Theory and Strategy in Histochemistry
Theory and Strategy in Histochemistry
Theory and Strategy in Histochemistry
With 74 Figures
Springer-Verlag
Berlin Heidelberg NewYork
London Paris Tokyo
HongKong Barcelona Budapest
Dr. Hans Lyon
Kebenhavns Kommunes Hvidovre Hospital
University of Copenhagen
Department of Pathology 134
DK-2650 Hvidovre
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned,
specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on
microfilms or in other ways, and storage in data banks. Duplication of this publication or parts thereof is only
permiued under the provisions of the German Copyright Law of September 9, 1965, in its current version, and a
copyright fee must always be paid.
© Springer-Verlag Berlin Heidelberg 1991
Softcover reprint ofthe hardcover Ist edition 1991
The use of registered names, trademarks, etc, in this publication does not imply, even in the absence of a specific
statement, that such names are exempt from the relevant protective laws and regulations and therefore free for
general use.
Product liability: The publisher can give no guarantee for information about drug dosage and applicatlOn thereof
contained in this book. In every individual case the respective user must check its accuracy by consulting other
pharmaceutical literature.
Data conversion by Danny L. Lewis, Buchproduktion Berlin with TEX
27/3020-543210-Pnnted on acid-free paper
Preface
If you want practical information on how to use this book please refer to "Note
to the Readers" p. VII.
Histochemistry and cytochemistry are essential tools in biomedical research
and routine service laboratories.
Most texts on histochemistry fall into one of two categories:
1. Encyclopaedic texts covering all or nearly all information available on the
whole or selected parts of histochemistry.
2. Reviews or surveys of methods found to be useful by the author(s).
While the former category often appeals to the more philosophically inclined
reader, direct guidance on the selection of technique may be difficult to find. In
contrast, the latter category are often excellent sources for details on how to
perform a particular method with a reasonable chance of success. Consideration
of the exact mechanism of staining, of possible reasons for failure, and of
alternative techniques are, however, frequently lacking.
This book is an introduction to the scientific basis of histochemistry and is
intended to provide a background for the selection and development of appro-
priate methods. It is not a "cook book" and readers expecting exhaustive
methodological descriptions will be disappointed.
Although most ofthe contributors to this book would not describe themselves
as histochemists, they have all at some time found it essential to develop a basic
understanding of histochemistry. This book contains the information they
would have greatly appreciated ready access to at that time.
It is our intention that this book should be useful to:
1. Scientists who would like to develop their understanding of histochemical
techniques.
2. Laboratories where histochemical methods are used (frequently or infrequen-
tly) in which it should provide assistance in the selection of appropriate
techniques.
The dedicated academic may read this book for reasons that are self evident.
However, other potential readers may be asking one or more of the following
questions:
1. Can the investigation or task I am undertaking be assisted by the application
of histochemical techniques?
VI Preface
The purpose of this note is to assist the readers in getting what they want out of
this book as quickly as possible.
We strongly recommend that all readers should familiarize themselves with
the overall organization of the book as set out in the initial Table of Contents
(p. XI).
The sequence of chapters presented reflects one kind of logical progression
through the intricacies of histochemistry. We do not necessarily recommend
that the book should be read in this order. The approach chosen by an individual
reader will depend on their specific interests. Chapters 2, 3, 31, and 32 may be
considered as appendices to the rest of the text; they will not generally be read in
their entirety but should be consulted as required.
Areas of interest may further be found by consulting one of the indexes. These
are: a. General index; b. Index of constituents induding chemical groups, cells,
and tissues detectable by histochemistry; c. Index of dyes induding chromogenic
reagents, pigments, and stains; d. Index of methods, histochemical and histo-
logical. Note that all references are made to Section numbers and not to pages.
Below we have outlined how different readers may wish to approach the
book.
t General Approach
r----
I
I
I
I
I
Additional
Background possibilities
Part" and 111
VIII Note to the Readers
Look through
and raad sa-
Raviaw Raad laetlons 01 Raad
Chaptars
4.10 and 23
t
Look through
I Chaptar 1 I
4 Laboratory Technician/Technologist/Scientific Officer
Raad
1
I
I I
I
L.._ Raad _.J Rlad
1 Chaptar 41 Part VII
5 Pathologist
a) Should I Use Another Technique?
cessary
IParts 11 & 11'1
Acknowledgements
The idea for this book is based on a Danish text which was published in 1985 by
DSR, Copenhagen, Denmark.
I am indebted to Professor Dietrich H. Wittekind, University of Freiburg,
who not only has introduced me to the field of standardization of dyes, stains,
and staining methods but who has also consistentiy encouraged me to prepare
the present book.
My thanks are due to all my coauthors for their never failing cooperation and
interest in preparing this book.
I gratefully acknowledge the help of Erik Hasselager, Royal Veterinary and
Agricultural University, Copenhagen, Denmark, in organizing the text on a
personal computer and for supervising and improving the lay-out of tables and
figures. I am very much indebted to Palle Jakobsen, Ferrosan Ltd., Denmark,
for correcting the formulae and equations. My thanks are also due to Erik
Schulte, University of Munich who undertook the task of reading the pre-
liminary drafts of all chapters and not only correcting evident mistakes but also
of supplying constructive suggestions for amendments and improvements.
Michael R. Barer has in addition to his valuable contributions to individual
chapters carried the burden of correcting the English language throughout for
which I am truly grateful.
I wish to thank the Department of Pathology, Hvidovre Hospital, University
of Copenhagen and all my colleagues here, past and present, for never failing
support and encouragement. I am especially grateful to Professor Per Christof-
fersen for giving me so excellent working conditions.
The painstaking work of the photographers Susanne 0stergaard, Hvidovre
Hospital, and Keld Ottosen, The Panum Institute, in preparing the many figures
and formulae is gratefully acknowledged.
I am most grateful to my secretary, Ulla Evald, for her never failing interest in
converting what at times has been loose thoughts and practically illegible notes
into a readable manuscript and who has never complained of having to rewrite
the same sections time and time again.
Finally, my thanks are due to the staff of Springer-Verlag, for their encourage-
ment and patience during the preparation of this book.
Hans Lyon
Table of Contents
Michael R. Barer
Department of Microbiology, University of Newcastle upon Tyne,
Newcastle upon Tyne, England
Bo van Deurs
Department of Medical Anatomy A, The Panum Institute,
University of Copenhagen, Denmark
Erik Hasselager
Department of Anatomy and Physiology, Royal Veterinary and Agricultural
University, Copenhagen, Denmark
Astrid K. N. Iversen
Institute of Medical Microbiology, University of Copenhagen, Denmark
Palle lakobsen
Ferrosan Ltd., Copenhagen, Denmark
Lars Kayser
Department of Medical Anatomy A, The Panum Institute,
University of Copenhagen, Denmark
Hans Lyon
Department of Pathology, Hvidovre Hospital, University of Copenhagen,
Denmark
Morten M011er
Department of Medical Anatomy B, The Pan um Institute,
University of Copenhagen, Denmark
Poul Prent0
Institute of Cell Biology and Anatomy, University of Copenhagen, Denmark
Erik Schulte
Department of Anatomy, University of Munich, FRG
Jakob Visfeldt
Department of Pathology, Rigshospitalet, University of Copenhagen,
Denmark
Mogens Vyberg
Department of Pathology, Aalborg Sygehus, Denmark
Dietrich H. Wittekind
Department of Anatomy 11, Albert Ludwig University,
Freiburg im Breisgau, FRG
Part 1
General Considerations
1 The Scope of Histochemistry
Aim of Histological Methods. The chief aim of these methods is to visualize and
differentiate between tissue components, not to determine the chemical composi-
tion.
A division between qualitative and quantitative histochemistry is widely recog-
nized.
1. Qualitative histochemistry is concemed with the occurrence and localization of
histochemically demonstrable components.
2. Quantitative histochemistry also assesses the amount of the individual chemical
components.
This book is mainly concemed with qualitative histochemistry. Distinetion is
sometimes made between histochemistry and cytochemistry, depending on whether
tissues or individual cells are being examined.
H. Lyon (Ed.)
Theory and Strategy in Histochemistry
© Springer Verlag 1991
4 H. Lyon, M.R. Barer
In principle all reactions that ultimately produce coloured products on tissue sec-
tions may be considered as histochemical. In such areaction it is necessary to
understand both the chemical and the histological aspects of the method. The
chemical aspects comprise adescription of the mechanisms involved, the speci-
ficity or selectivity, and the sensitivity. The histological aspects require that the
precision with which localization is achieved should be assessed. For areaction
to be designated as histochemical, an account of both chemical and histological
aspects must be given. We have adopted the following headings throughout this
book.
1.2.1 Mechanism
The following sub-headings have been used: reactive groups, reagents, reaction
products, reaction equation, and sources of error.
1.2.2 Selectivity
1.2.4 Localization
This is the central histological consideration which refers to the location of reaction
products. Changes in the location of the chemical compound under examination
may take place during fixation, dehydration, embedding, sectioning, and during or
after the histochemical reaction itself. The compound, one wishes to demonstrate,
the intermediate products during the histochemical reaction or the final product may
all diffuse. This may result in the compound or reaction product being completely
removed from the tissue, diffusely deposited throughout or even specifically bound
to other areas away from its site of origin.
According to Grimelius (1968), an assessment of the localization of the reaction
product may be made by:
1. The use 01 consecutive seetions. The histochemical reaction is performed on one
of two adjacent thin sections (1-2fLm). The second section is then stained using
a general oversight method (Sect.31.2) or a second histochemical procedure
whose reaction product distribution is weIl established.
2. Double-staining technique. Two different histochemical reactions are performed
consecutively on the same section. It is essential that the two methods used do
not affect each other qualitatively or quantitatively. The assessment is made by
taking a photograph after the first reaction has been performed and comparing
the result with that obtained after the second. If possible, the test should be
repeated with the reactions in reverse order.
3. Restaining technique. A photograph is taken after the first reaction as above.
The section is then destained, a second reaction performed, and the results
compared Interactions between the two reactions should be checked for as
with the double staining technique. Comparison is facilitated if thin sections
are used. This is a valuable method as it is often possible to perform three or
more reactions consecutively with destainings inserted between. Unfortunately,
the approach is limited by the almost unavoidable damage to the section during
destaining.
4. Differential count technique. In this method the number of cells demonstrated
by different reactions is compared.
6 H. Lyon, M.R. Barer
1.2.5 Controls
For the majority of histochemical methods, control reactions are required in order
to discriminate between true positive reactions and non-specific reactions. The
following control reactions are usually necessary:
1. A positive control reaction performed on a section from another block of the
same tissue or another tissue containing the chemical substance under consid-
eration. The substance should be at a concentration close to the relative limit
of detection (eliminates false negative reaction).
2. A negative control reaction performed on a positive control tissue block, a
consecutive section, or both. Here the control sections are subjected to the
same treatment as the test section but a key reagent in the reaction sequence is
omitted (eliminates a false positive reaction).
3. Blocking and/or extraction procedures. These assist in determining the selec-
tivity or specificity of the his.ochemical reaction.
2 The Structural and Chemical Basis for
Histochemistry
EpC
Fig. 2.1. Schematic model for the general organization of a multicellular animal. The epithelial
cells (EpC) define the external limits of the organism and rest on a basement membrane (BM),
and below this the extracellular matrix (M) is found with various fibre proteins embedded in
an amorphous ground substance. Mesenchymal or stromal (MC) cells are also present in the
matrix. In addition, blood and lymph vessels (BIL) are present, separated from the matrix by the
endothelium (EnC) and a basement membrane. A blood cell (Be) is shown in the vessel.
H. Lyon (Ed.)
Theory and Strategy in HiSlOchemistry
@ Springer Verlag 1991
8 H. Lyon. B. van Deurs. P. Pren~. E. Hasselager. E. Schulte
Fig. 2.2. Schematic drawing of a typical mammalian cell with a nucleus (NU). cell organelles. and
a centre for the synthesis of glyco- and lipoproteins (granular endoplasmatic reticulwn (GER);
smooth endopiasmatic reticulwn (SER); Golgi complex (GO); secretory granules (SG». elements
of the lysosomal system (endocytotic vacuole (EV); primary lysosome (LYl); secondary lysosome
(LY2» . peroxisomes (P). and mitochondria (M).
Although proteins, nucleic acids, carbohydrates, and lipids may occur as pure
substances in tissues, they are more frequently found either as molecular com-
plexes or mixed compounds (Table 2.2). For instance, nucleic acids + proteins =
nucleoproteins.
nucleic acids
-C
NUCLEOPROTEINS~
GLYCOPROTEINS ~ I
...-,- - - carbohydrates proteins
GL YCOLIPIDS PROTEOGL YCANS I I
---lipids------LIPOPROTEINS------'.
1-1
2.1.1 Water
Most of the cell is composed of water. It functions as a medium for diffusion and as
a solvent for the reactive molecules and ions in the cell (Sect.3.2.1). Physiological
processes take place predominantly in the aqueous environment, and water takes
part directly in many enzymatic reactions (Chap.24). Up to 5% of the water is
"bound" more or less strongly to the other components of the cell, particularly
proteins.
Water is an integral part of the structural organization of cells. The formation
and maintenance of cell membranes and protein conformation are mainly due to in-
teractions between water and both polar and non-polar residues in macromolecules.
This leads to aseparation with polar residues remaining solvated in the aqueous
phase, and non-polar residues of the protein or membrane becoming buried in
the inner, non-aqueous phase. This process is called "hydrophobic stabilization"
(Sect.4.S.5).
2.1.3 Proteins
Proteins are high molecular weight compounds (macromolecules) which split into
amino acids on hydrolysis. Distinction is made between the primary structure
(sequence of amino acids), the secondary structure (geometrical arrangement of
the polypeptide chain, e.g. a-helix, ß-pleated sheet (Sect.21.7) or tripie helix
(Sect2.4.1), tertiary structure (three-dimensional structure), and quaternary struc-
10 H. Lyon, B. van Deurs, P. Pren~, E. Hasselager, E. Schulte
With the exception of the terminal amino and carboxyl groups, the reactive
groups in proteins are represented by the side groups. A large proportion of the
side groups are aliphatic hydrocarbons whieh can take part in the formation of
hydrophobie bonds (e.g. with lipids). In the aqueous phase, these groups will tend to
orientate themselves towards the inner part of the molecule thereby reducing contact
with water to a minimum (cf. Fig. 12.1). In contrast, hydrophilie amino acid radicals
are usually found on the outer surface of the protein. At neutral pH both positively
(arginyl, lysyl, histidyl) and negatively (asparagyl, glutamyl) charged groups may
be present. Further bonding possibilities include: hydrophobie interactions (tyrosyl,
tryptophanyl, phenylalanyl), hydrogen bonding (tyrosyl, tryptophanyl, thiol), and
covalent bonds (thiol).
2.1.4 Lipids
Lipids are defined as all naturally occurring fat, oil, and wax-like substances, which
are soluble in chloroform, but insoluble, or colloid-forming, in water.
Hydrophobie lipids exist in the lipid phase only and give rise to a high sur-
face tension lipid-water boundary. Storage lipids are thus homophasie and sharply
delimited from the surroundings.
Lipids are classified as hydrophilie when, for example, one of the fatty acids in
a triglyceride is replaced with a hydrophilic molecule. Such molecules of course
still retain both hydrophobie and hydrophilic regions. This gives rise to a polarity
which confers the properties of low surface tension, tendency 10 form micelles
in water, and the possibility of organization into the double layers characteristic
of biological membranes, hence the term struetural lipids. In biomembranes the
hydrophilic lipids associate with neighbouring molecules and form a complex c!Üled
heterophasie lipid.
Occurrence. Some cell types (fat cells, sebaceous gland ceIls, cells in the adrenal
cortex, and other steroid producing cells, Schwann cells) contain relatively large
amounts of lipid. In consequence these tend to be weIl characterized in terms of
composition and function.
Other cell types can show marked ftuctuations in the amount of histochemi-
cally demonstrable lipid, either synchronous with natural changes in the functional
condition of the cells or as a result of pathological changes (e.g. epithelial cells of
the lactating mammary gland and the appearance of larger amounts of triglyceride
droplets in liver cells in mammals).
Even in cells where lipid is not normally demonstrable using light microscopy,
the amount of lipid is considerable (up 10 10-20% of the dry weight). Most of this
lipid (predominantly phospholipids and cholesterol) is located in the membrane
systems of the cello The rest (including triglycerides) is found dispersively bound
to cellular proteins and can become histochemically demonstrable as a result of
pathological changes. Sometimes it may be rendered visible by a histochemical
demasking process.
2.1.5 Carbohydrates
Monosaccharides. These are carbohydrates which cannot be split into lower molec-
ular weight carbohydrates by simple hydrolysis. They are either aldoses, i.e. contain
an aldehyde group, or ketoses, i.e. contain a ketone group. They are designated by
the number of carbon atoms as trioses, tetroses, pentoses, hexoses, and heptoses.
The most important monosaccharides are glucose, mannose, galactose, xylose, ara-
binose, fucose, ribose, and deoxyribose.
12 H. Lyon, B. van Deurs, P. Pren~, E. Hasselager, E. Schulte
Derivatives of monosaccharides.
a. Uronic acids are derived from the simple monosaccharides by substituting the
primary hydroxyl group with a carboxyl group. Examples are glucuronic acid,
mannuronic acid, and iduronic acid.
b. Deoxymonosaccharides are derived by substituting a hydroxyl group with hy-
drogen. An example is fucose.
c. Aminomonosaccharides are derived by substituting a hydroxyl group with an
amino group. Examples are glucosamine and galactosamine.
d. Acetylaminomonosaccharides are derived from aminomonosaccharides by ester
formation between acetic acid and the amino group. An example is N-acetyl
glucosamine.
e. Sulphate esters 01 monosaccharides are derived by forming sulphate esters be-
tween sulphuric acid and either a hydroxyl group or an amino group in the
monosaccharide.
f. Sialic acids are a group of naturally occurring N- and O-acyl derivatives of
neuraminic acid. Neuraminic acid is a carbohydrate derivative consisting of
nine carbon atoms in addition to an amino group and a carboxyl group. The
formula for N-acetylneuraminic acid is shown below.
CH 2 0H
91
HO-C-H
81
HO-C-H
~
OCOO
OH
HN OH
I
COCH 3
Table 2.5. Characteristics of glycoproteins and proteoglycans (modified from Pedersen, 1982,
p.251).
Endogenic
1. Haematogenic
a. haem
b. porphyrins
c. haemosiderin
d. acid haematins
e. bile pigments
f. Dubin-lohnson pigment
2. Lipofuscins
3. Melanins
4. Monoamine derived pigments
11 Exogenic
1. Carotenes
2. Carbon
3. Certain metal compounds (see Chapters 17 and 18)
The eukaryotic cell consists of nucleus and cytoplasm and is delimited by a plasma
membrane.
The nucleus consists of an apparently amorphous nucleoplasm or karyolymph
which contains the chromosomes and generally one or more nucleoli. The chro-
mosomes are usually to a high degree despiralized to an apparently unorganized
jumble of chromatin fibres.
The cytoplasm contains all the cell organelles (mitochondria, endoplasmic retic-
ulum, Golgi complex, lysosomes, peroxisomes, etc.). These are embedded in the
cytoplasmic matrix which in addition contains formed elements such as actin fila-
ments, microtubules, etc. Lipid droplets and glycogen granules may also be present.
The matrix contains a very large number of different soluble proteins including the
enzymes of glycolysis and most of the enzymes and other proteins which take part
in protein synthesis. The soluble phase of the cytoplasmic matrix, the cytosol, func-
tions as a medium of diffusion or transport for low-molecular soluble substances
in the cello
16 H. Lyon, B. van Deurs, P. Prentß, E. HasseJager, E. Schulte
In eukaryotic cells (i.e. all cells except bacteria and blue-green algae which are
prokaryotes) the genetic material, deoxyribonucleic acid (DNA), is present in a
membrane-bound nucleus in the form of chromosomes. One or more nucleoli are
present in the nucleus. The nucleolus is essentially a factory for ribosomes, which
are later released to the cytoplasm. The nucleus is delimited by the nuclear envelope
which consists of an outer and an inner membrane (Sect.2.2.3) separated by the
perinuclear space. The outer and inner membrane are connected by a number of
nuclear pores where the exchange of materials between cytoplasm and nucleoplasm
takes place.
2.2.2 Ribosomes
Protein synthesis takes place on ribosomes. t-RNA molecules with their associated
amino acids recognize complementary nucleotide triplets on m-RNA bound to the
ribosome. This process collects amino acids in the correct sequence on the m-
RNA. Ribosomes contain the synthetic machinery that catalyzes formation of the
polypeptide chain.
Ribosomes are approximately 20 nm in diameter and consist of roughly equal
amounts of RNA and protein. Synthesis of proteins ean take place on free ey-
toplasmic ribosomes e.g. haemoglobin synthesis in erythroblasts, or on riboso~es
situated on the cytoplasmic surface of the endoplasmic reticulum (rough or granular
ER, cf. Sect.2.2.4). The polypeptide component of all lysosom al enzymes, secre-
tory produets (glandular secretions and certain hormones) as well as most integral
membrane proteins (see below) are all formed in the rough ER.
