS for 5 min and treated with 1% osmium
tetroxide to increase the density of the precipitated DAB and improve localization of
DAB deposits.
Study of consecutive sections stained with GH or PRL antibodies showed immuno-
labeling for both hormones in the cytoplasm of the same cells, indicating the presence
of bihormonal mammosomatotrophs in this pituitary tumor.
To confirm the presence of bihormonal cells and assess subcellular localization of
GH and PRL in mammosomatotrophs, immunoelectron microscopy was performed
using the double immunogold method of Bendayan (48). In this protocol, it is essential
to carry out all incubations by floating. It is important that the reagents should not
reach the other face of the grid, to avoid excessive background. The incubations should
also be undertaken in a humidified chamber to prevent drying of the immunoreagents.
To localize two different antigens on the same ultrathin section by the immunogold
method, each reaction must be performed on opposite surfaces of the section. However,
when the double staining is carried out using primary antibodies raised in different
species, the two reactions to detect different antigens can be done on the same surface
of the ultrathin section. The gold-labeled secondary antibody is specific against different
species, and crosslinking does not occur. It is also important to consider that in double
immunostaining, the gold particles for the second antigen must be smaller than those
used for the first antigen.
Pretreatment of Sections
1. Before labeling, ultrathin sections are pretreated for 1 h in a saturated aqueous
solution of sodium metaperiodate.
2. Subsequently sections are washed several times in distilled water for 20 min.
Immunocytochemical Labeling
3. One side of the nickel grids are preblocked in 0.2 M PBS (pH 7.5) admixed with
0.2% cold water fish gelatin.
4. Next they are incubated at 37°C for 12 h with specific antisera directed toward
either GH or PRL.
5. Grids are washed at least four times in 0.2 M PBS (pH 7.5) admixed with 0.2%
cold water fish gelatin.
6. Subsequently, the grids are treated at 37°C for 1 h with gold-labeled, goat anti-
rabbit IgG (Biocell Research Laboratories, Cardiff, UK); particle diameters should be
either 10 or 20 nm.
7. Grids are washed again in 0.2 M PBS (pH 7.5) admixed with 0.2% cold water
fish gelatin.
For double immunostaining, the procedure is repeated on the other side of the grid
using another specific antibody and a colloidal gold conjugate of a different size. After
immunolabeling is completed, sections are stained with uranyl acetate and examined
on a Philips 410 LS electron microscope.
Electron microscopy demonstrates colocalization of GH and PRL in the same ade-
noma cells. The two different hormones can be localized within the same secretory
granules (Figs. 1 and 2).
402 Protocols
Fig. 1. Immunogold labeling in a GH- and PRL-producing human pituitary adenoma com-
posed of mammosomatotrophs demonstrating that adenoma cells with different granule sizes
are immunopositive for GH. Gold particles are 10 nm. Original magnification ×19,170.
20
Clonality Analysis
INTRODUCTION
Assessment of clonality is an important factor in differentiating reactive from neoplas-
tic lesions. In tumors with a known genetic defect, clonal expansion can be analyzed
by the identification of the genetic changes such as translocations in soft tissue tumors
or malignant lymphomas. For examining tumors without known genetic changes, a
more general approach is needed.
To compensate for the double amount of X-chromosomal genes in females, either
the maternally or paternally derived X chromosome has to be inactivated. This occurs
during embryogenesis by random methylation of one X chromosome in each cell (1,2).
This process is stable during subsequent cell divisions (3). Normal tissues in adult
females are therefore cellular mosaics differing by which of the two X chromosomes
is methylated. In contrast, neoplasms derived from a single somatic cell show a uniform
pattern of X chromosome inactivation, indicating cellular monoclonality.
The first approach toward analysis of tissue clonality by X chromosome inactivation
was based on isoenzymes of glucose-6-phosphate dehydrogenase (G6PD) (4). The low
frequency of heterozygosity (2%) for the G6PD isoenzymes and the need for fresh
frozen tissue, however, made this approach impractical. All the following approaches
take advantage of methylation-sensitive restriction enzymes that selectively cleave
nonmethylated DNA, thus permitting discrimination between the active (unmethylated)
and inactive (methylated) X chromosome.