Cells that are very active in protein synthesis have a high content of ribosomes
and t-RNA. The nuclear and nucleolar size and the amount of cytoplasmic RNA
are also important indices of protein synthesis. The relationships between these
indices and the amount of stored protein product may indicate whether a cell is
preparing for, actively engaged in, or leaving the synthetic phase of a secretory
cycle.
The membranes of all eukaryotic cells have the same basic organization regardless
of whether they are plasma membranes (plasmalemma, the membrane surrounding
the cell), or membranes of intracellular compartments such as the ER, lysosomes,
and mitochondria.
A membrane consists of a continuous ~9 nm thiek double layer 01 lipid
moleeules orientated with their polar, hydrophilic ends facing outwards. The in-
terior of the membrane is therefore hydrophobic. The lipids are chiefly phospho-
glycerides, but also include phosphosphingosides, glycolipids, and sterols, such
as cholesterol. While the pattern remains constant for individual membrane types
within one species, the relative lipid composition often shows variation between
membrane types within species and for the same membrane type between species.
Generally, sphingo- and glycolipids and sterols are abundant in plasma membranes
(cf. myelin).
Peripheral and integral pro teins are found in intimate association with the lipid
bilayer. The peripheral proteins form no direet links with the hydrophobie part of
the membrane, while the integral proteins penetrate this region to a greater or lesser
degree. Integral proteins contain one or more non-polar (hydrophobie) amino acid
sequences whieh permit this association to take place.
Membrane proteins of the plasma membrane, and to a lesser extent those asso-
ciated with other cellular membranes, are glycosylated, i.e. they are linked to short
side chains of sugars. The carbohydrate portion is outside the membrane proper,
18 H. Lyon, B. van Deurs, P. Prenll1l, E. Hasselager, E. Schulte
and for the plasma membrane, where it fonns the cell coat (or glycocalyx), the
carbohydrate chains most often tenninate with galactose or sialic acid. The plasma
membrane is a selectively penneable barrier and the integralproteins playa central
role in this function in the fonn of transport systems such as the Na/K-ATPase (the
"Na/K pump"). In addition, membrane proteins are important as antigenic sites
(e.g. blood and HLA type detenninants) and receptoTS.
Proteins destined for secretion, for the plasma membrane, for lysosomes, or for
the Golgi complex itself, all are delivered to the Golgi complex from the ER. The
Golgi complex appears in the electron microscope as a stack of 4-8 cisternae. In
the light microscope the complex may sometimes be demonstrated by reactions for
2 The Structura1 and Chemical Basis for Histochemistry 19
2.2.6 Peroxisomes
These organelles vary in size and shape and contain the enzyme catalase (the
marker enzyme far cytochemical demonstration of peroxisomes) and one or more
H2 0 2 -producing oxidases. The function(s) of the peroxisomes is still somewhat
unclear, but they are probably involved in lipid metabolism and, like the smooth
ER, detoxification (e.g. oxidation of ethanol to acetaldehyde).
2.2.7 Mitochondria
Mitochondria are found in all eukaryotic cells that can utilize 02, and in most cells
they are the main site of ATP production.
Although they are highly pleiomorphic, mitochondria are generally sausage or
thread shaped (0.2-0.6 pm wide and one or several pm long). They consist of
an outer membrane, which resembles other membranes, and an inner membrane
containing more protein than lipid. This inner membrane is folded into the mi-
tochondrial matrix, forming the mitochondrial cristae. Among the many proteins
of this inner membrane are the enzymes of the respiratory chain and the ATP
synthetase (mitochondrial "ATPase"). Many of the inner membrane enzymes, for
instance succinate dehydrogenase and cytochrome oxidase can be histochemically
demonstrated (cf. Chap.2S). The mitochondrial matrix contains a wide variety of
enzymes from the tricarboxylic acid cycle (Krebs cycle or citric acid cycle), fatty
acid oxidation, amino acid metabolism, etc. Many of these enzymes are of physio-
logical importance and exhibit a tissue-specific distribution. For instance glutamate
dehydrogenase, which is responsible for ammonia formation from amino acids
(glutamic acid) has a much higher (histochemical) activity in mammalian liver and
intestine than in most other tissues.
20 H. Lyon, B. van Deurs, P. Pren~, E. Hasselager, E. Schulte
Depending on cell type, the number of cristae formed by the mitochondrial inner
membrane and the amount of matrix may be more or less abundant. Cells with rela-
tively large ATP requirements (e.g. certain musele cells, cells of convoluted tubules
of the kidney, salt gland cells) have numerous cristae and relatively sparse matrix,
while cells which are more involved in intermediary metabolism (e.g. liver cells)
may have relatively few cristae, but abundant matrix. This morphological variety
reflects differences in the relative and absolute amounts of various mitochondrial
proteins and enzymes, which are often detectable by enzyme histochemistry.
Mitochondria are frequently distributed throughout the cytoplasm (e.g. liver
cells). However, in some transporting epithelial cells, which show a marked polar-
ity (e.g. tubule cells of the kidney), mitochondria may show a regionallocalization,
typically elose to the basal plasma membrane, where the ATP-consuming Na/K-
ATP-ase is localized. In heart musele and cross-striated muscIe, mitochondria lie
in elose proximity to the myofibrils, occasionally forming almost geometrical ar-
rangements (e.g. insect flight musele).
In Sects.2.2.2 and 2.2.5 the outward protein traffic of the cell has been outlined,
however, cells can also intemalize proteins. The process is referred to as endocytosis
and involves invagination of an area of the plasma membrane which is finally
"pinched off' to form an endocytic vesicle.
Endocytosis is divided into phagocytosis (e.g. a white blood cell ingesting a
bacterium) and pinocytosis (uptake of cellular solutes). Pinocytosis may be either
fluid-phase uptake or receptor-mediated uptake (an adsorptive uptake mechanism,
where the molecules in question are bound with a degree of specificity to receptors
on the cell surface prior to intemalization). Following endocytosis endocytic vesi-
eles fuse to form the endosome system, which comprises larger vacuoles as weIl as
sm aller vesieles and tubules. These structures continuously undergo the processes
of mutual fusion and formation by "pinching off' and may therefore be considered
as an interconnected compartment. The endosome system is somehow responsible
for the sorting of intemalized molecules with respect to their next destination (the
cell surface in the case of receptor cyeling; lysosomes for substances to be degraded
or receptors and ligands to be down-regulated; etc.). Histochemically endosomes
may be distinguished from lysosomes by incubating cells with a "marker" of en-
docytosis at low temperature (18°e in mammals). At this temperature endocytic
vesieles do not fuse with lysosomes. These structures may also be differentiated
by histochemical demonstration of lysosomal markers such as acid phosphatase.
The endocytotic path involves two main compartments: first endosomes, then
lysosomes. Avesiele containing both endocytosed material and lysosomal enzymes
is often called a secondary lysosome (in contrast to the primary lysosome, which
is derived directly from the Golgi complex). The secondary lysosomes essentially
form a recyeling system intermittently receiving new material from endosomes and
2 The Structura1 and Chemical Basis for Histochemistry 21
newly synthesized hydrolytic enzymes from primary lysosomes. How the fusion
with primary lysosomes is regulated is as yet unknown.
Secondary lysosomes which are no longer enzymatically active often contain
an indigestible residue and are called residual bodies. These may have the character
of pigment (e.g. lipofuscin) which can be demonstrated with special cytochemical
methods (Sect.18.2.2).
Endocytosis - intracellular digestion - is important from both a basic biological
and a cytochemical point of view, since many macromolecules internalized by
cells can be detected by enzyme- or immuno-histochemical techniques. Moreover,
currently available techniques enable a highly detailed description of the functions
and the malfunctions oflysosomes (cf. storage diseases, Sects.31.10.8 and 31.11.2).
The processes described above are called heterophagy, literally, "eating" of
material "different" from the ce1!. Autophagy, lysosomal degradation of some of
the cell's own organelles, also occurs. This may take place as part of the restruc-
turing of a cell, as a result of a cell injury, as an integrated part of morphogenesis
(programmed cell death or apoptosis), or sometimes as a response to starvation.
The cytoplasmic matrix contains several systems of fibre proteins, which are impor-
tant in the maintenance and restructuring of cell morphology, intracellular transport,
and in cell motility. The intracellular fibre proteins comprise the "cytoskeleton" of
the cells. The various elements of the cytoskeleton can be demonstrated using the
electron microscope. The most important fibre proteins are tubulins, actins, and
the proteins forming the intermediate filaments (e.g. keratins). These fibre proteins
constitute a very substantial part of the protein of the cell, actin alone up to more
than 10%.
Tubulins. These are globular proteins which associate to form tube-like fibre struc-
tures, microtubules, which may attain a length of several {Lm. Microtubules are im-
portant in the internal organization of the cell and in establishing "tracks" for the
intracellular movement of organelles and granules. The actual movement depends
on the interaction between microtubules, actin filaments, and "motor" proteins. The
microtubules play a similar role in chromosome movement in nuclear division. In
cilia microtubules form the so-called axoneme structure, with 9 microtubule dou-
bIets surrounding two single microtubules. The presence of the "motor" protein
dynein on the doublets ("dynein arms") enables these to slide relative to each other
leading to movement of the cilium.
Actins. These occur partly in the form of a free globular protein, G-actin, partly
as a filament protein, F-actin, which is a polymer of G-actin in a double helical
arrangement. There is a dynamic equilibrium between the two states of actin, in the
same way as for free tubulin and microtubules. F-actin comprises the greater part
of the thin filaments in the sarcomeres in striated muscle and in smooth muscle
22 H. Lyon, B. van Deurs, P. Prent/ll, E. Hasselager, E. Schulte
cells. Actin occurs in practically all cell types, e.g. epithelial cells and fibroblasts,
although it may be of a different subtype (as judged immunocytochemically). In
motile cells, and in cultured cells actin is frequenüy arranged in bundles, the so-
called stress fibres, immediately beneath the plasma membrane. Large amounts of
F-actin filaments also occur in microvilli where they are arranged parallel to the
long axis.
Myosin. This is a large protein molecule which, especially in muscle cells, forms
filaments, and which can be specifically demonstrated due to its ATPase~activity
(Sects.24.6.4, 31.11.6). Actin and myosin interact in cell movement and in muscle
contraction, forming actomyosin.
This section deals with the structure and chemical composition of bacteria. Qnly
sufficient details to understand the background for the mechanisms of the Gram
and Ziehl-Neelsen staining methods will be given (Sect.6.1.6). Bacteria (Fig. 2.3)
are prokaryotic cells, Le. cells without a nucleus, a nuclear membrane, or a mitotic
apparatus. The essential DNA of bacteria is found in a single circular molecule
(chromosome). Additional, often pathogenicity and antibiotic resistance related
functions are coded for on extrachromosomal circular DNA molecules of vary-
ing size. These may be present in multiple copies and are known as plasmids.
While these structures are not demonstrable histochemically, OUT recently devel-
oped ability to manipulate both eukaryotic and prokaryotic nucleic acids is heavily
dependent on plasmids. This technology forms the basis for the development of in
situ nucleic acid hybridization techniques which can be used to detect and demon-
strate genes or smaller base sequences at the cellular level (cf. Sect.20.6).
The bacterial cytoplasm may also contain granules consisting of neutral poly-
mers such as stareh, glycogen, or polyphosphate (volutin). There are no mitochon-
dria.
Bacteria have both a plasma membrane and a thick cell wall (Fig. 2.4). The
plasma membrane is structurally and functionally very similar to that found in
eukaryotic cells. The active transport processes it supports contribute to the devel-
opment of an inner osmotic pressure weIl in excess of that found in most fluid
environments (between 500 and 2,000 kPa). This would cause bacteria to burst if
they were not surrounded by a cell wall with considerable mechanical strength (Fig.
2.3). This strength is due to a large complex polymer, known as peptidoglycan or
murein.
Fig. 2.3. Diagram of a bacteriwn with granule (Gr), plasmid (PI), chromosome (Ch), cell mem-
brane (CM), cell wall (CW), ciliwn (Ci) and flagellwn (FI). The dotted rectangle is expanded in
Fig.2.4.
24 H. Lyon, B. van Deurs, P. PrentQ!, E. Hasselager, E. Schulte
:}-OM
J-Pg
-PS
6 ,I"'T""'-r,"""T,-,,.....,,""T""'1', }- C M
, ! ! , " .
A B
Fig. 2.4. Diagram of bacterial cell wall in Gram-positive (A) and Gram-negative bacteria (B).
Peptidoglycan (pg), cellular membrane (CM), outer membrane (OM), perip1asmatic space (PS).
The figures indicate the thickness of the layers in nm.
Development and application of the Gram staining method in the latter part of the
19th century fortuitously revealed fundamental properties of bacteria that are valu-
able both for classification and identification. The difference in staining between
Gram positive and Gram negative organisms appear to reflect differences in cell
wall structure revealed latterly by electron microscopy.
The peptidoglycan layer is a polymer found in all bacteria (with the exception
of mycoplasmas) and consists of alternating molecules of N-acetyl glucosamine
(NAG) and N-acetyl muramic acid (NAM) with tetrapeptide side chains (which
include D-amino acids) attached to the muramic acid residues. The side chains are
bonded together by a further pentapeptide bridge, effectively linking the NAM-
NAG polymer. The peptidoglycan layer is thus a single giant molecule. Gram-
positive bacteria have a thick peptidoglycan layer which comprises up to 90% of
the total cell wall whereas Gram-negative bacteria have a thin layer amounting to
between 5 and 20% of the cell wall (Fig. 2.4).
Both groups of bacteria may express proteinaceous and polysaccharide material
on their surfaces. In addition Gram-positive cell walls contain teichoic acids (Greek
teichos = wall) (glycerol polymer bonded together with phosphodiester bonds)
which form a part of the surface antigen structure. Gram-negative cell walls have
an additional layer outside the peptidoglycan known as the outer membrane. This
layer retains the standard unit membrane structure but also includes substantial
quantities of lipopolysaccharide (see Fig. 2.5). The latter are extremely toxic to
higher organisms and are referred to as endotoxins. Lipoproteins bind the outer
membrane down to the peptidoglycan layer. In the outer membrane matrix proteins
are arranged in groups so that they form "pores" through which small hydrophilic
molecules may pass. A good survey of the biochemistry of bacteria may be found
in Jawetz et al. (1989).
2.3.2 Mycobacteria
Mycobacteria are Gram-positive rods which show the property known as acid
fastness, Le. they can withstand decolourization with acid after staining with an
2 The Structural and Chemical Basis for Histochemistry 25
CW
111111111111111111111111111111111111 CM
11 1111111111 1111111111111111 11111111
Fig. 2.5. Diagram illustrating the molecular structure of the cell wall (CW) and the cellular
membrane (CM) in Gram-negative bacterium. The cell wall is composed of lipopolysaccharide
(Lps), outer membrane (OM), lipoprotein (Lp), and peptidoglycan (Pg).
2.4.1 Collagens
Collagens are the most abundant proteins in mammals (around 25% of total protein).
They usually form fibres which are themselves collections of smaller fibrils. These
in turn consist of aggregates of long (300 nm) tropocollagen molecules in a specific
arrangement. Electron microscopy usually reveals a characteristic cross striated
pattern in collagen fibrils with aperiodicity of about 67 nm.
The periodicity results from a staggered lateral association of the tropocollagen
molecules. Molecules are lined up along the fibrils with a 35nm gap in between.
As the molecular length is elose to 4.4 times the period interval this gap accounts
for the actual ratio of 5 periods per molecule (335nm/67nm). The 35nm gap is very
accessible to the metal stains used in negative staining procedures.
The tropocollagen molecule consists of three polypeptide chains (a-chains)
with a unique primary structure. Every third amino acid in the sequence is glycine
and the molecule has an unusually high content of proline and hydroxyproline. On
the basis of the primary structure, seven different a-chains (see Table 2.8), each
with the sequence gly-X-Y repeated 300-350 times, can be distinguished. Due to
steric restrictions from proline + hydroxyproline each of the three a-chains forms
an extended left-hand helix. In turn, these helices wind around each other to form
the greatly extended right-hand tripie helix of the tropocollagen molecule. The
intimate contact between the three helices is made possible because the glycine
residues are always nearest to the centre of the tripie helix. This conformation
allows the maximum number of intra- and interchain hydrogen bonds, with chain
ftexibility highly restricted by the proline and hydroxyproline residues.
Collagen structure is further stabilized by covalent cross-links between a-chains
and between neighbouring tropocollagen molecules. These cross-links are primarily
between lysyl and hydroxylysyl residues. As shown in Table 2.8, it is possible to
distinguish between at least five different collagen types depending on which of the
seven a-chains take part in the formation of the molecule. The different collagen
types contain carbohydrate in highly varying amounts. This carbohydrate, which
may be either galactose or glucosylgalactose, is bound by O-glycosidic bonds to
hydroxylysyl.
2.4.2 Reticulin
The term reticulin was originally used in relation to reticular fibres in the same
way as collagen is used in relation to collagen fibres, i.e. as a term for the protein
believed to be characteristic for the fibres. Later, reticulin became adesignation
for the PAS-positive material in reticular fibrils and basal membranes.
It is now well established that reticular jibrils are composed of type In collagen.
These fibrils are far thinner and contain more carbohydrate than type 1 collagen
fibrils. These two features are responsible for the histochemical differences between
the two fibril types. Characteristically, reticular fibrils are stained black with ox-
idative argyrophil methods (Sect.8.3.2), red with PAS (Sect.9.2.1), and yellow with
2 The Structural and Chemical Basis for Histochemistry 27
Basal membranes are structures which are found in the boundary between the con-
nective tissue matrix and the adjoining epithelial or endothelial cells (Fig. 2.1), and
around muscle cells and fat cells. In addition basal membranes are found between
the two celllayers in the glomeruli of the kidney. Light microscopic demonstration
can be achieved using the same methods as those applied to reticulin and collagen
fibrils (Sect.2.4.2). Examination by electron microscopy reveals an homogeneous,
light area called the lamina lucida direct1y basal to the plasma membrane of the
epithelial or endothelial cell. This consists predominantly of the peripheral portion
of the cell surface glycoproteins. Under this lies the lamina densa (basal lamina)
which is 20-80nm thick (about 30ünm in glomeruli). This comprises a network
of thin filaments in a finely granular ground substance. Below the basal lamina,
reticular fibrils are frequently observed forming a network (reticular lamina) and
continuing into the underlying connective tissue. The basal lamina is composed of
types IV and V collagen as well as other elements (e.g. laminin and a heparan
sulphate proteoglycan).
2.4.4 Elastin
Elastin (Table 2.8) is one of the two components in elastic fibri/s. The other com-
ponent consists of "elastic microfibri/s".
Elastin has a high content of the non-polar amino acids alanine, valine, leucine,
and isoleucine, and a considerable amount of tyrosine. The polypeptide chains show
random coil configuration (no defined secondary structure) and are bound together
by cross links in an unorganized three-dimensional network. The cross links in
elastin are characteristic and arise between four lysyl groups from four different
peptides. The cross links Can be isolated from elastin and consist of the for elastin
specific amino acids, desmosine and isodesmosine.
desmosine isodesmosine
28 H. Lyon, B. van Deurs, P. Prenllll, E. Hasselager, E. Schulte
2.4.5 Fibronectin
2.4.6 Laminin
2.4.7 Fibrin
Fibrin (Table 2.8) is a fibre protein which takes part in the formation of the fibrin
coagulum in healing processes and in clotting. Together with fibronectin (Sect.2.4.5)
it forms the matrix that enables the development of granulation tissue.
Fibrin is derived from fibrinogen which is a soluble plasma protein (Table 2.8).
Histoehemical demonstration is achieved with relative selectivity using a number
of trichrome methods (Sect.21.6), e.g. Mallory's PTAH and the MSB method.
Good results can also be obtained with the DMAB method for tryptophanyl groups
(Sects.9.4.5 and 9.4.6).
2.4.8 Proteoglycans
Glycosaminoglycan Occurrence
Hyaluronic acid connective tissue, skin, vitreus body, cartilage, synovial fluid
Chondroitin-4-sulphate cartilage, cornea, bone, skin, arteries
Chondroitin-6-sulphate cornea, bone, skin, arteries
Dermatan sulphate skin, blood vessels, heart
Keratan sulphate cartilage, cornea, intervertebral discs
Heparan sulphate lung, arteries, cell surfaces
Heparin lung, liver, skin, mast cells
30 H. Lyon, B. van Deurs, P. Prent0, E. Hasselager, E. Schulte
Fig. 2.6. Schematic representation of proteoglycan aggregate. The framed area is expanded on
the right. The proteoglycan aggregate consists of chondroitin sulphate (C) and keratin sulphate
chains (K) bound covalently to protein chains (P) which are in tum bound non-covalently to a
hyaluronic acid chain (H) by special link proteins (L).
Elastic micro- Glycoprotein with a high content of Elastic fibres (see elastin).
fibrils cysteinejcystine, and proline, but
no hydroxyproline. It is distribu-
ted as microfibrils along the sur-
face of the elastic fibres.
Fibronectin Soluble 220 kOa glycoprotein whicb forms Blood plasma and otber
fibronectin a dimer held together by disul- body fluids.
= "cold in- phide cross-links at the carboxyl
soluble glo- terminals of the two polypeptide
bulin H
subunits.
Insoluble Usuallya dimer, but may occur in a Partly pericellular, in elose
fibronectin = multimer form, when side groups con tact to tbe cell surface
•ceil spread- in the polypeptides form cross- in tbe form of fibrils, net-
ing factor H
links botb between tbe poly- works, or focal deposits.
peptide chains themselves and to Partly in tbe matrix cor-
collagen, fibrin, and proteo- responding to loose con-
glycans. nective tissue.
Laminin 1,000 kOa asymmetric g1ycoprotein Basal membranes, especially
molecule composed of three 200 in lamina lucida.
kOa A-chains and one 400 kOa
B-chain bound together by di-
sulphide bonds.
Fibrinogen Precursor to fibrin whicb is com- Soluble protein in blood
posed of tbree nodules bound to- plasma.
getber by two rods. Total lengtb
46 nm. Large content of aspar-
agyl, glutamyl, and tyrosyl-O-
sulpbate.
Fibrin Thrombin acting on fibrinogen Takes part in tbe formation
splits four arginyl-glycyl peptide of a coagulum in bealing
bonds and removes four low mo- processes and tbe forma-
lecular (10-20 amino acids) poly- tion of tbrombi. Togetber
peptides (fibrinopeptides) witb witb fibronectin it forms
many negative cbarges. The re- tbe matrix in granulation
mainder is a fibrin monomer tissue.
witb about 97% of tbe amino
acids of fibrinogen. Tbe mono-
mers immediately aggregate to
tbe insoluble polymer fibrin.
Fibrinoid Probably a mixture of fibrin and In vessel walls in collagen
some plasma proteins. diseases.
3 Reagents
Aprerequisite for successful and reproducible staining is the use of the correct
reagents. The composition, preparation, and storage qualities should be known. A
detailed knowledge of the chemical and physical properties of reagents and the
reactions for which they are to be used, is therefore highly desirable. Although
beyond the scope of this text, it is also mandatory to identify the specific health
risks associated with reagents and take appropriate precautions. The information
presented below is intended to give guidance to the reader in developing asound
approach to the preparation, storage and disposal of histochemical reagents.
3.1.1 Purity
The degree of purity of areagent is usually described using one of the following
terms:
• Technical quality: normal commercial quality, venale, crudum
• Purified: gereinigt, practicum
• Pure: purum, rein, reagent quality, and pharmacopoeia terms such as Ph. Eur.;
Ph. Helv.
• Very pure: purissimum, reinst.
• Analytical grade: analytical reagent (AR), pro analysi (p.a.), zur Analyse
"Technical quality" and "purified" should not be used in analytical work, be-
cause reagents labelled "purified" frequently contain remarkable quantities of con-
taminants. The choice between "pure" and "very pure" depends upon what the
presence of possible impurities may signify for the use.
H. Lyon (Ed.)
Theory and Strategy in Histochemistry
© Springer Verlag 1991
34 A.P. Andersen, H. Lyon, P. Jakobsen
Water quality. The purity of water from different sources is shown in Table 3.1.
A measure for the quality of water may be obtained by measuring the conductivity
of the water. This is indicated in Mn-1cm- 1 = ftScm-1, where Mn = megaohm;
ftS = microSiemens (SI unit system (Systeme International d'Unites».
Organie Conduetivity
Water quality Ions eompounds Gases (J.lSem- l ) at 25°C
This should always follow international and Iocal recommendations. Examples are
BSC (Biological Stain Commission, Sect.3.3.1O), ECCLS (European Committee
for Clinical Laboratory Standards), and DIN (Deutsche Institut für Normung). In
addition, it is expedient if the following is noted on the label:
1. The use of the reagent
2. Date of purchase or preparation
3. Special requirements regarding storage (e.g. refrigerator)
4. Date and method for discarding.
Most reagents contain unstable components that are susceptible to particular en-
vironmental influences. Important examples include the potential for oxidation by
atmospheric oxygen, uptake of carbon dioxide by alkaline solutions leading to the
formation of carbonates, and 'spontaneous' cleavage of macromolecules. Exposure
to light, particularly ultraviolet light, (a potent catalyst for many chemical pro-
3 Reagents 35
When reagents are discarded the regulations for protection of the environment
should be strictly adhered to. Choice of method depends on the concentration of
the substance, the rate with which it is degraded in nature, and its toxie effects.
Three practical approaches are widely used:
1. Sink-disposal using plenty of water to flush through the waste-pipe
2. Detoxification followed by sink-disposal
3. Collection and transportation for incineration
Individual laboratories should establish regulations for the collection and dis-
posal of refuse in accordance with local regulations.
3.2 Solvents
Only certain polar and non-polar molecular compounds are liquid at room
temperature. Ionic compounds, atomic lattice compounds, and metals, with the ex-
ception of mercury, are all solids under ambient conditions. In consequence the fol-
lowing types of solvents are available: polar (e.g. water), non-polar (e.g. hydrocar-
bons), and solvents containing both polar and non-polar groups with approximately
equal influence, termed amphiphilie, e.g. sodium laurylsulphate C12B2SS04Na. The
following rules may be applied to solvent-solute combinations:
• Ionie eompounds form strong intermolecular bonds but many will dissolve in
water and other very polar molecular compounds, as the ions can form stronger
bonds with the polar molecules
• Atomie lattiee eompounds (e.g. Si02) cannot be dissolved in either polar or
non-polar molecular solvents due to the strength of their in~r-atomic bonds
• Polar moleeular eompounds may be dissolved in other polar compounds
• Non-polar moleeular eompounds can only be mixed with other non-polar com-
pounds
From the practical point of view therefore, polar solvents have more widely
applicable dissolving properties than non-polar solvents since the latter can only
dissolve other non-polar molecular compounds. The suitability of a polar solvent
can be assessed on the basis of the dipole moment (p,) of the compound, its ability
to form hydrogen bonds, its dielectric constant (E), and its ability to form chemical
bonds with the dissolved compound.
Dipole moment
Solvent 1O- 30 xCxma Db
The greater the dipole moment the more strongly the compound will bind to
other permanent or induced dipoles. The dipole moment is, however, not the only
factor that determines solubility, the nature of any solvent-solute bonds formed is
also important. Both hydrogen bonds and complex bonds are relevant in this regard.
Compounds containing -OB -SB -NH2 >NB -COOB and -S03B are all good
at forming hydrogen bonds with each other and with water.
The ability of a polar solvent to dissolve other polar molecular compounds
can thus be assessed from the size of the dipole moment and the ability to form
3 Reagents 37
hydrogen bonds, while its ability to dissolve ionic compounds may be assessed
from its dielectric constant and ability to bind to the compound.
Table 3.3. Constants for the solvents water, ethanol and xylene.
3.2.1 Water
The Structure of Water. Water forms polar, non-linear molecules which are bound
together by hydrogen bonds in both solid (ice) and liquid forms.
38 A.P. Andersen, H. Lyon, P. Jakobsen
Water as a Solvent for lonic Compounds. Water is an excellent solvent for ionie
compounds (salts) due to its high dielectric constant and its ability to bind to ions.
When a salt is dissolved in water ions are formed. Their activity may be expressed
as the ionic strength of the solution using
1 2
1= "2L:CiZj
Water as a Solvent for Acids. Water is, due to its basie character, a good solvent
for acids. The following reaction takes place: A( 1) + H20 +::! B( 1) + H30+, where
A( 1) = acid and B( 1) = corresponding base. The concentration of H30+ in the
aqueous solution is expressed by pH which is < 7.
Water as a Solvent for Bases. Water is also a good solvent for bases with the
following reaction taking place:
B( 1) + H20 +::! A( 1) + OH-
The concentration of OH- is expressed using [H+] as [H+][OH-] = K w • fu a
solution of a base pH is > 7.
Li+ Na+ K+
Mi+ Ca2+ Ba2+
sol- CH3COO- Cl- Br- 1- NO;
Water as a Solvent for Hydrophobie Substances. While such substances are not
soluble in water on their own, they may form emulsions composed of micelles
after the addition of a detergent. A detergent is a substance which contains both a
hydrophobie radical and a strong hydrophilie group. It may thus bind both to water
and to the hydrophobie substance and thereby bring the latter into solution as an
oiVwater emulsion.
40 A.P. Andersen, H. Lyon, P. Jakobsen
3.2.2 Ethanol is a polar solvent (cf. Table 3.2). It is very versatile because it
contains both a hydrophobic group and a hydrophilic group which can fonn hy-
drogen bonds. Ethanol is a translucent, colourless, volatile liquid which is easily
inflammable. It should be kept in a closed container protected against light. The
commercial product is 96% v/v alcohol and absolute alcohol ~99% v/v. Ethanol is
freely miscible with water and is an excellent solvent for the majority of organic
and many inorganie compounds, but not for salts. The majority of lipids are very
sparingly soluble in ethanol.
3.2.3 Xylene is a purely non-polar or hydrophobie solvent and can be used for
dissolving hydrophobie substances such as paraffin. The commercial product is a
mixture of 0-, m-, and p-dimethylbenzene. It is a translucent, colourless, volatile
liquid. The vapours are noxious, and xylene may be absorbed~through the skin.
Inhalation and contact with xylene must be avoided. Xylene bums with a very
sooty flame and can fonn explosive mixtures with air. It should be kept in weH-
closed bottles protected against light. Xylene is fuHy miscible with ethanol, ether,
and chlorofonn.
3.3 Dyes
3.3.1 Definitions
3.3.2 Colour
green
Fig. 3.1. Complementary colour circle showing relationship between wavelength of absorbed (A)
and observed (C) colours. Ultraviolet (UV) and infrared (IR) regions.
E
0.8
0.4
o~~~~--~--~==~==~
40 0 50 0 60 0 70 0 nm
where there are three o--bonds, 1 7l"-bond, and two lone pairs, the energy levels will,
according to the molecular orbital theory be o--bonding, 7l"-bonding, n-non-bonding,
o--antibonding, and 7l"-antibonding.
The relationship between the energy levels of the bonds is shown in Fig. 3.3.
Since greater transitions in energy level require more energy input, it can be seen
that substances containing only o--bonds need light of higher energy (shorter wave-
3 Reagents 43
cr* antibonding
1t* antibonding
n non-bonding
1t bonding
cr bonding
length) to excite the electrons, while substances containing double bonds (rr-bonds)
and/or heteroatoms can be excited by light of lower energy (longer wavelength).
If a substance contains several 7l'-bondsllone pairs conjugated with each other,
the energy difference between 7l'-bonding and 7l'-antibonding molecular species is
reduced, and absorption occurs at a higher wavelength.
From benzene it is known that in cyclic structures the presence of a suitable
number of double bonds conjugated to each other gives the molecule its aromatic
character. The electrons involved are considered not to be localized in individual
7l'-bonds but delocalized over the whole molecule. A similar delocalization can
arise in planar, open systems containing conjugated double bonds and/or lone
pairs conjugated to the double bonds even if the substance does not have aromatic
characteristics.
In larger aromatic and heteroaromatic systems as wen as in non-aromatic sub-
stances containing many conjugated double bonds, (e.g. carotenoids), the energy
difference between the bonding and antibonding molecular orbitals is large enough
to allow absorption in the visible spectrum.
Alkanes win thus absorb light with a wavelength of approximately 150 nm,
while alkenes (>C=C<) will absorb at around 190 nm. For benzene, where there
is a delocalized cyclic 71' orbital over all six carbon atoms, absorption is seen at
203.5 and 254 nm.
In nitrobenzene
where the size of the delocalized cyclic 71' orbital is increased compared with
benzene absorption occurs at 268.5 nm.
In a dye such as Crystal Violet the size of the delocalized cyclic 71' orbital has
become so great that the absorption has moved into the visible part of the spectrum
at 590 nm.
44 A.P. Andersen, H. Lyon, P. Jakobsen
Chromophore groups. Groups that are able to absorb light in the ultraviolet or
visible part of the spectrum are called chromophore groups or chromophores. Ex-
amples of such groups are:
-N=N- azo
-N=O nitroso
-N02 nitro
>C=C< alkene
>C=O oxo
As seen in the previous examples it is necessary to have several chromophores
conjugated with each other before a compound will absorb light in the visible
spectrum.
Auxochrome groups. These are groups that are not able 10 absorb light in the
ultraviolet or visible spectrum by themselves; however, when they are bound to a
chromophore system they can change the light absorption 10wards longer wave-
length and at the same time impart a higher intensity to the absorption. Auxochrome
groups contain a lone pair of electrons that, in conjugation with a conjugated sys-
tem, are able to increase the size of the delocalized cyclic 'Ir orbital. Examples
are:
-OH hydroxyl
-SH thiol
- NH2 amino
-Cl, -Br,-I halogens
Substitutions in auxochrome groups often produce additional alterations in
colour. Substituents of this kind are called modifying groups (modifiers). Alky-
lation will cause a bathochrome shift, acylation a hypsochrome shift.
°
Br
When writing structural formulae for compounds containing a delocalized cyclic 'Ir
orbital it is weH known that classical symbols far bond valency give an erroneous
o
view of the structure. For benzene the classical Kekule-structures are:
©
where the ring designates the delocalized cyclic 'Ir orbital (the aromatic system).
46 A.P. Andersen, H. Lyon, P. Jakobsen
It is, however, frequently useful 10 be able to write the structures using the
classic formulae as these give a better idea of where possible charges may be
found and how large the delocalization iso
Formulae of this kind are called resonance formulae, and the true structure is
to be found somewhere between the different possible resonance formulae for the
substance. (The electron distribution in the substance corresponds to a superposition
of the electron distributions of the possible resonance formulae). This is shown by
connecting different resonance formulae for the same substance with a double
arrow.
For aniline the following resonance formulae can be written:
There are many possible resonance formulae for the more complicated dyes; it
is therefore not possible to predict which are the most probable. Some of the
resonance formulae for Acridine Orange can be written:
Note that in the rest of this book a compromise has been Made between the modern
ring formulae and the Kekul6 notation. Where relevant, the latter has been retained
to show the presence of chromophores such as the quinone configuration.
H~C~
CH 3
HO@::;:~Or' sQ(
CH,
HO
o N-N
CH 3
CH
3
The trans fonn is coloured because the molecule is planar and electron delocal-
ization can take place over the whole conjugated system including the lone pair in
the OH group. In the cis fonn the two benzene rings cannot lie in the same plane
as the size of the ortho-methyl groups hinders this confonnation (steric hindrance).
The cis compound thus appears colourless. Exposure to light causes a change from
the trans to cis fonn. On standing in the dark the cis compound reverts to the more
stable coloured trans compound.
Theory concerning the pattern and intensity of absorption in the visible and ultra-
violet spectrum for a selected compound is based on the molecular orbital theory.
Certain "rules of selection" exist that indicate whether an electron transition is per-
mitted or not (forbidden). These depend on the "symmetry" of the energy levels
involved in the transition. Forbidden transitions will often be seen as absorption
peaks with low intensity.
When examining the purity of a dye sampie it is common to record an ab-
sorption spectrum of the substance and compare this with data for the pure dye.
ldentity does, however, not constitute an absolute criterion of purity.
If the spectrum of the substance under investigation contains more absorption
maxima than the pure dye, clearly the substance is impure. If an identical spectrum
to that given for the pure dye is found, another (similar) substance with the same
absorption maxima may be present. Moreover, there may be absorptions that are
not observed with the spectrometer used, or there may be several closely spaced
absorptions that "melt" together to one and thus give a "false" impression of purity.
To obtain an absolute detennination of purity thin layer chromatography (TLC) or
high perfonnance liquid chromatography (HPLC) analyses should be made.
In the Colour Index (Society of Dyers and Colourists, 1971) the dyes are designated
by a 5-digit number, C.I. number and a specially constructed name.
The trivial names in more general use are given in Conn's Biological Stains
(Lillie, 1977). It should be noted, however, that there are numerous synonyms, i.e.
names which may be the same for several dyes. Biological Stains thus distinguishes
between e.g. C.!. 42585 Methyl Green and C.!. 42590 Ethyl Green. Ethyl Green
48 A.P. Andersen, H. Lyon, P. Jakobsen
is frequently sold as Methyl Green, and in this case the C.I. number is the only
certain way of detennining which dye is actually being used. The C.I. number
should therefore always be noted when ordering dyes.
Table 3.4. Dyes, dye precursors, and pigments c1assified according to chemical structure.
c.I.
group Formula (the presumed
no. Class Example active component)
3. Azo
A. Monoazo Orange G
c.I. 16230
B. Disazo Congo Red
c.I. 22130
C. Trisazo Chlorazol Black E
c.I. 30235
D. Tetrakisazo Sirius Red F3B
c.I. 35780
3 Reagents 49
C.I.
group Formula (thc prcsumed
no. Class Examplc activc component)
:-©
5. Azo coup I·mg components Naphthol AS·BI
~OH HCO
G0
CI. 37566
CO - N
No NO,
6. Tetrazolium salts Nitro blue tetrazolium N
CI. M. O~
~ °N_N,,~
I (--i2;
@{J~N~N
N-N
2 CI
CH 30 OCH 3
8. Arylmethane
A. Diphenylmethane
. . triphenylmethane
B. Dlammo Fast Green FCF
CI. 42053
HO
50 A.P. Andersen, H. Lyon, P. Jakobsen
CI.
group Formula (the presumed
no. Class Examplc active componcnt)
C Triaminotriphenylmethane Pararosanilin
CI. 42500
c.1.
group Formula (the presumed
no. Class Examplc activc component)
9. Xanthene
A. Aminoxanthene
a. Pyronin Pyronin Y
c.1. 45005
b. Succinein Rhodamine S
c.1. 45050
c. Rosamine Sulphorhodamine B
c.1. 45100
d. Rhodamine Rhodamine B
c.1. 45170
52 A.P. Andersen, H. Lyon,P. Jakobsen
C.I.
group Formula (thc prcsumcd
no. Class Examplc activc component)
B. Hydroxyxanthene Eosin
c.1. 45380 o
Br
10. Acridine
~N"/ -~ "'CH 3
14. Quinonei~ine
A. Indamme Toluylene Blue
c.1. 49410
H 2N
~'O·/CH'
0 N _ N "'CH
3
H 3C
HOSNSO
B. Indophenol Indo-oxine
No. c.1. no.
NO
/, ~
C.I.
group Fonnula (the presumed
no. Class Example active component)
b. Rosinduline Azocarmine B
c.1. 50090
c. Safranin Safranin 0
c.1. 50240
D.Oxazin Gallocyanin
C.1. 51030
E. Thiazin
16. Anthraquinone .
A. HydroxyanthraqulDone Alizarin Red S
c.1. 58005
o
54 A.P. Andersen, H. Lyon, P. Jakobsen
C.I.
group Formula (the presumed
no. Class Examplc activc component)
B. Aminoanthraquinone Kernechtrot
c.1. 60760
~OH
~SO;
o OH
17. o 0
©CX:©
Indigo Indigo
c.1.73000
K
H H
N N/- N
x©9-~"-?wx
NlrNX
I CH3
+/N_ CH3
""N-
x = -CH 2 -S-C
CH 3
\
CH 3
3 Reagents 55
c.1.
group Formula (the presumed
no. Class Example activc componcnt)
Reactive Dyes. This new group should be added to the list. According to Rys
and Zollinger (1972), a reactive dye is defined as a coloured substance contain-
ing a group that is able to form a covalent bond between a carbon atom in the
dye ion or molecule and an oxygen, nitrogen, or sulphur atom in a hydroxyl, an
amino, or a thiol group respectively in the substrate. The first reactive dyes which
appeared in 1956 were the procions which contain the reactive group 2-amino-4,6-
dichlorotriazine with two labile chlorine atoms:
+ 2R-OH -
+ 2 Hel
An example is:
56 A.P. Andersen, H. Lyon, P. Jakobsen
Procion Brilliant Yellow M-6G, C.I. 18971, that has been used for the demonstration
of newly formed dentine. Note, that the dye can also be c1assified as a monoazo
dye.
+
Diazonium Salts. These are salts of the type R-N=:N, X-. The bases are usually
quite unstable, and this is often the case also with the simple halogen salts. So-
called stabilized diazonium salts are double salts with zinc chloride, borotrifiuoride,
sodium hydrogen sulphate, or naphthalene-l,5-disulphonic acid They are relatively
stable at room temperature if kept dry. In solution however, especially at alkaline
pH, stability is very limited.