First, Southern blotting of the restriction fragment length polymorphisms (RFLPs)
in the phosphoglycerate-kinase 1 gene (PGK-1) and the hypoxanthine phosphoribosyl
transferase (HPRT) gene was applied (5). The variable repeat sequence in the Dx255
locus identified by the M27 β probe made clonality analysis more practical because
of a high heterozygosity rate of approx 90%, but still all these Southern blotting
approaches required high quantities of DNA and hence fresh tissue. The identification
of a trinucleotide repeat (CAGn in exon 1 of the androgen receptor gene (HUMARA)
(6) with nearby methylation-sensitive restriction endonuclease sites for HpaII and HhaI
(7) made a polymerase chain reaction (PCR)-based approach possible. These enzymes
are sensitive to methylation of the cytosine residues in their recognition sequence; they
only cut the DNA strand if their recognition sequence is not methylated (8,9). Exon
403
404 Perren and Komminoth
Fig. 1. Exon 1 of the androgen receptor gene. The vicinity of the short tandem repeat
(STR) to the differentially methylated restriction sites for HhaI and HpaII permits a specific
amplification of the methylated (inactive) allele by restriction enzyme digestion prior to PCR.
Fig. 2. Flow diagram showing expected PCR results in a monoclonal (left) and polyclonal
(right) cell population. The active allele (white) of exon 1 of the androgen receptor gene is
cleaved because the enzyme restriction sites are not methylated. Thus the methylated inactive
allele is selectively amplified (hatched). The different sizes permit discrimination of the two
alleles after gel electrophoresis. (Disappearance of one band after restriction enzyme digestion
indicates monoclonality).
PROTOCOL: Clonality Analysis 405
REFERENCES
1. Lyon, M. (1961) Gene action in the X-chromosome of the mouse (Mus musculus L.).
Nature 190, 372–373.
2. Gartler, S. M. and Riggs, A. D. (1983) Mammalian X-chromosome inactivation. Annu.
Rev. Genet. 17, 155–190.
3. Tsukada, M., Wade, Y., Hamade, N, et al. (1991) Stable Lyonization of X-linked pgk-1
gene during aging in normal tissues and tumors of mice carrying Searle’s translocation. J.
Gerontol. 46, B213–B216.
4. Fialkow, P. J. (1977) Glucose-6-phosphate dehydrogenase (G-6-PD) markers in Burkitt
lymphoma and other malignancies. Haematol. Bluttransfus. 20, 297–305.
5. Vogelstein, B., Fearon, E. R., Hamilton, S. R, and Feinberg, A. P. (1985) Use of restriction
fragment length polymorphisms to determine the clonal origin of human tumors. Science
227, 642–645.
6. Tilley, W. D., Marcelli, M., Wilson, J. D., and McPhaul, M. J. (1989) Characterization and
expression of a cDNA encoding the human androgen receptor. Proc. Natl. Acad. Sci. USA
86, 327–331.
7. Allen, R. C., Zoghbi, H. Y., Moseley, H. B., et al. (1992) Methylation of HpaII and HhaI
sites near the polymorphic CAG repeat in the human androgen-receptor gene correlates
with X chromosome inactivation. Am. J. Hum. Genet. 51, 1229–1239.
8. Makula, R. A. and Meagher, R. B. (1980) A new restriction endonuclease from the anaerobic
bacterium, Desulfovibrio desulfuricans, Norway. Nucleic Acids Res. 8, 3125–3131.
9. Garfin, D. E. and Goodman, H. M. (1974) Nucleotide sequences at the cleavage sites of
two restriction endonucleases from Hemophilus parainfluenzae. Biochem. Biophys. Res.
Commun. 59, 108–116.
10. Enomoto, T., Fujita, M., Inove, M., et al. (1994) Analysis of clonality by amplification of
short tandem repeats. Carcinomas of the female reproductive tract. Diagn. Mol. Pathol.
3, 292–297.
11. Gale, R. E., Mein, C. A., and Linch, D. C. (1996) Quantification of X-chromosome inactiva-
tion patterns in haematological samples using the DNA PCR-based HUMARA assay.
Leukemia 10, 362–367.