+
Diazonium salts only contain one -N=:N group, while tetr~nium salts contain
two, and hexazonium salts three. Despite this, the term diazonium salts is widely
used to denote the whole group.
Many of the salts are colourless, but when the chromophore azo group - N=N-
is formed on coupling to tissue bound aromatic amines or phenols or with enzy-
matically liberated naphthols and naphthylamines colour arises. Many diazotizable
amines are already dyes, e.g. Safranin, Pararosanilin, Fast Garnet GBC base, and
Fast Black K base. In these cases the new azo dye is usually far more intense
and is stable towards treatment with acid. Choice of a suitable diazonium salt for
a particu1ar histochemical purpose can be very difficult when confronted with the
jungle of currently available commercial products. Some examples of frequently
used diazonium salts are given in Table 3.5.
Table 3.5. The colour of some coupling products with different diazonium salts (modified after
Lillie and Fullmer, 1976, p. 248).
Mol.
weight Entero- Colour on coupling with
Name c.1. no. (amine) chromaffin ß-naphthol naphthol AS protein
Table 3.6. Properties of different stabilized diazonium salts (modified after Lillie and Fullmer, 1976,
p.148).
Table 3.6 indicates the coupling rate, the stability in solution, and the preferred
application for the stabilized salts listed in Table 3.5.
Azo Coupling Components. This group of reagents are not dyes in themselves
but, on enzymatic hydrolysis, they produce phenols, naphthols, or aromatic amines.
These are transformed into insoluble azo dyes and deposited as precipitates at the
site of the enzyme activity by coupling to diazonium salts either immediately
(simultaneous coupling) or later (post-coupling).
The azo compound arising from the azo coupling reagent on coupling with a
diazonium salt should fulfil the following criteria:
1. Insoluble in water, acid, and base and preferably also in ethanol and xylene
2. Form a finely granular precipitate
3. Give a precise localization
4. Show high substantivity to proteins (Sect.1.2.3)
Early azo coupling reagents did not meet these requirements. The azo com-
pounds of a-naphthol are moderately soluble in water, have only slight substan-
tivity, and form coarse precipitates, while those of ß-naphthol are only slightly
soluble in water, have a finely granular precipitate, and moderate substantivity,
but are easily dissolved in an alkaline solvent. The azo derivatives of 6-bromo-2-
naphthol and 6-benzoyl-2-naphthol are insoluble in water, have a finely granular
precipitate, and moderate substantivity, but are destroyed in alkali.
In Table 3.7 a survey of modern azo coupling reagents is given. The azo com-
pounds formed with these are a11 insoluble in water, form very finely granular
precipitates, show pronounced substantivity, and give a precise and sharp localiza-
tion.
Naphthol AS 37505 263.3 Formula 3.43 phosphate alkaline and acid phosphatases
acetate esterase
sulphate arylsulphatase
OH chloroacetate protease of chymotrypsin
(-CO-CH 2 CI) Iike type
OOCO-NH-@ phenylpropionate protease of chymotrypsin
(-CO-CH 2CH 2C6 H s ) Iike type
Naphthol AS-E 37510 297.7 Formula 3.44 phosphate alkali ne and acid phosphatases
OH
OOCO-NH-@-CI
Naphthol AS-AN 37516 308.3 Formula 3.45 phosphate alkaline and acid phosphatases >
:-a
OH
~
OOCO-NH-@-NO, J
?=
Naphthol AS-TR 37525 311.8 Formula 3.46 phosphate alkaline and acid phosphatases ~
~
:-cl
OH H3k
OOCO-NH\QtCI f
t.>
Table 3.7. (Continued).
OH H3k
00 CO-NH-\Qr-CH,
Naphthol AS-CL 37531 327.8 Formula 3.48 phosphate alkali ne and acid phosphatases
oo::-N:~
CI
oo::-:~~
Br
U\
\Cl
Table 3.7. (Continued). ~
OO::-:~~CI
OCH 3
OO::_NH~~O> )-
~
2-aminoanthroquinone amide 393.4 Formula 3.52 phosphate alkaline and acid phosphatases
of 2-hydroxy-3-naphthoic
acid
IF
;:t:
o ~
ß
:-0
.....
OOX::-NH~ o
~
g
w
{
Aminoazobenzene amide of 367.4 Formula 3.53 phosphate alkali ne and acid phosphatases
2-hydroxy-3-naphthoic acid
OH
00 CO-NH-@-N~N-©
5,6,7,8-ß-tetralol-carboxylic 317.4 Formula 3.54 phosphate alkaline and acid phosphatases
acid ß-naphthylamide
©©:::-NHOO
ß-naphthylamine 142.2 Formula 3.55 alanyl derivative aminopeptidase
leucyl derivative aminopeptidase
ooNH'
0\
-
Ri
NH
©:d95 '
N
H
NH
' >
~
©:d95 N
I
CzH s
f
;:t:
j
:-c
.....
S-
O'
'"
g
3 Reagents 63
Schiff's reagent is used for the specific demonstration of aldehyde groups. The
reaction with ketone groups may effectively be ignored as this occurs much more
slowly (1~20 h against 1~20 min). Uses are listed in Table 3.8.
Basic Fuchsin is used to prepare the tradition al Schiff' s reagent. Altematively,
Pararosanilin, New Fuchsin, or possibly Rosanilin are now frequently preferred.
While it can be bubbled direcdy through the dye solution, S02 is usually prepared
from bisulphite, metabisulphite, dithionite or thionyl chloride, which are added as
salts, alone or in combination. Coloured impurities are then removed from the
primary decolourized product using activated carbon. It is important that this step
should not be performed before treatment with sulphite. The finished reagent should
be quite clear without colour and should smell of S02. pR should be between 1.2
and 2.5.
The precise chemical composition of Schiff's reagent is under debate. For many
years with Pararosanilin as the starting material, it was believed that Schiff' s reagent
Table 3.8. Histochemical reactions that are completed by the demonstration of aldehyde groups
with Schiff's reagent.
Demonstrated Demonstrated
Pretreatment groups* substance Reference
* Schiff's reagent demonstrates all aldehyde groups present, inc1uding surplus fixative (see 9.2.1).
64 A.P. Andersen, H. Lyon, P. Jakobsen
11
More recent investigations, however, suggest that the reagent is really the leu-
cosulphonic acid (1) itself (Gill and Jotz, 1976). While refrigeration is unnecessary,
Schiff's reagent should be kept in a dark bottle. The activity of the reagent can
easily be tested by mixing a few drops with one drop of fonnalin. The magenta
red colour should appear immediately. A number of other dyes and fluorochromes
can be used to prepare Schiff-type reagents after treatment with S02. In contrast
to the original Schiffs reagent, these are, however, often coloured. Those prepared
from Thionin (blue) and Acriflavine (fluorescent) are amongst the most frequently
used.
Commercial dye products show varying degrees of purity. In the U.S.A. a voluntary
arrangement has been made so that manufacturers of dyes can send their products
for investigation to the Biological Stain Commission (BSC). BSC has placed cer-
tain physical and chemical demands on the dyes and also tests them in different
standardized staining procedures. If a dye batch fulfils the demands a certificate is
issued stating that the batch contains the dye and that this has astated degree of
purity. A copy of this certificate accompanies the individual sampies of this batch
when sold by the producer. The physical and chemical tests perfonned by BSC
include a spectrophotometric detennination of the absorption maximum of the dye
and an evaluation of the symmetry of the absorption peak (see Fig. 3.2).
As stated above, only thin layer chromatography (TLC) or high perfonnance
liquid chromatography (HPLC) may be accepted as absolute detenninations of
purity. In addition, the BSC carries out an assay of the amount of pure dye in
the sampie. This is performed by spectrophotometry, by precipitation, or in the
majority of cases by a redox titration with titanous chloride, TiCI3.
3 Reagents 65
The chief use of enzymes is for the hydrolytic extraction of specific molecules in
tissue sections. Many of the same considerations as those discussed in relation to
the demonstration of enzyme activity (see Chaps.23-25) also apply in this context
The result of the enzymatic extraction is assessed by comparing the staining results
before and after the enzyme extraction. The result depends partly on the fixation
method and partlyon the choice of staining reaction.
As with other procedures, fixation should, as far as possible, preserve both the
concentration and the localization of the substrate prior to enzyme extraction. The
fixative should not limit access of the enzyme to the substrate or inhibit the enzyme
itself. Finally, the fixative should preserve the substrate in such a manner that the
enzymatic reaction products will be able to diffuse freely out of the tissue.
The method for visualization of the substrate before and after the enzyme
extraction should have a relatively high selectivity for the substrate in question.
Furthermore, the relationship between intensity of the colour arising and the amount
of substrate present in the tissue section should be as elose as possible. Any form
of differentiation of the final product must therefore be avoided, even if this means
that an aqueous mountant must be used.
66 A.P. Andersen, H. Lyon, P. Jakobsen
Although the specific characteristics of the tissue and the choice of staining proce-
dure are the central issue in demonstrating histological features, the pretreatment
and processing following staining also make important contributions to the final
result. The reactivity of all relevant tissue components must be taken into consid-
eration for each step of the process.
Dye Affinity. The affinity of a dye for a tissue element is a measure of the ability
of the dye to bind to a given tissue as opposed to staying in solution. In this
context, affinity takes into account the influences of all aspects of the staining
system (tissue section-dye-solvent-additives). An excellent review on the problems
relating to dye affinity is given by Horobin (1982a) in chapter 4 of his book. The
key factors involved in determining dye tissue affinity are reviewed in Table 4.1.
TIssue sections predominandy consist of solid material. The retained solids may
show very substantial variations in stainability dependent on their content of reac-
tive groups and the accessibility of these. Some reactive groups may be charged
(Chap.6); e.g.:
H. Lyon (Ed.)
TbeolY. and Strategy in Histochemisuy
© Springer ~rlag 1991
70 A.P. Andersen, P. Prent«l, H. Lyon
Component Properties
4.2 Dye
The struetural features of a dye that are important in determining its tissue staining
properties are eonsidered below.
-so;
-COO- anionie dyes
-0-
+ .
-NH3 and denved eationie dyes
The degree of ionization will depend on pH and the dielectrie eonstant of the
solvent.
These groups include earboxyl, hydroxyl, and nitrose groups. See further Seet.7.1.
Easily polarizable dyes with a large dipole moment ean form intermolecular (van
der Waal) bonds to the tissue (Seet.4.5.4). Partieularly pronouneed polarizability and
large dipole moments are seen in dyes which eonsist of several aromatie subunits,
as is the ease for polyazo dyes. Asymmetry of the moleeule also eontributes towards
inereasing the dipole moment.
Decisive for the binding of the dye to the tissue is the ratio between its hydrophobie
and hydrophilie groups. The greater the hydrophobie or aromatie part of the dye is,
72 A.P. Andersen, P. Prent9), H. Lyon
the greater will be the tendency for a dye dissolved in water to bind to hydrophobic
areas in the tissue.
Phenol, amino, azo, and carboxyl groups can all fonn hydrogen bonds with anal-
ogous groups in the tissue. In aqueous solution the hydrogen bonds with water
molecules will nonnally be dominant unless other fonns of binding between dye
and tissue are taking place at the same time. The binding is usually stabilized by a
hydrophilic-hydrophobie interaction between dye, solvent, and tissue (Sect.4.5.5).
4.3 Solvent
4.4 Additives
Additives to the staining solution can be acids, bases, inorganic salts, organic
solvents, or organie compounds such as urea or dimethyl sulphoxide (DMSO).
Some inorganie ions such as Na+, Mg2+, and Cl- can influence the dye-tissue
affinity for anionic and cationic dyes by competing for tissue binding sites (Criti-
cal electrolyte concentration, Sect.6.1.2). Other ions including Al3+ and <;r3+ are
hydrated and can fonn complexes with dyes and/or tissue ligands. These salts have
also a pH-effect.
4 General Theory for Tissue Staining 73
Some organic additives act as detergents and consequently reduce the surface ten-
sion of the solution. Others may alter dye activity by inducing micelle formation.
The different kinds of bonds which can be formed between dye and tissue are
enumerated below. In addition, approximate sizes of bonding energy are given in
kJ/mol.
Anhydrous In water
crystals
(f = 1) (f = 80)
1. lonic attraction 800 (max.) 10
2. Covalent bond 400 400
3. Complex bond 400 400
4. Intermolecular bonds 4-40 4-40
In polar solvents, and especially in water, ionic forces are much weaker than
800 kJ/mol as expressed by the equation
F = qlq2
f X r2
where F is the attraction force between two ions with charges ql and q2 separated
from each other by the distance r. f is the dielectric constant (Sect.3.2).
For water f is 80 which reduces the attraction from about 800 (for two univalent
ions) to about 10, or of the same order of magnitude as hydrogen bonds and ion-
dipole forces (Sect.4.5.4). The latter two are of course the main intermolecular
forces in water. Between certain di- or trivalent metal ions and water molecules
the ion-dipole forces are especially strong and may result in the formation of
complex bonds where in fact the interatomic distance is comparable to that in a
salt crystal (Sect.?1).
The ratio of the radii of the cation and the anion determines how many an-
ions surround a given cation (the coordination number). The closer the ions can
approach each other the stronger the attraction. It is evident that charged groups in
macromolecules, or in dye ions and macromolecules, may interact even though they
are not directly opposed. For instance, this is the case in the cooperative bonding
which takes place between DNA and histone, and probably also in the bonding be-
tween tissue components and dye ions. To avoid the precise term, "ionic bond",the
74 A.P. Andersen. P. Prent\'!. H. Lyon
terms "salt bridge" or simply "electrostatic attraction" are frequently adopted. The
possibility for salt bridge formation will thus be determined by:
1. The charge of the dye ion
2. The charge of the reactive groups
3. The dielectric constant, ionic strength, and pH of the solvent.
In general, the hydrated macromolecules, including the majority of proteins, do
not show any greater tendency to form ionic bonds with each other than with the
small organic or inorganic ions normally present in the cells and tissues.
This bond may be considered as a transition between an ionic bond and a covalent
bond. The compound consists of a central atom to which are bound two or more
ligands, depending on the coordination number.
The central atom is usually a metal ion with a small ionic radius and high
density of charge which exerts a great force of attraction. The ligands are ions or
molecules with a free pair of electrons which are drawn towards the central atom.
The ligands are bound directly to the central atom by the joint possession of
electrons. The complex compound formed often deviates in properties (colour, solu-
bility, ionizability, reactivity) from its constituent substances. Examples of complex
bonds are:
to existing or induced dipoles, weak bonds may form between intact molecules.
These are tenned intennolecular forces or van der Waal forces.
Intennolecular forces (van der Waal forces) are divided into:
a. London forces = dispersive forces (frequently the expression van der Waal
forces is used only for the London forces)
b. Dipole bonds which are subdivided into:
1. Dipole-dipole bonds between pennanent dipoles, including the hydrogen
bond
2. Dipole-dipole bonds between pennanent and induced dipoles.
London Force or Dispersive Force. This is the weakest kind of bond between
molecules (bonding energy about 4 kJ/mol). While there is a factQ!, of r- 2 in ionic
bonds, this factor is r- 6 , for London forces, where r is the distance between the
atoms involved. The bond arises between symmetrical, neutral molecules without
any dipole moment. In these molecules, due to the asymmetric movement of the
electrons around the nucleus, there will be short-lived induced dipoles that give
rise to attractive forces. Under these circumstances involved molecules will be so
close to each other that repellent forces between the electrons will be balanced by
the attractive forces. Dispersive forces, when they are not reinforced by pennanent
dipoles (see below), will thus constitute the only van der Waal attraction between
the molecules. Due to the extreme dependence on distance (r- 6 ), dispersive bonds
are very much affected by temperature.
Dipole Bonds. Dipole bonds (bonding energy as a rule between 10 and 20 kJ/mol)
can arise between two molecules with dipole moments or between one molecule
with a dipole moment and a second in which a dipole moment can be induced,
e.g. a 7r-electron system. The force of attraction F can be expressed in the same
manner as for ionic bonds, though the dependence on distance varies according to
whether it is an ion-dipole bond or a dipole-dipole bond.
F ion-ion = ql~2
r
F ion-dipole = ql ~2
r
F dipole-dipole = f1.1~2
r
q = ionic charge; f1. = dipole moment; r = distance between charges; F = force of
attraction.
A particularly strong dipole bond is the hydrogen bond (bonding energy about
40 kJ/mol). This occurs in molecules where a hydrogen atom is covalently bound
to a strong electronegative atom, e.g. oxygen, nitrogen, or sulphur.
Atomic groups that can participate in hydrogen bonds include:
-OH, -NH2, = NH, = N-, and >C=O
In addition to their importance for solubility in water, hydrogen bonds also play
a part in the mutual association of dye molecules and in dye-tissue binding. Both
76 A.P. Andersen, P. Pren~, H. Lyon
The majority of blocking and deblocking reactions belong to one of the following
6 groups: Oxidation. reduction. formation of acid derivatives. hydrolysis of acid
derivatives. deamination, and addition processes.
5.1.1 Oxidation
The use of oxidants can be subdivided into the use of weak oxidants (e.g. iodine.
bromine. hydrogen peroxide. or a feme salt). medium strength oxidants (e.g. per-
formic acid and persulphate), and strong oxidants (e.g. permanganate and chromium
trioxide) (Table 5.1).
H. Lyon (Ed.)
Theory and Strategy in Histochemistry
@ Springer Verlag 1991
78 M.R. Barer, H. Lyon
5.1.2 Reduction
Reduction is used for several very different histochemical purposes (Table 5.2). Ex-
amples of useful reducing agents include formaldehyde, hydroquinone and sodium
or potassium borohydride. Of these, the last two are by far the strongest and are
essential where reduction of oxo groups is required.
As shown in Table 5.3, the formation of acid derivatives (ester, amide) comprises
acylation, which is performed with an acid anhydride or an acid halogenide, and
alkylation, which is performed with an alcohol (usually methanol).
5.1.5 Deamination
reversed by short treatment with a weak oxidant (Table 5.1) with a weak acid or
base, or by longer treatment with Schiff's reagent (9.7.2). Hydroxylamine blockade
may be reversed with an oxidant of medium strength (Table 5.1).
Tissue bound
group Reaction Compound formed Use Reference
Chemical Groups that Can Be Ionized. Table 6.1 lists the different chemical
groups in tissue that can be ionized detailing name, fonnula (the ionized fonn
underlined), occurrence, and approximate pKA.
Table 6.1. Chemical groups which can be ionized. The ionized form is in italics.
Group Occurrence
R'-O OH R'-O 0
"/p~/ "
/p~
/ + H
R·.:....O 0 R"-O 0
The demonstration of chemical groups that can be ionized naturally falls into
two:
• Demonstration of tissue cations
• Demonstration of tissue anions.
In 1877, according to Lison (1960), p.273, Ehrlich introduced the tenns ba-
sophil, acidophil, and neutrophil tissue components depending on their differential
staining by the so-called basic or acid dyes or a mixture thereof. Basic dyes are
now tenned cationic dyes, and acid dyes are called anionic dyes (Puchtler et al.,
1985).
Table 6.2 lists the different groups and substances that may be negatively charged.
pKAS are also given.
Two preconditions must be met in order to stain these groups:
H. Lyon (Ed.)
Theory and Slralegy in Histocbemistry
© Springer Verlag 1991
82 P. Prenw.s, M.R. Barer, H. Lyon, E. Hasselager
Table 6.3. The basophilia of different substances (modified from Lison (1960),
p.275).
Qnly very few tissue components are stainable by cationic dyes at practically
all pH-values. They possess an absolute basophilia. Examples are the granules
in mast cells (heparan sulphate pKA < 0). Allother tissue components exhibit
a relative basophilia (phosphate pKAI-2, carboxyl pKA3-5), i.e. their stainability
with cationic dyes depends upon the staining conditions. These include:
a. pH of the solution
b. Type and concentration of the dye
c. Staining time and temperature
d. Buffer type and concentration
e. Kind of treatment following staining.
Post-treatment is particularly problematic. After the dye bath it is usual 10 wash
briefty in either water or in a buffer at the same pH as the stainiftg solution. If
the wash is short the staining result is only slightly affected. If ethanol is used for
dehydration, pronounced extraction of cationic dyes frequently takes place. Use
of 2-propanol, tertiary butanol (2-methyl-2-propanol), or acetone may reduce this
problem (See also Sects.16.2 and 16.3).
Acid Protein. In those cases where the additional reactions for nucleic acids, acid
carbohydrates, and lipofuscin are negative, a weak basophilia can be due to an
acid protein. This can be confirmed with a masked basophilia or metachromasia
reaction (Sect.21.4.2).
Uric Acid or Urate Deposits. These are seen in humans with gout (tophi) and also
frequently in invertebrates. They are strongly basophilic and can only be identified
by negative reactions for all other substances.
84 P. Prent~, M.R. Barer, H. Lyon, E. Hasselager
6.1.1 Metachromasia
o
\>;;;y
11 yello w
\1
11 /
I' /
\I "
\\ //
\\ ////
orange
I\~
I"
\'
\I
\I
1\
~ 800
'\
uv IR
Fig. 6.1. Complementary colour circle showing relationship between absorbed (A) and observed
(C) colours. Ultraviolet (UV) and infrared (IR) regions.
The metachromatic spectral shift is shown for two dyes:
Safranin 0
orthochromatic
metachromatic
--t metachromatic absorption shift
Toluidine Blue
orthochromatic
metachromatic
metachromatic absorption shift
6 Staining of Macromolecules on the Basis of Charge 85
E
,.
I'
,'
I 1
I 1
I 1
I 1
I 1
I 1
I 1
, 1
I
: \1 1
I 1
I 1
, 1
I 1
I 1
I 1
I 1
I I
I
I
I
1
\I
I
!
Fig. 6.2. Absorption spectra for Toluidine Blue under different conditions. Extinction (E).
Toluidine Blue in n-butanol (1)
Toluidine Blue in water
2.7 x 10-5 mol/l (2a)
2.0 x 10-4 " (2b)
3.0 X 10-3 (2c)
around 10- 2 " (3) (heparin added)
terferenee from the nucleic acid bases as weIl as the interaetion between phosphate
groups and protein-amino groups (partieularly histones). The planar dyes (Tolui-
dine Blue, Aeridine Orange), whieh may exhibit metaehromasia, tend to interealate
between the heteroaromatie bases when staining non-denatured DNA.
As illustrated above, metaehromasia is very sensitive to interferenee. Aleohols,
detergents, relatively low sah eoneentrations, as weH as eompounds sueh as urea
and dimethyl urea, are all able to impair or abolish metaehromatie binding between
dyes and polyanions. With salts the impairment of metaehromasia is probably due to
a eompetition between metal ions and dye anions for binding sites on the polyanion.
All the other interfering eompounds reduee the surfaee tension of water and thereby
diminish the hydrophobie interaetion between dye and water.
Metaehromasia is thus the result of an ordered aggregation of dye moleeules
that brings the hydrophobie boundary between dye and water down to a minimum
(see Fig. 6.3).
6 Staining of Macromolecules on the Basis of Charge 87
Fig. 6.3. A model of metaehromatie bonding between eationie dye and polyanions
o water molecule
.) hydrophobie boundary towards water
0- hydrophobie residue
EB eation
e- anionie site
eationie dye
The Alcian Blue dyes are water soluble copper phthalocyanin compounds which
were introduced into the textile industry in 1944. Their special advantage was that,
after the cloth had been stained, they could be converted to insoluble pigments
(Monastral Fast BIue in the case of Alcian Blue 8G) by treatment with alkali.
The formula shown applies to Alcian Blue 8G (previously 8GX) from I.C.1.
(Scott, 1973a). The product contained up to 40-50% boric acid as a stabilizer and
comparable amounts of sodium chloride and sodium sulphate. These compounds
can be removed by gradually diluting a 2-5% aqueous Alcian Blue solution with
five times its volume of acetone. Pure dye is precipitated (Scott, 1972a; altematively
McAuliffe, 1983) by this procedure. In some Alcian Blue products other dyes have
been added in varying quantities.
6 Staining of Macromolecules on the Basis of Charge 89
For histochemical applications Alcian Blue sampies must fulfil the following
criteria (Scott and Mowry, 1970):
a. Solubility in water to a concentration of at least 5% w/v
b. A 1% w/v solution in 0.025 mol/l acetate buffer pH 5.7 with MgCh added
at 2.0 mol/l should not precipitate on preparation and only very slightly after
standing for 24 hours at room temperature
c. The spectrum of an aqueous solution should have a Amax of 623 nm with a
shoulder at 674 nm. A solution in ethanol has Amax at 674 nm and a shoulder
at 608 nm (Scott, 1970)
d. In paraffin sections (e.g. formalin fixed large intestine or trachea) stained with
Alcian Blue at pH 2.5 the nuclei and cytoplasm should remain unstained
Schenk (1981) gives the requirements of the Biological Stain Cgmmission for
certifying Alcian BIue. Lake (1980) discusses the possibilities for acquiring Alcian
Blue products of reasonable quality and purity.
Alcian Blue is used for the differential demonstration of acid proteoglycans
and acid glycoproteins. Selectivity is controlled by varying either the pH or the
inorganic electrolyte content of the staining solution.
Mechanism. In contrast to the majority of other dyes (SectsA.2 and 4.5), dye
binding is purely electrostatic.
In aqueous solution the Alcian Blue ions form approximately spherical dimers
with 6-8 charges. These are electrostatically isolated against further formation of
aggregates. The large ionic diameter and the high charge limit the ability of the
dye to stain protein-rich complexes such as chromatin and ribosomes.
Sulphate Containing Carbohydrates. These are often protein-bound and this may
delay or completely inhibit the binding of Alcian Blue. The problem is particu-
larly pronounced with heparin in mast cells and chondroitin sulphate in cartilage.
Deposits of calcium salts constitute a further impediment to staining in the latter
instance. The polyanion-protein complex is so dense that it impedes penetration
by Alcian Blue. The high density is due to the charge saturation of the sulphate
groups which reduces the electrostatic field and results in decreased binding of
water. Reduction of both Alcian Blue staining and metachromasia by protein can
be abolished by treatment with protease, deamination, or by adding 0.1-0.3 mol/l
MgClz to the dye solution (see Alcian BIue CEC). Sulphomucins almost or com-
pletely free of protein can show a similar increase in staining after the addition
of salt and lowering of pH. In this instance, these conditions reduce the mutual
repulsion between Alcian BIue ions and allow them to bind to polyanions at eloser
90 P. Prent~, M.R. Barer, H. Lyon, E. Hasselager
intervals (i.e. increase total dye binding). The protein blocking effect is most im-
portant when staining at pHs under 3, when carboxyl groups will be extensively
deionized.
After dehydration, carbohydrates containing carboxylic groups are far more
intensely stained with Alcian BIue than with most other cationic dyes. This is
because the high charge and high molecular weight of Alcian Blue renders the
dye only slightly soluble in ethanol and hence non-extractable. In addition, the
carboxylic group containing carbohydrates (e.g. many epithelial mucins) are only
slightly bound to basic proteins and are therefore easily accessible.
Alcian Blue can stain carboxylic groups down to approximately pH 2. ~taining
at pH 0.5-1 is therefore a reliable indication for the presence of sulphate groups.
The section is washed after staining with a suitable acid buffer solution or - al-
ternatively - the non-bound Alcian Blue can be removed by applying a piece of
filter paper before washing with water. Washing directly in water will result in
Alcian Blue being removed so slowly from the tissue that considerable staining
of carboxyl proteoglycans can take place as the carboxyl groups become ionized
when the buffer is removed.
Sensitivity. This is high. The Alcian Blue methods can demonstrate considerably
lower concentrations of acid carbohydrates than the metachromasia methods. A
further increased sensitivity can be achieved by treating the sections with diluted
alkali after staining. Alcian Blue will, as previously mentioned, be transformed
into an insoluble pigment that is called Monastral Fast Blue. This process releases
negative charges on the polyanion so that they can react with more Alcian Blue
(Scott, 1972b).
Alcian Blue CEC Method. ControHed amounts of a salt, usually MgCh, are added
to a solution of 0.05% w/v Alcian Blue in 0.025 mol/l acetate buffer pH 5.8 (Scott
and Dorling, 1965). In the absence of MgCh there will be a rather diffuse blue
staining of the tissue as weH as staining of acid carbohydrates. The addition of
small amounts of MgCh (0.1-0.2 mol/l) ensures that proteins, sialomucins, and
polycarboxyl proteoglycans are not stained. Larger amounts of MgCh cause a
progressive decrease in the staining of sulphated heteroglycans. The more sulphate a
compound contains the higher electrolyte concentration required to abolish staining.
The effect is shown in Fig. 6.4 where it is also apparent that the concentration
interval of MgCh over which dye binding shifts from maximum to zero for a
given polyanion, is quite sharp. This concentration interval gives rise to the term,
critical electrolyte concentration (CEC).
Mechanism. CEC has been investigated in experiments like the one shown in Fig.
6.4 and similar experiments with "models", i.e. "spots" of different polyanions
applied to filter paper both with Alcian Blue, other cationic dyes, and the simple
6 Staining of Macromolecules on the Basis of Charge 91
organic cation, cetyl pyridinium ion from cetyl pyridinium chloride (l-hexadecyl
pyridinium chloride)
Scott and Dorling (1965) and Scott (1973b) have on this basis put forth a hy-
pothesis for the mechanism on staining with stain solutions with added electrolytes.
A poly anion will be precipitated from an aqueous solution by the addition of larger
organic cations, e.g. alkaloids, detergents, dyes, and antibiotics. The reaction is an
exchange of ions, and by mathematical treatment of the law of mass action they
reach the conc1usion that the complete removal of bound dye and subsequent precip-
itation of the poly anion takes place inside a very narrow interval of concentration,
or more correct1y ionic strength, of the metal salt. This is in agreement with the
results of the experiments, but many of the chemical, physieal, and mathematical
suppositions are doubtful (Horobin and Goldstein, 1974).
Fig. 6.4. Alcian Blue binding to different polyanions in tissue sections as a function of MgCI2
concentration. The curves illustrate the CEC phenomenon.
Staining intensity (I), Hyaluronate (A), ONA or RNA (not protein masked) (B), Chondroitin
sulphate (C), Heparin (0).
Experiments have shown that the CEC principally depends on the type of
charged group in the polyanion, so that:
CEC(-COO-) < CEC(=Ü2P02-) < CEC(-OS03-)
For many dyes other than Alcian Blue contributions from other kinds of bonds, e.g.
polar or hydrophobie bonds, will cause deviations from the simple CEC-pattern. If
the CEC is used for comparing dyes it is worth remembering that the dye powder
should only contain the dye salt, and that the concentration of the dye ions should
be low (10- 3-10- 4 molll). Table 6.5 gives a simplified survey of some CEC-values.
92 P. Prent~, M.R. Barer, H. Lyon, E. HasseJager
Table 6.5. CEC-values (molfl MgCI 2 ) for the binding of cetyl pyridinium chloride and
different cationic dyes to selected polyanions on filter paper (modified from Scott, 1967).
Aldehyde Fuchsin was originally introduced for the demonstration of elastic fibres
(Gomori,1950a). It is prepared by dissolving Pararosanilin in 70% viv ethanol
and adding hydrochloric acid and paraldehyde (2,4,6-trimethyl-l,3,5-trioxane). The
solution is "ripened" for about 5 days by standing at room temperature. It should
be used within a few days thereafter.
N=CHCH 3
while the staining of pancreatic B-cell granules may be explained on the basis of
covalent bonding (Cole and Nettleton, 1988).
Selectivity. Aldehyde Fuchsin stains nuclei and RNA-rich cytoplasm only slightly
or not at all. At pH 1-2 an intense staining of sulphate containing proteoglycans
and sulphomucins is obtained. Elastin is also stained strongly. In addition, Alde-
hyde Fuchsin stains neurosecretory substance, several kinds of APUD cells, and
hepatitis B antigen (HBsAg). For these purposes the staining is preceded by an
oxidation of the tissue section, usually with pennanganate. While this is necessary
for neurosecretory substance, routine use of oxidation is discouraged by Mowry
(1978) who recommends a longer staining time (up to 4 hours) instead. It has now
been demonstrated (Mowry and Kent, 1988) that true Aldehyde Fuchsin (Aldehyde
Pararosanilin) is able to stain purified insulin, oxidized or not. Goodresults depend
on the quality of fixation, for which Bouin's picric acid fixative can be recom-
mended. On fonnaldehyde fixed material greatly improved results can be achieved
by an after-fixation of the non-deparaffinized sections in Bouin's fluid (Lyon and
Prent9l, 1980).
Blocking. This is as for Alcian BIue.
Staining with the cationic dye, Victoria BIue (C.1. 44045) probably involves a sub-
stantial degree of non-ionic binding. An analysis of nuclear staining from alcoholic
solution has been made by Schulte et al. (1988).
The two cationic dyes Methyl Green (C.1. 42585) and Pyronin Y (C.1. 45005) are
used together in the same staining solution. At pH 2.5-4.5 a differential staining
of DNA with Methyl Green and RNA with Pyronin is seen.
Pyronin Y
Methyl Green
94 P. Pren~, M.R. Barer, H. Lyon, E. Hasselager
Crystal Violet
Mechanism. Binding of the two dyes to nucleic acids is not purely electrostatic,
hydrophobie interactions also play a key role. Methyl Green is intercalated in the
DNA double helix by hydrophobie bonding between the aromatic ring of the dye
and the purine and pyrimidine bases of DNA. It is likely that the two charged
groups of the dye form salt linkages with a phosphate group from each DNA
strand (Scott, 1967). When Methyl Green is bound to RNA hydrophobie bonding
to the nucleotide bases may take place but the predominantly non-helical struc-
ture precludes intercalation. Pyronin Y and other planar dyes with a single charge
bind efficiently to both single and double stranded nucleic acids. On the other
hand, Methyl Green, which is non-planar and carries two charges, binds much less
efficiently to single stranded nucleic acids. Although binding to double stranded
nucleic acids is more efficient, it does not reach the same level as Pyronin. In
consequence, it is possible to determine a ratio between the two dyes where native
DNA is stained by a high relative concentration of Methyl Green while all other
nucleic acids are stained with Pyronin. A suitable molar ratio is 2.6 (H~yer et al.,
1986).
Selectivity. It is worth noting that the method employs two cationic dyes simul-
taneously. The normal roles for basophilia are invalid in these circumstances. The
dominance of Pyronin staining at pH < 2 is probably due to denaturation of the
DNA double-helix. Between pH 2 and 4, Pyronin staining is, apart from RNA,
only found in highly sulphated structures such as cartilage matrix and mast cell
granules. This rarely gives rise to interpretive problems as these structures give
rise to an orange metachromasia. At pH above 4.0-5.0 an increased staining of
protein occurs and the value of the method for the RNA demonstration becomes
very dubious. If the method is carried out at the correct pH and temperature and
consideration is given to the purity of the dyes, Methyl Green is fairly selective
for DNA. Commercial sampies of Methyl Green are, to greater or lesser degrees,
always contaminated with Crystal Violet. The staining solution should therefore be
purified by repeated extraction with chloroform before use. The same result may
be obtained by filtering an aqueous solution through polyamide (Andersen et al.,
1986). Another problem with the method is that differentiation, or extraction, is
very considerable when the standard dehydration procedures are used. Dehydration
must therefore be performed either by drying in an oven at 37°C or with I-butanol
(H~yer et al., 1986).
6 Staining of Macromolecules on the Basis of Charge 95
Cationic dyes are particularly useful for the demonstration of nuclei in eukaryotic
tissues. They are also widely used for the demonstration of bacteria and other
microorganisms.
The Mechanism and Selectivity of Nuclear Staining. The pH of the staining so-
lution should normally be between 3 and 4 in order 10 ensure complete ionization
of DNA. RNA in nucleoli and cytoplasm is also stained under these conditions, as
are mast cell granules, some goblet cells, cartilage matrix, and connective ti$sue
matrix (10 a limited extent). These latter components, however, usually show some
degree of metachromasia (Sect.6.1.1). Blocking of phosphate groups in DNA, due
to tightly associated oppositely-charged nucleoprotein amino groups, often pre-
vents adequate staining of nuclei under low pH conditions. Considerably more
intense nuclear staining may be achieved under these conditions after deamination
(Sect.9.5.1). Using a higher pH, staining intensity is increased, partly due to the
carboxylic groups in proteoglycans, glycoproteins, and proteins becoming ionized,
and partly due to a diminution of the blocking effect. Extraction of cationic dyes
during dehydration can be avoided by using acetone, tertiary butanol, or I-butanol
instead of ethanol or by placing the slides in athermostat oven at 37°C with a
desiccant such as silica gel.
Gram-Staining. This staining method was named after the Danish physician, Chris-
tian Gram, who described a method for the selective demonstration of bacteria in
tissue sections in 1884. The technique involved a differentiation step that left some
bacteria stained so that they could easily be seen on a background of decolourized
nuclei and other cell components.
The Gram method consists of the following steps:
1. Staining with Crystal Violet, a cationic dye that binds to negatively charged
components of bacterial cells
2. Iodine treatment leading to complex formation with Crystal Violet
3. Differentiation: Using 95% ethanol, the iodine-Crystal Violet complexes are
retained by Gram-positive bacteria and extracted from Gram-negative bacteria
4. Counterstaining using a second cationic dye (usually giving red or pink staining)
restains the Gram-negative organisms and has little or no effect on Gram-
positive organisms
96 P. Prent~, M.R. Barer, H. Lyon, E. Hasselager
The result of the Gram stain both in terms of dye retention and bacterial mor-
phology is of major importance in the detection and identification of bacteria. An
understanding of the mechanisms involved and familiarity with the possible range
of staining results is essential if misinterpretations are to be avoided.
Mechanism of the Gram Method. The cationic dye binds to phosphate groups in
bacterial nucleic acid. Any cationic dye that can be precipitated by treatment with
iodine, may be used. Generally, Methyl Violet or Crystal Violet are used. Gram-
positive and Gram-negative bacteria take up dye in a similar pH dependent fashion.
Staining time may be varied from 15 sec to 5 min. One minute is excellent.
Mordanting with Iodine. This makes the stain less soluble in alcohol and water so
that it is relatively resistant to extraction during the subsequent processing stages.
Lugol's solution (I 1 g, KI 2 g, in 300 ml dist. water), which is widely used, becomes
progressively more acidic on contact with air so that the amount of reactive iodine
halves over the course of 30 days. An increased iodine content makes the storage
period less critical. The dye-iodine complex is formed in a molecular ratio of 1:1
tq 1:2 and is identical for Gram-positive and Gram-negative bacteria.
Differentiation. This is achieved because Gram-negative bacteria are destained
more quickly than Gram-positive bacteria. 80-100% ethanol is recommended as,
at concentrations below 60%, destaining occurs at the same rate in both groups of
bacteria. As water dissolves the bound iodine, the specimens must not be rinsed
with water for prolonged periods following iodine treatment.
Methanol and acetone destain rapidly. These reagents are widely used in routine
bacteriology where staining of tissue sections is rarely performed and where rapid
results are required. Considerable experience is essential if over destaining is to
be avoided. The higher alcohols act too slowly but 95% ethanol gives a gentle
elution time-course allowing variation between 15 sec and 3 min depending on the
application.
d. The penneability theory: The cell walls have different compositions in the two
types of bacteria conferring different penneabilities towards alcohol, iodine,
and the iodine-dye-precipitate
The penneability theory is the most widely accepted. Gram-positive cells have
a much thicker layer of peptidoglycan than Gram-negatives (Sect.2.3.1). Removal
of the cell wall (e.g. in spheroplast preparations made by treatment with peni-
cillin) renders "Gram-positive" organisms susceptible to destaining. In addition,
peptidoglycan is probably extensively hydrated. Thus removal of water during ex-
posure to organic solvents is likely to compact the cell wall resulting in a reduced
penneability to the dye-iodine-complex.
Heating. Staining is speeded up by hearing, because the lipid content in the bacterial
wall becomes more fluid at higher temperature. The phenol in carbol Fuchsin gives
an increased lipophilia and aids the passage of the dye through lipids.
Counterstaining. Picric Acid gives good contrast in bacterial smears. For paraffin
sections Al-haematein, Methylene Blue, or Janus Green are used.
The net charge of a protein depends on the balance between ionized amino groups
and ionized carboxyl groups. Whether the groups are ionized or not depends on
the immediate environment (e.g. tissue section) and the pKAS of all the ionizable
groups (for amino groups see Table 6.6). In general, proteins are positively charged
at low pH (deionization of carboxyl groups) and negatively charged at high pH
(deionization of amino groups) (see Fig. 6.5).
Group pK A at 25°C
pH« pI
pH below 4 [ + + +
r
J
- -
pH=pl
pH approx. 6-7 [ +
I
+
I
+
pH» pI
pH over 8 [ r
Fig. 6.5 • The influence of pH on the charge of a protein (pI rv 6-7).
+ cationic site
- anionic site
The pKA -values for different ionizable groups in proteins are shown in Tables 6.1
and 6.6. It is evident that the pKA -value of an individual group may vary according
to position in relation to other groups in the protein (e.g. histidyl residues show
values between 5.6 and 7.0). Thus the primary, secondary and tertiary structure
of the protein are all relevant. Fig. 6.6 shows the relationship between ionization
and pR for protein bound carboxyl groups with pKA -values of 3, 4 and 5 and for
histidyl (pKA 7), lysyl (pKA 10) and arginyl (pKA 12) groups.
The number of ionizable groups clearly depends on the content of amino and
carboxyl group hearing amino acids in different polypeptides. Some examples are
given in Table 6.7, and the pR dependency of the net charge is shown in Table 6.8.