12. Tonon, L., Bergamaschi, G., Dellavecchia, C., et al. (1998) Unbalanced X-chromosome
inactivation in haemopoietic cells from normal women. Br. J. Haematol. 102, 996–1003.
13. Uchida, T., Ohashi, H., Aoki, F., et al. (2000) Clonality analysis by methylation-specific
PCR for the human androgen-receptor gene (HUMARA-MSP). Leukemia 14, 207–212.
14. Kubota, T., Nonovarna, S., Tonoki, H., et al. (1999) A new assay for the analysis of X-
chromosome inactivation based on methylation-specific PCR. Hum. Genet. 104, 49–55.
15. Volante, M., Papotti, M., Roth, J., et al. (1999) Mixed medullary-follicular carcinoma.
(1999) Molecular evidence for a dual origin of tumor components. Am. J. Pathol.
155, 1499–1509.
PROTOCOL: Clonality Analysis 407
MATERIALS
DNA Extraction of Paraffin Tissue
1. Extraction buffer
a. 100 mM Tris buffer, pH 8.0 (Trizma-Base, Sigma, St. Louis, MO).
b. 5 mM EDTA (Triplex III, Merck, Dietikon, Switzerland).
c. 0.5% NP 40 (Nonidet + P40, BDH, Poole, UK).
d. 0.5% Tween 20 (Merck).
2. Phenol (Fluka, Buchs, Switzerland).
3. Isoamylalcohol/chloroform 1:24 (Merck).
4. 100% ethanol (Merck).
5. 70% ethanol (Merck).
6. 3 M Na-acetate, pH 5.2 (Fluka).
7. Glycogen, 20 mg/mL (Roche, Rotkreuz, Switzerland).
8. Proteinase K, 50 mg/mL (Roche).
9. Tris buffer, 10 mM, pH 8.3 (Sigma).
10. Freezer at −20°C.
11. Freezer at −80°C.
12. Centrifuge.
13. Cold room at 4°C.
14. Spectrophotometer at 260λ and 280λ.
PCR Amplification
1. Oligonucleotide primers, 20–50 pmol/µm (Table 1).
2. dNTP, 2 mM (Roche).
3. MgCl2, 25 mM (Perkin-Elmer, Rotkreuz, Switzerland)
4. PCR buffer II, 10× (without MgCl2, Perkin-Elmer).
5. Taq DNA polymerase (Boehringer Mannheim, Mannheim, Germany) (see Note 2).
408 Perren and Komminoth
Table 1
Oligonucleotide Primers
Primer Position Orientation Sequence
AR 1F 221–240 Sense 5′ GAGGAGCTTTCCAGAATCTG 3′
AR 1R 454–439 Antisense 5′ GATGGGCTTGGGGAGA 3′
Gel Electrophoresis
1. TBE 10×: 0.89 M Trizma base (Sigma), 0.89 M boric acid (Sigma), 0.02 M EDTA Tritriplex
III (Merck).
2. Acrylamide, 40% (Bio-Rad, Glattbrugg, Switzerland).
3. Bisacrylamide, 2% (Bio-Rad).
4. Urea (Merck).
5. APS, 10% (Sigma).
6. Tetramethylethylenediamine (TEMED; Sigma).
7. Blue stop loading buffer:
a. 95% formamide (Sigma).
b. 20 mM EDTA (Merck).
c. 0.05% bromphenol-blue (Sigma).
d. 0.05% xylencyanol (Sigma).
Silver Staining
1. Glass well.
2. Rain-X anti-Fog (Unelka, Belgium).
3. Saran wrap foil (Dow).
4. Ethanol, 10% (BDH).
5. HNO3, 1% (Merck).
6. AgNO3, 0.2% (Merck) (dissolve 2.02 g AgNO3 in 11 dH2O and filter).
7. Filter (Nalgene, Nalge Nuc, Rochester, NY).
8. Developer:
a. 0.28 M Na2CO3, 0.02% formaldehyde
b. 29.6 g Na2CO3 (Merck) in 800 mL dH2O
c. 540 µL formaldehyde (Merck).
d. dH2O to get 11.
9. CH3COOH, 10% (Merck).
10. EDTA, 50 mM pH 8.0 (Merck).
11. dH2O.