100 P. Pren~, M.R. Barer, H. Lyon, E. Hasselager
r(!f\\
100
I pK A
50,
o L-~~~~~~ __ ~~ __ ~ __ ~ __ ~ ____ ~ __
o 2 4 6 8 10 12 14 pH
Fig. 6.6. Per cent ionization as a function of pH for protein-bound carboxyl and amino groups
(pKAS specified). Degree of ionization in per cent (I), histidyl (bis), lysyl (lys), arginyl (arg).
Ionizable groups
Amino
Carboxyl Histidyl Lysyl Arginyl
pK A 3 4 5 7 10 12 Total
Polypeptide A 30 0 2 1 6 1 40
B 12 17 7 8 4 2 50
C 5 5 18 17 6 9 60
D 2 11 11 12 10 4 50
E 1 6 19 6 16 12 60
F 2 3 5 8 7 35 60
Table 6.8. Relationship between net charge and pH for the polypeptides listed in Table 6.7.
Polypeptide
A B C D E F
pH pI* 2.4 3.5 6.4 6.3 9.0 12.6
0 8+ 14+ 32 + 26 + 34 + 50 +
1 8+ 14+ 32 + 26 + 34 + 50 +
2 5+ 13 + 31 + 26 + 34 + 50 +
3 7- 6+ 29 + 24 + 33 + 49 +
4 19 - 7- 23 + 17 + 28 + 46 +
5 23 - 17- 13+ 8+ 18 + 43 +
6 24 - 22 - 4+ 2+ 9+ 39 +
7 24- 26 - 5- 4- 5+ 36 +
8 24- 29 - 11- 9- 3+ 33 +
9 26 - 30 - 14 - 11- 0 31 +
10 28 - 32 - 16- 15 - 6- 29 +
11 31 - 34- 19 - 19 - 13- 22 +
12 31 - 35 - 23 - 22- 20- 8+
13 32 - 36 - 27 - 24 - 25 - 6-
14 32 - 36 - 28 - 24 - 26 - 10 -
* pI = isoelectric point.
Charge
120
80
40
-120
Fig. 6.7. Relationship between pH and net charge for haemoglobin.
Compound Isoelectric pH
Lysozyme 11.0
Cytochrome C 9.5
Histones 9.0-11.0
Haemoglobins 6.0-7.0
Globulins 6.4-7.2
Myosin 6.2-6.6
Insulin 5.3-5.4
Collagen 4.0-5.0
Pepsin <I
102 P. Prentf/l, M.R. Barer, H. Lyon, E. Hasselager
A high salt concentration may suppress the binding of a dye to an ionized group (see
Sect.6.1.2). This is utilized in the demonstration of e.g. elastin and amyloid where
6 Staining of Macromolecules on the Basis of Charge 103
The principle reason for binding is almost certainly the electrostatic attraction (ionie
bond, salt link) between the charged dye and the charged tissue component.
This fundamental principle is, however, heavily complicated by other factors.
Some will reduce the dye binding - e.g.:
a. Competition with other ions in the staining solution (see critical electrolyte
concentration, Sect.6.1.2)
b. Masking of tissue groups by electrostatic interaction (e.g. histone blockade of
the phosphate groups of nucleic acids, Sects.6.1.1 and 6.2.5)
c. Steric inhibition (permeability, Sect.6.2.3). The charges in the tissue can be
placed in such a manner that the dye ions are unable to come into contact with
them. For instance, this can occur if an area surrounding the attractive charge
is electrostatically repellent to the dye. Occasionally the concentration of dye
matter is so high that this has a direct influence on the penetration of the dye.
In both cases the expression "decreased permeability" is sometimes used
Other factors can increase dye binding - e.g.:
d. The formation of bonds other than ionic bonds between dye and tissue compo-
nent. These different categories of bonds are enumerated in 4.5
Under conditions where ionic bonds are unable to assert themselves (e.g. high
content of electrolytes) or are weakened (e.g. low pH with cationic dyes) these other
weaker bonds predominate. Examples include the selective methods for demonstra-
tion of elastic fibres and methods for the demonstration of amyloid (Sects.21.5 and
21.7).
Staining experiments on gelatin with different dyes at pH 4 and pH 9 (Horobin
and Bennion, 1973) showed that many cationic dyes stained gelatin to varying
degrees at pH 4, just as many anionic dyes stained gelatin at pH 9. In both cases
staining must be attributed to the formation of "non-ionic" bonds. Horobin and
Bennion also showed that the possibilities for a dye to form non-ionic bonds in-
creases with increasing molecular weight. Since, in their study, higher molecular
weight generally meant more aromatic rings, more substituents in the rings, and
possibly more heteroatoms, it can be seen that the possibilities for the formation
of van der Waal bonds should increase in parallel (Sect.4.5.4).
M~chanism. This involves binding to positive charges in the tissue, i.e. amino
groups in proteins.
Selectivity. This depends upon pH and to some degree on the anionic dye used.
An aqueous solution of the dye is frequently used. This may give rise to a rather
low staining intensity. By adding a little acetic acid (lowers pH) to the stain, or by
dissolving the dye in a buffer at pH 4, intense staining is achieved of red blood
ceIls, eosinophil granules in eosinophils, muscle ceIls, and the cytoplasm of other
cells as weIl as various connective tissue fibres. Note also that nuclei, particularly
nucleoli, are stained by anionic dyes. This is particularly clear if staining with a
cationic or metal complex dye is omitted.
6.3.2 Mixtures of Anionic and Cationic Dyes and their Histological Use
In principle, on staining with an anionic and a cationic dye in the same solution,
e.g. Azure A-Eosin or a Romanowsky type stain, all structures with ionized groups
should be stained. By changing pH, different colour patterns may therefore be ob-
tained. This allows conclusions to be drawn concerning the content of different
molecules in the tissue. Such methods are weIl suited for the identification of dif-
ferent classes of leukocytes, intracellular parasites, and different physiological or
pathological conditions where changes arise in the amount of nucleic acids or in
their degree of conjugation (e.g. secretory phase in gland ceIls, acidophil bodies,
Mallory bodies). The majority of anionic and cationic dyes precipitate as insoluble
complexes on mixing. However, for a few dye combinations, complex formation
is less pronounced particularly when the mixture contains organic constituents (al-
cohols, acetone or, dimethylsulphoxide (DMSO)).
A commonly used combination is Azure A-Eosin which is applied in a buffered
solution containing acetone. Staining results are heavily dependent on the pH (Table
6.10) and fixation procedure used (Sect.6.2.4).
Table 6.10. Examples of staining resuIts using Azure A-Eosin.
Tissue component pR 1 2 3 4 5 6 7 8 9
The Romanowsky group of dye mixtures are very similar to Azure A-Eosin
(including: May-Grünwald, Giemsa, Wright, and Leishman). These mixtures are
particularly used for staining blood smears, but are also of value for the examination
of other cytological preparations. In addition to Eosin they contain Methylene
BIue andlor various oxidized derivatives thereof. The term "polychrome Methylene
Blue" is frequently used for oxidatively demethylated Methylene Blue.
The most important dyes which arise on oxidative demethylation of Methylene
Blue are Azure B, Azure A, and Thionin (see Fig. 6.8).
N'@SON
H_ +_H
0
H-
N/. ,/ ~
-H
The dyes are frequently sold in a solution which, regardless of the dec1ara-
tion on the label, inevitably comprises a mixture of the different dyes in varying
proportions. The dyes are usually dissolved in methanol. It should be noted that
the oxidation processes may continue in such preparations, especially if the pH is
above 6.6.
The so-called Romanowsky-Giemsa-effect, (i.e. purple staining of nuc1ei and
other structures), cannot be obtained with Eosin alone or with derivatives of Methy-
lene Blue alone, but only when an interaction between Eosin, Azure B, and tissue
components is possible. Azure B can to some extent be substituted with Azure A,
but not with any of the other thiazin dyes. Eosin (tetrabromoftuorescein) cannot be
substituted with Eosin B (dibromodinitroftuorescein) but can be substituted with
tri- or dibromoftuorescein, tetrachloroftuorescein, or with tetraiodoftuorescein (Ery-
throsin B). For a standardized Azure B Eosin procedure see Appendix A.
7 Staining Involving Metal Complex Dyes
P. Prent(J, E. Schulte
Metal complex dyes C"Dye Lakes") are frequently used as substit!ltes for cationic
dyes in the staining of chromatin. The relative stability of staining in the face of
subsequent processing steps is an important property.
Many metal ions are able to form complexes with water, ammonia, cyanide, and
various organic compounds. The resultant compounds are termed complex or co-
ordination compounds. They are principally formed by metals with a small ionic
radius that exert a particularly strong attraction for negative ions or dipoles such
as water or ammonia CO and N respectively form the negative pole). The number
of atoms which may bind to the central metal atom is given by the coordination
number which is usually twice the charge; for instance, an aluminium ion binds
six water molecules forming a so-called hexaqueous ion.
The atoms or atomic groups, which bind to the central metal atom, are called
ligands. These ligands are characterized by having a free electron pair which is
drawn towards the central atom. The bond is therefore intermediate between ionic
attraction and a covalent bond.
The binding of water molecules to aluminium in the aluminium hexaqueous
ion causes such an extensive displacement of electrons that the electron density of
the O-H bond is significantly diminished. The bound water molecules thus have a
lower pKA, and their tendency to release protons is increased.
H. Lyon (Ed.)
Theoty and Strategy in Histochemistry
@ Springer Verlag 1991
108 P. Prent~, E. Schulte
The ring-shaped complex fonned in this way is called a chelate. Stable chelates
with planar dyes are generally pentagons or hexagons.
The meta! ions used most frequently for producing meta! complex dyes are
AI3+, Fe3+IFe 2+, and C?+. The stabilities of the metal complex dyes correspond
to a certain extent to the stabilities of the corresponding aqueous ions. While a
water molecule is exchanged every minute or so in the Al- and Fe-aqueous ions,
the turnover is many hours for the Cr-aqueous ion.
This difference in the rate of ligand turnover is accompanied by corresponding
differences in the required staining time for the metal complex dyes. This indicates
that tissue binding may be partly coordinate or chelate in nature.
Table 7.1. Dyes which may form histologically useful chelates with
aluminium, iron, andJor chromium ions.
Gallein 45445
Coerulein S 45510
Chromogen I 57030
Rhodizonic acid
Alizarin 58000
Alizarin Red S 58005
Purpurin 58205
Anthrapurpurin 58255
Rubeanic acid 58055
Anthracene Blue SWR 58605
Alizarin Cyanin BBS 58610
Aluminon 43810
Solochrome Cyanin R 43820
Chromoxane Pure Blue BLD 43825
Chromoxane Pure Blue B 43830
Helio Fast Rubin BBL 60760
(Nuc1ear Fast Red)
Alizarin Blue S 67415
Carminic Acid 75470
Gallocyanin 51030
Modem Violet N 51025
Gallamin Blue 51045
Celestin Blue B 51050
Brazilein 75280
Haematein 75290
Morin 75660
Mac1urin 75240
Fluorone Black
Chromoxane Cyanine R 43820
Twenty of the 28 dyes listed in Table 7.1 contain orthodiphenol and/or semi-
ortho-quinone, 4 contain orthohydroxy-carboxyl, and the remainder have a variety
7 Staining Involving Metal Complex Dyes 109
of different ligand groups (e.g. dithioamide in rubeanic acid). These groups are
thought to participate directly in the complex.
The two most important ligand dyes, Haematein and Gallocyanin, both have
diphenol and semiquinone groups. The importance of the aromatic nitrogen-car-
boxyl group in Gallocyanin for chelation is debated.
Dyes with diphenol or semiquinone groups are weakly ionized in aqueous
solution at neutral pH. The ionization entails an increase of the dye resonance and
a shift of the absorbance towards Ion ger wavelength. The resonance is increased at
higher pH, and some of the ligand dyes are used as pH indicators (e.g. Haematein
and Alizarin).
The binding of metal aqueous ions to diphenol and semiquinone groups has
the same effect as an increase in pH, namely ionization of a phenol group. When
forming chelates most ligand dyes therefore exhibit a colour shift and an increase
in colour.
This property is exploited in histochemical reactions for tissue bound metals
(Sect.17.5.2) where the chelates are insoluble. Some ligand dyes form soluble
chelates with certain metal ions. In pure aqueous solution these dyes are mostly
weak or very weak dyes. Stain intensity is dramatically increased by chelation with
the metal ion which also substantially determines the selectivity of staining. For
instance, Chromium-gallocyanin and Chromium-haematein are both selective but
slowly acting nucleic acid stains, while chelates of iron and these two dyes are less
selective but rapid histological stains for nuclei and other structures.
The following discussion will be limited to metal complex dyes containing
Haematein or Gallocyanin.
OH OH OH
HO HO HO~O
OH LQJ
OH o OH
7.2.1 Chromium-Haematein
When boiled with chrome alum, both Haematein and Gallocyanin form stable
chelates which predominantly stain nucleic acids at low pH. Chromium-haematein
(Cr-haematein) is somewhat less selective than Chromium-gallocyanin (Cr-gallo-
cyanin) and the blue colour less bright. The staining procedure and resistance
towards subsequent acid treatment is the same as for Cr-gallocyanin.
7.2.2 Aluminium-Haematein
In general aluminium (Al) forms labile chelates, and the staining result using
Aluminium-haematein (Al-haematein) is strongly dependent on the molar ratio
of aluminium to Haematein (formerly called the mordant ratio), pH, and ionic
strength (Sect.3.2.1; Figs. 7.1, 7.2, and 7.3). Furthermore, Al-bound or free Hae-
7 Staining Involving Metal Complex Dyes 111
matein may act as anionic dyes provided pH is not too low (see Figs. 7.1 and 7.3).
An AI-Haematein solution with a low AI:Haematein ratio and a high Haematein
concentration may therefore stain basic proteins and enterochromaffin granules.
Heparin
/
Duodenal
mucin
300
.003 .03 .3 3 30 [AIJL
/'friaemateinJ
Fig. 7.1. Al-haematein staining of maeromolecules on filter paper at pH 3.5: The infiuence of
[AI]/[Haematein] molar ratio.
This ratio is indieated on the abseissa, and obtained by varying the aluminium eoncentration
and keeping the Haematein eoneentration fixed at 3.3 mmol/l. Staining intensity (I) based on
photometrie measurements (Prentj!l, unpublished).
Fig. 7.2. Al-haematein staining of macromolecules on filter paper at pH 3.5: The influence of
electrolyte concentration.
[Al] = 10 mmol/l, [Haematein] = 3.3 mmol/l. Staining intensity (I) based on photometrie mea-
surements (Prent0, unpublished).
1 DNA
2 histone
3 heparin
prc.
2 3 4 5 pH
Fig. 7.3. Al-haernatein staining of tissue constituents in histone extraeted. untreated. or DNA
extraeted duodenal sections: The influenee of pH. [Al] = 10 mmol/l. [Haernatein] = 3.3 mmol/l.
Staining intensity (I) based on photometrie measurements (Prent0. unpublished).
1 nuelei in trypsin-treated sections
2 duodenal mucin (untreated sections)
3 triehloroacetie acid-treated sections
prc. precipitate.
that, in making the desired AI-Haematein solution from a new dye batch, one may
have to experiment, especially with the oxidation step.
At pH 3.5 and with an AI:Haematein ratio of about 1 both nucleic acids, ba-
sic proteins. and various negatively charged heteroglycans and mucins are stained.
7 Staining Involving Meta! Complex Dyes 113
The now obsolete "mucihaematein" method for acid mucins (Mayer, 1891) has an
AI:Haematein ratio of about 1.1 and a pH of 3.5-4. The AI-Haematein staining of
acid carbohydrates is more reddish violet than the nuclear staining. While reminis-
cent of metachromasia, this effect is probably not due to dye aggregation but may
indicate that the binding is not due to complex formation.
At a pH below 2.5 and an approximate AI:Haematein ratio in excess of 10,
relatively selective nucleic acid staining may be obtained. Broadly speaking, a
surplus of aluminium salt modifies staining in the same way as the addition of
MgCh for example (see Critical electrolyte concentration, Sect.6.1.2). In addition
to macromolecules, inorganic constituents, especially iron and calcium deposits,
may be stained by Al-haematein. This may be due to metal substitution or may
involve direct formation of chelates between Haematein and the tissue-bound metal
ion (see Sects.17.7.1 and 17.7.4).
Table 7.2. The theoretical composition of some AI-Haematein solutions used in histology (assuming
pure reagents and total conversion of Haematoxylin to Haematein without overoxidation).
Differentiation and Blueing. Al-haematein is far less stable towards acid extraetion
than ehromium ehelates. Tissue bound Al-haematein ean therefore be extraeted
with acid (e.g. differentiation following overstaining). In contrast, Al-haematein is
eompletely stable towards water and ethanol. This is a eonsiderable advantage if
the tissue is to be mounted in a hydrophobie mountant after staining.
Before dehydration, seetions are always washed in a weak base, often tap water.
This proeess gives rise to blueing (brownish-red Al-haematein changing to clear
blue), an expression of a further ionization of the aluminium bound Haematein eor-
responding to the formation of aluminium hydroxide by the aluminium-aquoions.
In histologieal praetiee Al-haematein is often eombined with a red anionie dye,
usually Eosin. This eombination is exeeedingly weIl suited for general histologieal
studies as the nuelei appear blue against a reddish background showing variations
in intensity.
7.2.3 Iron-Haematein
Blueing. Nuclei stained with pure Ferrous-haematein become deep blue after rins-
ing in alkaline tap water while nuclei stained with Ferric-haematein (and possibly
ferric ions) show a less distinct colour change as ferric hydroxide is reddish-brown.
Brown to reddish-brown nuclear staining is the norm if the section has been treated
with an oxidizing agent such as picric acid after Ferrous-haematein. In a suitable
ratio, ferric and ferrous hydroxide are dark brown to black and this feature, com-
bined with the shift in Haematein colour, can give rise to black or nearly black
nuclei with certain Iron-haematein stains.
Two-Bath Methods. In the past a two-bath method was commonly used for "Iron-
haematoxylin" staining. For this procedure the seetion is treated with a ferric salt so-
116 P. Prent~, E. Schulte
lurion (30 min-many hours) followed by was hing and 0.5% aqueous, non-acidified
Haematoxylin solution for 30 min or longer. Tissue bound ferrie ions gradually
oxidize Haematoxylin to Haematein and the final result is a mixed FerrouslFerrie-
haematein staining. Altematively, due to the relatively high pH of the Haematoxylin
solution, the overall effeet may be that of an impregnation affeeting praetically all
structures in the tissue. The final result is achieved by an extraction, "differenti-
ation", with an acid, an acid oxidizing agent, or a ferric salt solution. Providing
fixation is adequate, most cell and tissue structures ean be seleetively stained us-
ing this procedure. Satisfactory results are, however, difficult to aehieve and the
method has been largely supereeded by more satisfactory approaehes.
A procedure in which the section is first overstained and then differentiated
is called a regressive stain (in contrast to a progressive stain). Single-bath Fe-
haematein and Al-haematein are sometimes used regressively. In this situation the
staining solution is not acidified.
Haematein can bind to many other metals besides Cr, Al, and Fe. For example,
Copper-haematein (Cu-haematein) and Lead-haematein (Pb-haematein) have some-
times been used in histological staining.
In an equimolar mixture of an aluminium salt and an iron salt, Haematein binds
predominantly to iron. This explains why Al-haematein stains tissue iron deposits.
For similar reasons zinc and eopper may also be stained, while the staining of
calcium deposits probably involves a more complex mechanism.
7.3 Chromium-Gallocyanin
With increasing pH the carboxyl group becomes ionized first followed by one
of the phenol groups; the overall charge changes from positive at pH 1.6 to negative
at around pH 4.0 (Baker, 1958, p.215).
Boiling with chromalum for between 2 and 20 min yields (1 + 1)- and (2+ 1)-
chelates between Gallocyanin and hydrated chromium ions; longer bolling times
favour the formation of (2 + I )-complexes.
The spectral absorption maximum of this Chromium-gallocyanin solution is
around 575 nm. Even after boiling for 20 min there is a small proportion of free
Gallocyanin which may cause a reddish staining (metachromasia) of sulphate con-
taining polysaccharides or a weak reddish staining of cytoplasmic proteins. Free
Gallocyanin can be easily demonstrated by thin layer chromatography (Horobin and
Murgatroyd, 1968). Attempts to prepare reagents with no free Gallocyanin result
in the formation of histochemically unsuitable Chromium-gallocyanin complexes.
We believe that the most selective reagent for staining nucleic acids is prepared as
follows (Einarson, 1951; Sandritter et al., 1963; Berube et al., 1966, and Horobin
and Murgatroyd, 1968):
Dissolve 150 mg pure Gallocyanin and 15 g KCr(S04h in 100 ml H20, boll
for 10 min, allow to cool to room temperature, filter and adjust pH to pH 1.6
with H2S04. Staining time at room temperature is 24 h (16-100 h). The active
Chromium-gallocyanin complexes can be isolated by precipitating with ammonia
(Berube et al., 1966), flushing the precipitate and redissolving it in 1 mol/l H2S04.
Many different structures have been suggested for the chelates (Einarson, 1951;
Sandritter et al., 1963; Horobin and Murgatroyd, 1968). It is highly probable that
chromium is attached to the orthodiphenol group in Gallocyanin thus forming a
hydroxyquinonimine chelate. Cr-gallocyanin would then have the same configu-
ration as the chromium-chelates of Gallamine Blue (C.I. 51045), Gallo Blue E
(C.!. 51040), and Celestin Blue (C.I. 51050) where the only differences from Gal-
locyanin are that the carboxyl groups have been changed to amides or methyl esters
and thus cannot be ionized. However, it is suggested that in the (2 + 1)-complex
aminocarboxylate chelation takes place (Marshali and Horobin, 1972). Gallamine
BIue and Celestin Blue have been repeatedly tested as substitutes for Cr-gallocyanin
(Proescher and Arkush, 1928; Gray et al., 1957).
Binding to Nucleic Acids. Cr-gallocyanin stains DNA and RNA an intense blue
colour. As polyphosphoric acid is stained in exactly the same way (Jobst and
Sandritter, 1964) the nucleic acid binding probably involves phosphate groups.
Cr-gallocyanin behaves therefore, in histochemical terms, like a cationic dye.
The association between Cr-gallocyanin and nucleic acids is highly resistant
to both ethanol and acid extraction. This property and the extended time-course
of staining (around 100 h to maximum) are both consistent with the view that
complex bonds form between C?+ and nucleic acid-phosphate. In contrast, free
Gallocyanin binds rapidly (minutes) to proteins or polysaccharides and does not
resist acid or ethanol extraction.
118 P. Pren~, E. Schulte
Selectivity. If perfonned correct1y, blue Cr-gallocyanin staining is, with the excep-
tion of keratohyalin granules, highly selective for nucleic acid phosphate groups
and for polyphosphates. The method has, however, been accused ofpoor selectivity
based on the finding that RNase-treatment sometimes (particularly after fonnalde-
hyde fixation) fails to remove all the Cr-gallocyanin positive material in the cy-
toplasm and the nucleolus. It must be remembered that fonnaldehyde is a protein
stabilizing agent (Sect.12.2.1) and may therefore hinder the free access of RNase
to ribosomal RNA. These factors should be set against the relative insensitivity of
staining to protein blockade (Shires, 1967). In contrast, when acid alcoholic fixa-
tives are used, RNA hydrolysis is nonnally complete and Cr-gallocyanin does not
stain either cytoplasm or nucleoli.
Sensitivity. The molar extinction coefficient for Cr-gallocyanin is high, and the
Gallocyaninlphosphate ratio approaches 1:1 (Kiefer,I970). The method is there-
fore very sensitive provided the staining time is maintained. The reader is cau-
tioned against reducing the staining time by incubating at elevated temperature.
Cr-gallocyanin is a complex dye, and staining results from a delicate balance be-
tween several variables including temperature. Elevated temperature combined with
the standard low pR staining conditions leads to substantial hydrolysis of RNA be-
fore it is immobilized by Cr-gallocyanin binding. The long staining time makes
Cr-gallocyanin at present unsuitable for application in a routine staining method.
Rusain and Watts (1984) have recommended a modified Cr-gallocyanin stain-
ing procedure using staining times between 5 and 10 min at pR 1.0. Although
their technique has been applied as a quantitative nuclear stain in automated cytol-
ogy, the theoretical background of the dye-substrate-interaction has not been fully
elucidated.
Quantitation. The method is weIl suited for the photometrie determination of total
nuc1eic acid eontent (cf. Seet.28.8.4). RNase treatment allows the direet determi-
nation of DNA eontent, and RNA eontent may then be ealeulated by subtraetion.
The Beer-Lambert law (Seet.3.3.2) is obeyed throughout the visible speetrum (San-
dritter et al., 1963) and the acidlaleohol resistanee is a c1ear advantage for both
precision and reproducibility.
Considerable fading takes place in stained seetions. This is c1early detectable
after 3 weeks and stabilizes at about half the original extinetion at around 12
weeks. This process occurs regardless of whether the seetions are proteeted from
light or not (Vejlsted, 1972). Both microseopy and photometry should therefore be
performed within 14 days of staining.
8 Staining Based on Reductants and Oxidants
Redox reactions, apart from those involving enzymes, are widely used in the
demonstration of miscellaneous cell and tissue components. Staining depends on
the presence of groups capable of reducing or oxidizing staining reagents. Promi-
nent examples include the demonstration of reticulin and neurofibrils using silver.
Red +=!: Ox + e-
H. Lyon (Ed.)
Theory and Strategy in Histochemistry
© Springer Verlag 199J
122 M.R. Barer, H. Lyon
Thiol cysteine, free and protein bound disulphide cystine, free and
pro tein bound
Phenol tyrosine, free and protein bound;
thyroxine; adrenaline; noradren-
aline; dopamine; melanin
Pyrrole tryptophan, free and protein
(indole) bound; serotonin
Ascorbic acid dehydroascorbic acid
Alkene unsaturated lipids, induding lipo- glycol, aldehyde
fuscin
Aldehyde induced by acid hydrolysis ofDNA; carboxyl
by oxidation of 1,2-g1ycol or 1-
hydroxy-2-amino-groups, pos-
sibly alkene
There are several very sensitive methods for the demonstration of reduetants. These
include:
1. The ferrie-ferrieyanide method
2. Silver salt methods
3. Methods using tetrazolium salts
In contrast, methods for demonstrating oxidants are, in general, not as sensi-
tive. Examples inelude, nadi (a-naphthol + dimethyl-p-phenylenediamine), leueo-
diehloroindophenol, benzidine, and DOPA. These will not be diseussed further in
this text.
Mechanism. Ferrie ions (provided by FeCh) are redueed to ferrous ions, Fe3+ -+
Fe2+, and the latter are bound by eomplex bonds to ferrieyanide:
K+ + [Fe(CN)6]3- + Fe2+ -+ K[Fez(CN)6]
The eomplex formed is very insoluble and is ealled Prussian Blue or Turnbull
BIue. The precise eomposition of the eomplex is determined by the ratio between
ferrous and ferrie ions (Seet.17.7.4). Lillie and Donaldson (1974) have shown that
ferrieyanide aets as an oxidant in alkaline solutions, but not at pH values under
7. However, ferrie chloride easily oxidizes tissue thiol groups, noradrenaline, and
serotonin at pH 2.5, the standard pH for the ferrie-ferrieyanide mixture. These re-
8 Staining Based on Reductants and Oxidants 123
actions are consistent with the redox potentials for ferric and ferricyanide ions. By
chan ging the molar ratio between ferric chloride and ferricyanide, as weH as the
reaction time, it is possible to alter the selectivity of the method. As weIl as thiol
groups, melanin, lipofuscin, ascorbic acid, serotonin, adrenaline, and noradrenaline,
also give positive reactions; demonstration of adrenaline and noradrenaline, how-
ever, depends on prior treatment with a dichromate containing fixative. Formalde-
hyde fixed material may be used for the demonstration of melanin, lipofuscin, and
serotonin.
Selectivity. The method demonstrates all groups that are able to reduce ferric
to ferrous ions in acid solution. Schmorl (1928), p.205, described the use of the
method for the demonstration of lipofuscin. Naturally, ferrous ions in the tissue
will also be demonstrated (Sect.l7.7.4).
These methods involve the reduction of silver ions yielding a fine granular black
precipitate. The methods are called argentaffin if a tissue component can reduce a
silver salt to metallic silver in the dark and in the absence of a reducing agent. If
exposure to light or addition of a reducing agent is required, then the method is
called argyrophil. Due to the electron density of silver it is possible to utilize these
methods for electron microscopy (Sect.27.2.1).
The equilibria
[Ag(NH3hl+ ~Ag+ + 2NH3
and
Mechanism. Red + Ag+ ~ Ox + Ag. In practice silver nitrate is not easily reduced
at pH 4, while at alkaline pH, the complex silver compounds are far more easily re-
duced as, for example, bases may act as catalysts in the oxidation of aldehyde. Thiol
groups reduce neither silver nitrate nor silver complexes. Argentaffin methods can
be divided into those that directly demonstrate true argentaffin structures and those
that demonstrate induced reducing groups resulting from suitable pretreatment.
Examples of the latter are PASM (periodic acid-silver methenamine) where
the aldehyde groups arising on oxidation with periodic acid are demonstrated by
reduction of silver methenamine (Sect.9.2.2) and "black Feulgen" where aldehyde
groups induced in DNA by acid hydrolysis (Sect.9.9) are demonstrated by reduction
of a suitable silver compound. Grocott's method for the demonstration of fungi
(1955) includes oxidation with chromium trioxide, silver methenamine, toning,
and fixing. The mechanism is probably identical to that proposed for PASM.
Sensitivity. This varies from low for silver nitrate to high for silver complex
methods, although the latter are never as sensitive as the ferric-ferricyanide method.
Localization. This is precise when the reaction takes place in the dark and in the
absence of reductants.
Mechanism and selectivity. The mechanism is not fully elucidated but certain
features are common to all methods. There are also a number of similarities to the
photographic process from which several of the steps derive their names. The key
steps are:
1. Oxidation
2. Reduction
3. Sensitization
4. Silver bath
5. Reduction
6. Toning
8 Staining Based on Reductants and Oxidants 125
7. Fixing
A survey of the key steps in a number of argyrophil methods is given in Table
8.2.
Oxidation. Potassium permanganate is usually the oxidant of choice, but periodic
acid and chromium trioxide can be used. The aim is to produce reducing groups
in reticulin. Possible groups involved include hydroxylysyl residues in collagen
and glycol grOUPS in the carbohydrate matrix surrounding the reticulin fibres. In
principle, all tissue homoglycans and glycoproteins can give rise to aldehyde groups
and indeed, goblet cell mucin and hepatocyte glycogen granules do become rather
black. Alkene groups in lipofuscin can also give rise to aldehyde groups. This
seems to occur particularly if the permanganate solution is not acidified With a
permanganate solution pH under 1.5, manganese is reduced from an oxidation state
of +7 to +2 instead of +4 (manganese dioxide), and the reducing step can then
be omitted.
Sensitizing. In this step the tissue components to be demonstrated are made more
"sensitive" to the subsequent silver bath. Many sensitizing reagents have been pro-
posed including: salts of uranium, iron, chromium, and silver as well as hydrogen
peroxide, iodates, iodine, acetic acid, and ammonia. Mechanisms could involve par-
tial blocking or deblocking of active groups, oxidation, reduction, and pH-induced
changes, but the exact effects are unknown. In some methods (e.g. Bodian's neu-
rofibril method), omission of this SteP makes Hule difference, while in others (e.g.
Gordon and Sweets' reticulin method) it is essential.
peripheral
reticulin oligodendroglia nerve cells nerve ends neurofibrils
Steps in the (Gordon and (DeI Rio-Hortega, astrocytes (Ramon y with processes (Bielschowski, and nerve fibres
process Sweets, 1936) 1917) Cajal, 1913; 1916) (Golgi, 1875) 1904; 1909) (Bodian, 1936)
1. Oxidation KMn0 4
2. Reduction oxalic acid
3. Sensitization iron al um NH 3 K 2 Cr 207 AgN0 3 NH 3
4. Silver bath silver diammine silver diammine HAuCI 4 + HgCI 2 1 AgN0 3 in H 2 0 Silver diammine silver proteinate
5. Reduction formaldehyde formaldehyde light light light sulphite and hydro-
quinone
6. Toning (+) + (+) +2
7. Fixing + + + +
Method for the demonstration of
1. Oxidation
2. Reduction
3. Sensitization AgN0 3
4. Silver bath silver diammine AgN0 3 in H 2 0 AgN0 3 in H 2 0 silver proteinate
5. Reduction formaldehyde sulphite and hydroquinone sulphite and
hydroquinone hydroquinone ~
6. Toning (+) + ?c
7. Fixing I:X!
+ +
1 Gold and mercury instead of silver; 2 followed by renewed oxidation in oxalic acid.
~
p::
+ = the step is part of the method; ( + ) = the step is optional; - = the step is not included.
f
8 Staining Based on Reductants and Oxidants 127
the tissue and silver ions or complexes in solution. The negatively charged tissue
groups relevant in this context include, sulphate in carbohydrates, phosphate
in nucleic acids, and carboxyl in carbohydrates and proteins. Quantitatively,
the carboxyl groups of proteins are the most important when alkaline silver
solutions are used. Scopsi and Larsson (1986b) have suggested that the silver
proteinate in Bodian's method binds to lysine residues in neurofilaments and
also in argyrophil cells such as gastrin (0) cells in the antrum and glucagon (A)
cells in pancreas. The possibility that this reaction shares a common chemical
background with unspecific immunocytochemical staining is discussed.
Toning. In this step silver precipitate is replaced by gold. As gold has a higher
potential than silver in the electrochemical series the following process occurs on
exposure to gold salts:
3Ag + Au3+ -) 3Ag+ + Au.
Clearly, the total amount of metal in the section falls as a consequence of this
reaction. Due to the differences in atomic weight the mass of the gold will be
approximately 2/3 the mass of the silver. The staining intensity is therefore reduced,
and the shift in colour virtually abolishes non-specific background staining. This
can be an advantage, but, in the reticulin methods for example, staining of collagen
(golden-brown) is so reduced that the methods cannot be used for studies of mature
collagen. In some methods, such as Bodian's neurofibril method, the reduction in
intensity is so pronounced that a further reduction step, involving gold ions or
complexes bound to the negative charges in the tissue, becomes necessary. The
gold atoms formed precipitate on the gold nuclei from the preceding gold bath. It
should be noted that the process becomes irreversible following toning. Sections
should therefore be controlled in the microscope immediately following the primary
reduction, as it is possible to remove all the silver using a solution of a ferric salt
at this stage.
Toning with palladium (0.05% hexachloropalladate in 4 mol/l HCl) for 10 min
has been suggested as a much less expensive substitute for gold (Krikelis and
Smith, 1988).
Fixing. This is the final step in the majority of argyrophil and some of the ar-
gentaffin methods. The section is treated with sodium thiosulphate (hypo) which
removes un-reduced silver or gold by complex formation.
Ag+ + 2S2Ü~- -) [Ag(S4Ü6)]3-
If this step is omitted then exposure to light may produce new precipitates by
reduction of bound metal ions, and previously stained structures (e.g. spirochaetes)
may become unstained.
128 M.R. Barer, H. Lyon
8.3.4 Autometallography
In aseries of papers Danscher and coworkers (see for instance Danscher, 1984)
have presented the concept of autometallography. As described above metallic sil-
ver catalyzes the reduction of silver ions to metallic silver in the presence of a
reducing agent. A similar catalytic activity on the reduction of silver ions is dis-
played by metallic gold, metal sulphides, and metal selenides. In autometallography
the tissue sections are covered by a photographic emulsion (silver bromide), dried,
and exposed to a chemical developer (reducing agent). The developer penetrates
the emulsion picking up silver ions released from the crystals and continues into
the tissue that now contains both Ag+ and reducing molecules. If catalysts (silver,
gold, metal sulphides, or metal selenides) are present in the tissue they become
"physically developed", meaning that they become surrounded by shells of metal-
lic silver. This reaction continues exponentially until terminated by an acid stop
bath (1 % acetic acid) before the section is transferred to the fixation bath.
The autometallographic method thus represents a technique for amplifying ac-
cumulations of gold, silver, metal sulphides, and metal selenides. The method of
Timm, described in Sect.17.7.11, is based on this principle. Amplification of gold
deposits (gold-silver methods) are used in both light and electron microscopy for
the amplification of colloidal gold as used for instance in immunohistochemistry
(Sects.26.2.4 and 27.3).
9 Staining Involving Covalent Bonds
This chapter covers histochemical methods concemed with the following reactive
groups:
1. Alkene (Sect.9.1)
2. Glycol (Sect.9.2)
3. Thiol and disulphide (Sect.9.3)
4. Aromatic and heteroaromatic cyclic ring systems (Sect.9.4)
5. Amine (Sect.9.5)
6. Guanidyl (Sect.9.6)
7. Aldehyde (Sect.9.7)
8. Acid groups (Sect.9.8)
9. Purine-N-C1-deoxyribose glycoside (DNA, Sect.9.9)
H. Lyon (Ed.)
Theory and S1rategy in Histochemistry
@ Springer Verlag 1991
130 H. Lyon, M.R. Barer, E. Schulte
This method is useful in the investigation of insoluble lipid derivatives (e.g. lipo-
fuscin).
-7 0 hO
H-C HC-C~
"- J "-
O-OH O-OH
Performic acid Peracetic acid
U sing performic acid or peracetic acid ninety per cent of the double bonds will
be converted to glycol groups. The glycol groups can then be further oxidized
to aldehyde groups by periodic acid (Sect.9.2.l)
2. Demonstration of aldehyde groups with Schiff's reagent (see Sect.9.7.1)
Selectivity. This is fairly good but DNA will also react due to acid hydrolysis of
the purine-glycoside bonds (Sect.9.9).
9.1.30z-Schiff
9.1.4 UV-Schiff
This method, where ultraviolet irradiation for 3-4 hours converts alkene to aldehyde
groups which subsequently can be demonstrated with Schiffs reagent, is far more
specific than the peracid oxidation methods. It is more fully discussed in Sect 19.6.8.
With dichromate oxidation C=C bonds are broken and complexes form with chro-
mium. These can be demonstrated with Haematein (Sect.19.6.7).
I I
H-C-OH H-C=O
I + [oJ --+- + HO
2
H-C-OH H-C=O
I I
I I
H-C-OH H-C=O
I + [0] --+- + NH 3
H-C-NH 2 H-C=O
I I
o o o o
2. Demonstration of the aldehyde groups with Schiff's reagent (Sect.9.7.1)
Selectivity and Results. Using 0.5% w/v (0.022 molll) periodic acid (pH 2.1) for 5
min atroom temperature the oxidation of 1,2-glycol and l-amino-2-hydroxyl groups
leads exdusively to the formation of aldehyde groups. Alkene groups sometimes
react, but this is probably only when partial oxidation to 1,2-glycol groups has
already taken place. Higher temperature, increased concentration, or prolonged
reaction time with periodic acid can lead to further oxidation to carboxyl groups.
Ovadia and Stoward (1971) investigated the relationship between glycol oxida-
tion by periodic acid after defined times and cellular glycogen content. Even after
an oxidation time of 12 days it was found that not more than 24 % of these groups
were oxidized. The authors believed this to be due to the formation of anionic com-
plex intermediates formed between periodate ions and the outer glycosyl residues
of the precipitated glycogen aggregates. The negative charges formed in this way
would prevent free periodate ions from penetrating into the interior of the aggre-
gates. This repulsion would increase as the amount of glycogen oxidized increased.
According to the authors this would explain the linear plateau of the oxidation-time
curve. This plateau appears after oxidation times of 5-15 min (Gahrton and Yata-
ganas, 1976; Halkrer-Kristensen and Ingemann-Hansen, 1979). Ovadia and Stoward
(1971) were further able to show that the addition of electrolytes, such as sodium
or magnesium chloride, (0.2 molll) increases the velocity and extent of oxidation
of glycogen in liver sections. They proposed that the cations decrease the repulsion
between periodate ions and the negatively-charged carbohydrate-periodate com-
plex. This interpretation fits in well with observations made on the use of critical
electrolyte concentration (CEC, cf. Sect.6.1.2).
Preformed aldehyde groups in the tissue and additional reactive aldehyde groups
induced by aldehyde, oxidative, or acid fixation can also give rise to a positive PAS-
reaction. In addition, several lipids and hydroxylysyl groups in collagen may also
react. (Pretreatment of the sections with sodium or potassium borohydride (NaBH4,
KBH4) or with glycine followed by washing in water may eliminate false positive
reactions). If the PAS-stained material is not also Sudanophilic, then lipids can be
excluded as a source of the PAS-positive reaction; if it is Sudanophilic, then partly
oxidized unsaturated lipids (lipofuscin) or glycolipids must be considered. The
9 Staining Involving Covalent Bonds 133
Blocking. Hydroxyl groups, amino groups and thiol groups are blocked by acyla-
tion (Sect.5.1.3), for example:
a. Acetylation with acetyl chloride or acetic anhydride with added anhydrous pyri-
dine
b. Benzoylation. Benzoyl chloride with added anhydrous pyridine
c. Tosylation. Aqueous p-toluenesulphonyl chloride
Primary and secondary amines react forming amides.
R
""-N-CO-R + HCI
R/
Hydroxyl (alcohol and phenol groups) and thiol groups react forming esters and
thiol esters respectively
R-OH + R-COCI ---t R-O-CO-R + HCI
R-SH + R-COCI ---t R-S-CO-R + HCI
Selectivity, Sensitivity, and Blocking. These are all as for PAS (Sect.9.2.1).