METHODS
DNA Extraction of Paraffin Tissue
Tissue Preparation
Identify tumor and normal tissue on hematoxylin and eosin-stained section
1. Prepare 10-µm serial sections.
2. Use 3–10 sections (according to size of the area) to scrape off the desired area and transfer
the tissue to an 1.5-mL Eppendorf tube.
PROTOCOL: Clonality Analysis 409
Proteinase K Digestion
1. Add 500 µL extraction buffer.
2. Add 5 µL proteinase K.
3. Incubate at 54°C overnight in a shaker.
4. Repeat steps 2 and 3 twice to achieve complete digestion.
5. Heat at 95°C for 10 min to inactivate proteinase K.
Phenol/Isoamylalcohol/Chloroform Purification
1. Add 500 µL phenol and vortex well.
2. Spin for 2 min at 13,100g (14,000 rpm).
3. Pipet supernatant in new Eppendorf tube (only take the aqueous phase)
4. Add 250 µL phenol and 250 µL isoamylalcohol/chloroform and vortex.
5. Spin for 2 min at 13,100g (14,000 rpm) and pipet supernatant in a new tube.
6. Add 500 µL chloroform and vortex.
7. Spin for 2 min at 13,100g (14,000 rpm) and pipet supernatant in a new tube.
DNA Precipitation
1. Add 1000 µL cold (−20°C) ethanol 100%, 50 µL Na-acetate, and 1 µL glycogen.
2. Vortex well and incubate at −20°C overnight.
3. Spin for 10 min at 14,000 rpm in a cold centrifuge (4°C).
4. Gently drain off the ethanol (DNA is visible as small pellet).
5. Add 500 µL cold ethanol 70%; vortex briefly.
6. Spin for 10 min at 13,100g (14,000 rpm) in a cold centrifuge (4°C).
7. Gently drain off the ethanol.
8. Air-dry pellet for 15 min; alternatively, Speed Vac for 5 min.
9. Add 20–50 µL Tris buffer (according to expected amounts of DNA).
10. Dissolve DNA at 65°C for 10 min.
DNA Concentration
1. Pipet 5 µL of DNA and 495 µL ddH2O in a new Eppendorf tube.
2. Measure O/D at λ260 nm and λ280 nm.
3. DNA quantity of the stock solution equals λ260 × 5 (in µg/µL).
4. Ratio of λ260:λ280 is a measure of DNA purity and should be between 1.8 and 2.
5. Prepare DNA working solutions by adding ddH2O to get a concentration of 0.1 µg/µL.
410 Perren and Komminoth
3. Add 1 µL glycogen.
4. Vortex well and precipitate at −80°C for 1 h.
5. Spin for 10 min at 13,100g (14,000 rpm) in a cold centrifuge (4°C).
6. Gently drain off the ethanol (DNA is visible as small pellet).
7. Add 55 µL cold ethanol 70%; vortex briefly.
8. Spin for 10 min at 13,100g (14,000 rpm) in a cold centrifuge (4°C).
9. Gently drain off the ethanol.
10. Air-dry pellet for 15 min; alternatively, Speed Vac for 5 min.
11. Add 10 µL Tris buffer (giving an approximate DNA concentration of 0.07 µg/µL).
12. Dissolve DNA at 65°C for 10 min.
PCR Amplification
PCR Preparation
1. Use 3 µL of digested DNA (approx 0.2 µg).
2. Add 5 µL PCR buffer 10×.
3. Add 3 µL MgCl2.
4. Add 5 µL dNTP.
5. Add 33.8 µL ddH2O.
6. Add 1 µL of each primer.
7. Add 0.2 µL of Taq polymerase.
Cycling Conditions
1. Initial denaturation and activation of Taq polymerase at 94°C for 300 s.
2. 35 cycles of 3–5 min.
3. Denaturation at 94°C for 45 s.
4. Annealing at 59°C for 45 s.
5. Extension at 72°C for 45 s.
6. Final extension at 72°C for 300 s.
Gel Electrophoresis
1. For 150 mL of gel use 15 mL of TBE 10×.
2. Add 20 mL acrylamide 40%.
3. Add 20 mL bisacrylamide 2%.
4. Add 63 g urea.
5. Add 150 mL dH2O and stir till urea is dissolved.
6. Add 1 mL ammonium persulfate (APS) 10%.
7. Add 70 µL tetraethylenediamine (TEMED).
8. Pour gel between glass plates.
9. Let polymerize at room temperature for 1 h.
10. Preheat gel at 80 W for 30 min.
11. Mix 20 µL of PCR product with 2 µL blue stop.
12. Load probes.
13. Run at 80 W for 2 h.
Silver Staining
1. Wipe glass well with Rain-X to get a hydrophobic film.
2. Pour ethanol 10% in glass well. Use enough solution to have the gel immersed at all times;
after incubation, remove the solution using a vacuum pump.
3. Transfer gel to glass well with ethanol using foil; incubate for 5 min with gentle agitation.
4. Wash 2× in dH2O for 10 min each to remove urea.
5. Incubate in HNO3 for 3 min.
412 Perren and Komminoth
6. Wash 2× in dH2O.
7. Incubate in AgNO3 for 30 min.
8. Wash 2× in dH2O.
9. Rinse with developer; immediately remove precipitate.
10. Add developer again and incubate until the bands are visible.
11. Incubate in CH3COOH 10% for 10 min to stop development.
12. Incubate in EDTA for 5 min.
13. Wash 2× in dH2O.
14. The stained gel can be archived in sealed plastic foil.
NOTES
1. HhaI and CfoI are isoschizomeres. In our settings, the most consistent results were obtained
with HpaII alone compared with HhaI and a mixture of both enzymes.
2. The use of a Taq polymerase of a different company requires adaptation of annealing tempera-
tures.
3. Use alcoholic eosin; exposure to acid will cause DNA damage.
4. For analysis of very small tissue areas, laser-based microdissection can be used (PALM Laser-
Microbeam Systems, Germany) with modified protocols for DNA extraction and restriction
enzyme digestion (15).
Index
413
414 Index
Confocal laser scanning microscopy (CLSM) Electron microscopy in situ hybridization (EM-
(cont.), ISH) (cont.),
organelle visualization, 168 preembedding method (cont.),
overview, 165, 166 light microscopy, 130, 131
probe detection, 167, 168 prolactin transcripts, 135–137
reflectance confocal images, 167 protease treatment, 130
transmittance images, 167 EM-ISH, see Electron microscopy in situ hybrid-
Cryofixation, see Immunoelectron microscopy ization
Cyclin D1, lymphoma expression and immunohis- EMA, see Epithelial membrane antigen
tochemistry, 289, 290 Endocrine cell immunohistochemistry,
Cytokeratin, tumor immunohistochemistry, 324– broad-spectrum neuroendocrine markers,
326, 369 bombesin, 366
Cytomegalovirus (CMV), in situ hybridization, 36, chromogranin A, 363
49 chromogranin B, 363, 364
gastrin-releasing peptide, 366
D Leu7, 366
DiGeorge syndrome, fluorescent in situ hybridiza- neural cell adhesion molecule, 366
tion, 74, 98 neuroendocrine-specific protein reticulons,
DNA, 367
hybridization analysis principles, 5, 6 neuron-specific enolase, 364, 366
laser capture microdissection sample extraction, peptidylglycine a-amidating monooxygenase,
24, 25 366
structure, 1 PGP9.5, 366
DNA microarray, proconvertases, 364
laser capture microdissection, 19, 20 Rab3A, 367
principles of analysis, 8, 9 SNAP-25, 367
synaptophysin, 364
E neuroendocrine cells,
EBV, see Epstein-Barr virus dispersed neuroendocrine system, 361
Electron microscopy, see Electron microscopy in situ markers, overview, 361–363, 367, 369
hybridization; Immunoelectron microscopy versus nonneuroendocrine cells, 361
Electron microscopy in situ hybridization (EM- nonneuroendocrine markers, 369
ISH), Epithelial membrane antigen (EMA), tumor
approaches, overview, 129 immunohistochemistry, 327–329, 345,
biotinylated probe preparation, 130 347
nonembedding method using ultrathin frozen Epstein-Barr virus (EBV), in situ hybridization,
sections, 134 63–65
postembedding method, overview, 35, 36, 49
advantages and limitations, 133, 134 ERP, see Estrogen receptor protein
electron microscopy, 133 Estrogen receptor protein (ERP), tumor immuno-
fixation, 132 histochemistry, 334
hybridization and washes, 132, 133 Ewing’s sarcoma, fluorescent in situ hybridiza-
preembedding method, tion, 98, 99
advantages and limitations, 132
F
electron microscopy, 131
fixation, 130 α-Fetoprotein (AFP), tumor immunohistochemis-
growth hormone transcripts, 131 try, 331
hybridization and washes, 130 FISH, see Fluorescent in situ hybridization
immunohistochemistry combination, Fluorescent in situ hybridization (FISH),
advantages and limitations, 138, 141, 142 advantages, 67
growth hormone, 138 controls, 80, 84
LR White resin utilization, 139–141 fixation effects, 81, 92
sectioning, 137 fluorescence microscopy, 92, 93
staining, 137 historical perspective, 67, 68, 91
416 Index
Fluorescent in situ hybridization (FISH) (cont.), Fluorescent in situ hybridization (FISH) (cont.),
interphase analysis, probes (cont.),
cytology specimens, telomere probes, 73
counterstaining, 110 types, overview, 68, 69
cytologic specimens, 108 unique-sequence probes,
denaturation, 109 definition, 73
detection, 109, 110 malignancies, 75–77, 96
fine needle aspirates, 108 microdeletion syndromes, 74, 75
hybridization, 109 nomenclature, 77
nuclei disaggregation, 108 protease treatment, 92
pretreatment, 108 tyramide signal amplification, see Tyramide
probe preparation, 109 signal amplification
touch preparations, 108 Follicle-stimulating hormone (FSH), immunohis-
washes, 109 tochemistry and tyramide signal amplifica-
formalin-fixed paraffin-embedded tissues, tion, 270
110, 111 FSH, see Follicle-stimulating hormone
overview, 91, 92, 94, 96, 97
malignancy analysis, G
denaturation, 84, 86, 88 Gastrin-releasing peptide (GRP), neuroendocrine
detection, 85, 87–89 cell immunohistochemistry, 366
dewaxing, 84, 85 GCDFP-15, see Gross cystic disease fluid pro-
fixation, 84, 85 tein-15
hybridization, 85–88 Gene, number in humans, 1
materials, 84, 85 GFAP, see Glial fibrillary acidic protein
protease treatment, 84, 86 GH, see Growth hormone
washes, 85, 87, 88 Glial fibrillary acidic protein (GFAP), tumor
metaphase FISH, immunohistochemistry, 326, 327
overview, 79, 91, 94, 96 Glucose-6-phosphate dehydrogenase, isoforms
solid tumors, and clonality analysis, 403
culture, 107 Gold labeling, see Immunoelectron microscopy;
harvesting of cells, 107 Tyramide signal amplification
materials, 106, 107 Gross cystic disease fluid protein-15 (GCDFP-
specimen, 107 15), tumor immunohistochemistry, 334,
whole blood culture, 349
blood collection and culture, 103 Growth hormone (GH),
chromosome spread preparation, 104, 105 double labeling with prolactin using
harvesting of cells, 103, 104 immunogold,
materials, 102–104 labeling, 400, 401
tips, 105, 106 pretreatment of sections, 400, 401
nuclear counterstaining, 99 specimen preparation, 400
prenatal diagnosis, 96, 97 electron microscopy in situ hybridization of
principles, 79, 80, 91 transcripts, 131, 138, 140
probes, microvessel immunohistochemistry, 174, 178
painting probes, GRP, see Gastrin-releasing peptide
chromosomal translocation identification, 79
historical perspective, 77, 78 H
indirect versus direct detection, 78, 79 HCV, see Hepatitis C virus
metaphase FISH, 79, 94, 96 Helicobacter pylori, in situ hybridization, 38, 51
nomenclature, 79 Hepatitis C virus (HCV), in situ hybridization, 36
spectral karyotyping, 79 Herpes simplex virus (HSV), in situ hybridiza-
α-satellite probes, tion, 49, 50
chromosome identification, 71, 72 HHF-35, tumor immunohistochemistry, 342
nomenclature, 72, 73 HHV, see Human herpesvirus
selection, 80 Histiocytosis, immunohistochemistry, 307, 308
Index 417