134 H. Lyon, M.R. Barer, E. Schulte
Occurrence. Proteins containing the amino acids cysteine and cystine. For example,
keratin and certain neurosecretory substances have a high content of these.
R-SH
The essence of these methods is the formation of a covalent bond between sulphur
in a thiol group and mercury in an organic chromogenic mercury compound such
as Mercury Orange or Merbromin (dibromohydroxymercurifluorescein).
9 Staining Involving Covalent Bonds 135
g
HO
R-SH +
CI-H9-@-N~N~ ~
HO
R_S_H9~N~N~ + Hel
Selectivity. The organomercurial dyes have been claimed to be specific for thiol
groups (Barka and Anderson, 1963, p.33; Lillie and Fullmer, 1976, p.238). Horobin
and Flemming (1982), however, have clearly demonstrated that Mercury Orange
also stains DNA and RNA. Cowden and Curtis (1984) found that both lipids and
proteins are stained by Mercury Orange after N-ethylmaleimide blockage of thiol
groups and have suggested that this was due to solvent partition. Horobin (1984)
concludes that the selectivity of Mercury Orange for thiol groups is questionable.
HO~ ~OH
~S-S~ + R-SH ----
~OH ~OH
R-S-S~ +S~
H
136 H. Lyon, M.R. Barer, E. Schulte
Selectivity and Results. The first step in the DDD reaction is specific, however,
the second step is only moderately selective as diawnium salts couple with all
aromatic amines and phenols. Serotonin can give a particularly strong reaction.
Mechanism. All these reagents react with thiol groups in tissue sections in a similar
manner to N-ethylmaleimide. In the case of APM the bound maleimide derivative
is then diazotized and coupled at room temperature (cf. Sect.9.4.3). On binding to
thiol the two coumarin compounds (CPM and DACM) become fluorescent (bright
blue) and no diazotization re action is necessary.
Selectivity. These methods appear specific for thiol groups. The reagents are, how-
ever, often difficult to obtain and contamination may make purification necessary.
Occurrence. Aromatic groups include protein bound phenol groups (tyrosyl), im-
idazole groups (histidyl), and indole groups (tryptophanyl) and also occur in a
number of biogenic amines, such as adrenaline, noradrenaline, dopamine, sero-
tonin, and histamine.
Demonstration. Table 9.2 gives a survey of reactions for these groups. The azo
coupling method will be described in detail because of its widespread application.
9 Staining Involving Covalent Bonds 137
The azo coupling re action for aromatic amines and phenols is used in histochem-
istry for the localization of tyrosyl, histidyl, and tryptophanyl radicals in proteins.
In addition, it can be used to demonstrate bilirubin (Sect.18.4.11) and biogenic
amines (Sect.18.4.1O) as weIl as phenols and naphthols arising from different en-
zymatic reactions in which esters of phenols and naphthols are used as substrates
(Sects.24.1.2 and 24.1.3).
The site of coupling is determined by the coupling reagent, pH, and the substi-
tution pattern of the aromatic ring.
BIocking. This can be achieved with acylation of phenol and amino groups (Sect.
9.2.1). Saponification leads to deblocking (Sect.5.1.4).
138 H. Lyon, M.R. Barer, E. Schulte
Selectivity and results. The coupling reaction is selective for phenol, indole, and
imidazole groups. The second step should be controlled by staining a parallel
without the coupling step. This will reveal how much Azure A is bound to sulphate
and phosphate groups. The method can be made selective for histidine groups by
a preceding nitration step (see Sect.9.4.4).
Mechanism. The phenol group in tyrosine is diazotized and the resultant tissue
bound diazonium salt is demonstrated with an aromatic amine or phenol:
1. Formation of diazonium salt in the tissue:
9 Staining Involving Covalent Bonds 139
N=N
HO=@-R +
+
140 H. Lyon, M.R. Barer, E. Schulte
Blocking.
a. Acylation of phenol groups (Sect.9.2.1) reversible by saponification (Sect.9.2.1)
b. Treatment with tetranitromethane
This reaction for histidyl was described by Lillie and Donaldson (1972).
Mechanism.
1. Nitration in
a. Nitric acid/acetic anhydride/acetic acid or
b. Tetranitromethane/pyridine/water
Here tyrosyl and tryptophanyl groups are nitrated so that coupling to diazonium
salts is no Ion ger possible at these sites. Histidyl groups are unaffected by this
procedure
2. Coupling with diazosulphanilic acid
3. Demonstration of sulphonic acid groups with Azure A at pH 1
Steps 2 and 3 thus correspond exact1y to the reactions described in Sect.9.4.1.
Selectivity. This is high (see Sect.9.4.1). With incomplete nitration a weak trypto-
phanyl reaction cannot be excluded.
This reaction for indole groups (tryptophanyl) was described by Glenner and Lillie
(1957).
Mechanism.
1. Coupling with p-dimethylaminobenzaldehyde (p-DMAB):
9 Staining Involving Covalent Bonds 141
©ci H
Selectivity. The first step in the reaction is dependent on one of the carbon atoms
in the pyrrole ring being unsubstituted. Hence tetrapyrrole pigments, where all the
carbon atoms are substituted, do not react (Sect.18.1.1). DMAB will, however,
condense with several different compounds; reactions with phenols and aromatic
amines are brown, yellow, or red, while blue colour is specific for tryptophanyl.
©ci H
rQYCOOH
~NH2
142 H. Lyon, M.R. Barer, E. Schulte
Adams (1957) and Olenner (1957) alm ost simultaneously described methods for
the demonstration of tryptophanyl, which differ only in detail.
Mechanism. The first step is identical to the first step in Sect.9.4.5, Le. coupling
with p-dimethylaminobenzaldehyde (DMAB); the second step consists of treatment
with sodium nitrite in hydrochloric acid (Adams) or of hydrochloric acid and acetic
acid (Olenner). This is not a nitrosation but an oxidation where a stable, more
intensely blue dye complex is formed. It is postulated that new rings are formed.
Selectivity. The blue colour is specific for protein bound tryptophan as described
under the selectivity of step one in Sect.9.4.5. The tissue morphology is frequendy
rather badly preserved. especially in Adams' modification, due to the harsh treat-
ment in step two.
Sensitivity. This is high, but not as good as with the postcoupled benzylidene
reaction.
Mechanism.
1. Oxidative deamination. The amino group is removed, possibly forming ammo-
nia, and the adjacent carbon atom is oxidized to an aldehyde group:
ox
R-CH2NH2 -f R-CH=NH -f R-CHO + NH3
The oxidative deamination can be performed with ninhydrin, alloxan, or chlor-
amine-T, the latter being optimal.
9 Staining Involving Covalent Bonds 143
Alloxan Ninhydrin
Chloramine-T (Burstone, 1959) is the sodium salt of N-chloro-p-toluenesul-
phonamide. It is used at pH 7.5 at 37°C for 6 hours.
Selectivity. Free amino acids are not present in paraffin sections. The only reactive
group corresponding to that shown in step 1 (above) is the €-amino group in
lysine. The reaction with terminal amino groups, which has been emphasized in
the literature, is in doubt. Secondary aliphatic amines should, in principle, react like
primary amines, while tertiary and aromatic amines do not react See Sects.9.2.1 and
9.7.1. regarding the selectivity of the second step 2 (demonstration of aldehyde).
R-CHzNCl z - R-CH=NCI,
'",
R-CHO
Particularly large amounts are found in certain histones, protamines, and lyso-
zyme.
Demonstration. Methods for the demonstration of the guanidyl group are based
on the classical Sakaguchi reaction in which a reddish colour develops on protein-
bound arginine after treatment with an alkaline solution of I-naphthol and sodium
hypochlorite.
I-Naphthol is oxidized by NaCIO to 1,2- and 1,4-naphthoquinones.
9 Staining Invo1ving Covalent Bonds 145
~
OH
o
2 + 4 CIO
The resultant condensation product is unstable and rapidly breaks down to a soluble
product:
The method thus poses several practical problems,notably, the solubility of the
final product in organic solvents, the weak colour, and a pronounced tendency to
fade.
These problems have, however, mostly been solved by the advent of the naph-
thoquinone-sulphonate, and 9,1O-phenanthrenequinone methods.
Even if cleavage of the cyclic product takes place between N and R, the added
barium ions ensure that the product does not diffuse as the barium compound is
insoluble.
Blocking. Specific blocking can be achieved with alkaline solutions of the following
compounds:
CHO
@-CO-CO-@ I
CHO
.-N>=
....... N
NH-R
Selectivity. The reaction is specific for guanidyl groups. The cyclic product grad-
ually breaks down and the preparations can only be considered stable for 1 hour.
R
""c=o
/
H
Aldehyde
Occurrence. Aldehyde groups are only present in very small amounts in tissue.
In young vertebrates lysinal aldehyde is found in elastic tissue arising from the
oxidation of lysine during the synthesis of elastin. The great importance of the
aldehyde group is that, unlike the ketone group, it can be induced in several different
macromolecules and lipids by a number of pretreatments (see Table 9.3. below).
9.7.1 Demonstration
Schiff Reagent. Schiff's reagent (see Sect.3.3.9) has been of such importance that
Lison (1960), p.159, has aptly called it the chameleon ofhistochemistry. According
to Kasten (1960), it reacts with aldehyde groups through its sulphinic acid groups
(carbonyl reaction) forming covalent bonds:
148 H. Lyon, M.R. Barer, E. Schulte
NH-S0 2 -CHOH-R
OH
I
H-C-SO OH
I 2
R
-
S03-
I
H-C-OH +
I
R
The sulphonic acid group must be removed at this point and this can be achieved
by increasing pH to around neutrality. Unbound Schiff's reagent/Basic Fuchsin is
normally removed by washing in running tap water. This restores the paraquinone
structure of the molecule. In critical work this can be avoided by interposing a "sul-
phite rinse" comprising a solution of potassium metabisulphite or an approximation
to Schiff's reagent without the Pararosanilin. This allows removal of a11 the non-
reacted Schiff's reagent from the section without any possibility of this forming
Pararosanilin. Altematively the section can be washed in 0.01 mol/l hydrochloric
acid.
Metal salts. Ferric ions are reduced by aldehyde groups (see ferric ferricyanide
reaction Sect.8.3.1), while silver ions are reduced to metallic silver (Sect.8.3.2).
9.7.2 B10cking
Process Product
a. Oxidation carboxyl
b. Reduction alcohol
c. Addition of cyanide hydroxynitrile
d. Addition of hydrogen sulphite a-hydroxysulphonic acid
e. Addition of arylamines Schiff's base
f. Addition of hydroxylamine oxime
g. Addition of hydrazine hydrazone
h. Addition of dimedone condensation product
b. Reductants. Aldehydes and ketones are easily reduced to primary and secondary
alcohols respectively. Sodium and potassium borohydrlde are frequently used
for this purpose
The net result is addition of hydrogen cyanide and the formation of a hydrox-
ynitrile. This blockade can be broken using periodic acid
d. Hydrogen sulphite (bisulphite) reacts with aldehydes to form a-hydroxysul-
phonic acids:
The hydrogen sulphite compound is easily broken (deblocked) with weak acid
or alkali, this also occurs after extended (2-3 hours) treatment with Schiff's
reagent.
e. Arylamines, such as aniline, can condense with tissue bound aldehyde groups
forming Schiff s bases:
_ H~C
ICH ~H3C
CH 3
HÖC 3 3
R-CHO + 2
l CH
o 0 o OH HO 0
CH
I
R
This seetion includes a diverse series of reactive groups for which the single com-
mon characteristic is a tendency for protons to dissociate leaving a negative charge.
Since their demonstration depends on this charge it is logical to consider them to-
gether. The groups comprise sulphate, sulphonate, phosphate, and carboxylate.
9.8.10ccurrence
The sulphate group, R-OS0 20-, occurs in a number of proteoglycans (e.g. chon-
droitin sulphates and heparin) and in sulphomucins. The sulphonate group,
R-S020-, is not found naturally in tissue but can be introduced into proteins
by oxidation of thiol groups and possibly also of disulphide groups.
The phosphate group is found as phosphate diester in nucleic acids and phos-
pholipids and as phosphate monoester in phosphoproteins, respectively:
152 H. Lyon, M.R. Barer, E. Schulte
0- 0-
I I
R-O-P-O-R R-O-P-OH
11 11
o o
The carboxylate group, R-COO-, occurs as a side group in proteins (asparagyl
and glutamyl) and in small amounts corresponding to the C-terminal carboxyl. It
also occurs in sialic acid (gangliosides and sialomucins), uronic acid (proteogly-
cans), and in free fatty acids Ce.g. lipofuscin, cf. Sects.18.1.2 and 18.4.1).
9.8.2 Demonstration
The demonstration of acid groups is performed using cationic dyes (Sect.6.1) and
metal complex dyes (Chap.7). These dyes are bound to the negatively charged
groups, and therefore depend on ionization.
9.8.3 Blocking
Table 9.5. Optimum hydrolysis times for the Feulgen reaction using 1 mol/1 HCI at 60°C.
Notes:
a. The required time for hydrolysis of the glycoside bonds is highly dependent on
the choice of fixative (Böhm, 1968; Deitch et aZ., 1968)
b. The use of 5 mol/l HCI at room temperature yields maximum staining intensity
after 15-20 min hydrolysis virtually independent of fixative. The phosphate
ester bonds in DNA appear less susceptible to hydrolysis and extraction is
therefore reduced. If hydrolysis is extended for 12 h, a minimal decrease in
staining occurs and this is really only clearly discemible after 18 h. HydroZysis
at room temperature is therefore strongZy recommended
The required time for hydrolysis of the glycoside bonds is also influenced by the
compactness of the nuclear chromatin. Highly condensed chromatin is less sensitive
to acid hydrolysis than decondensed chromatin (Böhm et aZ., 1968; Kiefer et aZ.,
1973; Duijndam and van Duijn, 1975). Consequently, different cell types show
different hydrolysis curves. Photometrie comparison of Feulgen staining intensities
between different cell types is therefore invalid unless staining has been performed
after the optimum hydrolysis time for each cell type to be compared.
Aldehyde demonstration. After hydrolysis, the purine bases dissociate, and free
Cl-hydroxyl groups appear on the deoxyribose molecules. These deoxyribose mole-
cules occur in both furanose and aldehyde forms:
I I
O~H O~
o H o H
I I
-o-p=o -o-p=o
I I
OCH 2 OCH 2
I I
9 Staining Involving Covalent Bonds 155
The equilibrium is strongly displaced to the left, but as the reaction with the
aldehyde reagent is irreversible (Sect.9.7.1), all Cl-atoms will eventually react.
Selectivity. The glycoside and phosphate ester bonds in DNA and RNA are hy-
drolyzed at different rates as shown below:
Purine-N-Cl-pentose Pyrimidine-N-Cl-pentose Phosphate
glycoside glycoside ester
DNA ++ o +
RNA + o ++
++ rapid hydrolysis
+ slow hydrolysis
o no hydrolysis
This means that, under appropriate hydrolysis conditions, in DNA only purine-
N-Cl-deoxyribose glycoside bonds are broken, while RNA is decomposed to nu-
c1eotides. Under routine conditions of hydrolysis RNA-fragments are extracted
from cytoplasm and nuc1eoli.
Sensitivity. Although the method is very sensitive it does not demonstrate mito-
chondrial DNA.
Localization. As phosphate ester bonds in DNA remain intact and Schiff's reagent
binds covalently to the induced aldehyde groups in "apurlnic acid", there are no
problems with diffusion of reaction products. Feulgen' s nuc1eal reaction therefore
gives quite precise localization of DNA provided that the nuc1eic acids are precipi-
tated during fixation and remain insoluble during the subsequent tissue processing.
H.Lyon
Certain aspects of living cells may be studied directly using the light microscope.
However, for most work where chemical and structural information is sought, it is
necessary to "stabilize" the cells or the tissue before the examination can take place.
This Can be achieved either by freezing the tissue or by using chemical treatments
that act upon or react with tissue components, notably protein. A detailed consid-
eration of the former approach is given in tissue processing II,freezing (Chap.ll)
while further discussion of chemical treatments can be found in tissue process-
ing III, fixation, general aspects (Chap.12) and N, fixation, applied (Chap.13).
After the stabilizationlimmobilization process the tissue must be dehydrated and
the water substituted with a suitable solid (embedding) substance. These processes
are discussed in tissue processing V, embedding (Chap.14). The special consid-
erations relevant to hard tissue (i.e. bones and teeth) require a dedicated seetion
(tissue processing VI, hard tissues, Chap.15). Finally, the treatment of seetions after
staining is discussed in tissue processing VII, posttreatment (Chap.16).
In every histochemical analysis of tissue the foHowing points must be consid-
ered:
• What is to be demonstrated?
• Which histochemical method is best suited?
• What demands do these factors place on the pretreatment of the tissue?
• Is a freezing procedure or a chemical fixation preferable?
• In the latter case, which is the optimum fixative and how should the fixation
take place?
H. Lyon (FA.)
TheOlY and Strategy in Histochemistry
© Springer ~rlag 1991
160 H. Lyon
In Figs. 10.1-5 below, the options for preparing biological material for micro-
scopic examination are presented in schematic form.
Imprint Monoeeliuar
Air drying
I
t
•
Ineubation lor demonstration 01 Chemieal fixation
!
enzymatie aetivity, antigens, ete.
i
Possible postfixation proeedure Staining
n~tI"
Hydrophilie Hydrophobie
n~tI"
Hydrophilie
mounting
Hydrophobie
mounting mounting mounting
medium medium medium medium
l t !
Light mieroseopy
Vibratome
Tissue
chopper Cryostat
t
Vapour
fixation
l
Dehydration
Incubation
Possible
postlixation proeedure
Dehydration
Epoxy resin
I
i
30 ·50nm 111 m
seetions seetions
i
Staining
Contrast
enhaneement i
,
(metal salt Dehydration
impregnation
i
Hydrophobie Hydrophobie Hydrophilie
Air drying mounting mounting mounting
medium medium medium
t i i
Light mieroseopy
Fig. 10.4. Further treatment of Vibratome, Tissue Chopper, and cryostat sections.
!
10 Tissue Processing: I. Overview 163
, t
Emb.d""
l
,
Freezing
(eryoslat)
2 - 20 flm
Paraffin
l
4-10flm
Celloldin
l
4-10flm
,
Melhaerylale
H20 miseible
1 - 5 11m
i
1 flm
Epoxy resin
I
t
30 - 50 nm
seetions seetions seetions seelions seetions seetions
Staining Contrast
,
enhaneement
(metal sal!
impregnation)
Dehydration
Hydrophobie
mounting
Hydrophilie
mounting
l
Air drying
medium medium
i
Light mieroscopy
Fig. 10.5. Further treatment of cryostat sections and of material embedded in paraffin, celloidin,
methacrylate, and epoxy resin_
11 Tissue Processing: 11. Freezing
H. Lyon (Ed.)
Theory and Slrategy in Histochemistry
© Springer Verlag 1991
166 M. M{611er, P.E. H{6yer
ameter less than 1 nm). To achieve this, cooling would have to occur at a rate of
105o C per second. Since this is not possible in practice, the technical objective is
to keep the inevitable ice crystals as small as possible.
The so-called "cryoprotective" agents provide some assistance in this regard
by raising the point of recrystallization. These agents fall into two groups: low
molecular weight compounds such as glycerol, glucose, and dimethylsulphoxide
(DMSO), which penetrate into cells, and high molecular weight compounds such
as dextran and polyvinylpyrrolidone which do not
The extra- and intra-cellular fluids both have an osmotic activity equivalent to a
0.9% w/v solution of NaCI in water (300 mOsm). As tissue is cooled (Franks, 1978),
extracellular fluid starts to freeze at slightly below -10°C. Initially, salt-free ice
crystals are formed so that the remaining salt solution becomes more concentrated
and highly osmotically active. This preferential freezing of pure water continues
down to -21°C; only then does freezing of the salt solution take place.
The increased extracellular tonicity during freezing draws water out of cells and
may lead to pronounced shrinkage. This problem can be minimized by rapid freez-
ing. The dynamics of intracellular freezing, which starts later, are more complex
and less weIl understood. In general, however, the same considerations regarding
crystal size and speed of freezing apply. The fastest freezing times that have been
achieved with biological tissue are between 5 x 1()l and 104o C per second, a factor
10 below the theoretically desirable rate. Moreover, these results have only been
achieved with exceedingly small pieces of tissue (less than 0.1 mm3 ). It is therefore
absolutely vital that small pieces of tissue should be used. In practice this means
that at least one dimension (the thickness) is held to a minimum.
Good results are facilitated by making the contact between tissue and the heat
exchange surface as intimate as possible. This may be achieved by using two
different approaches either alone or in combination:
l. Placing the tissue on a polished metal surface (copper, aluminium, or silver)
with good heat conducting properties
2. Placing the tissue in a cold liquid medium. The medium must be carefully
chosen so that it does not boil on contact with the tissue and form a mantle of
gas around the "warm" block. For instance, if a tissue block is placed directly
into liquid N2, freezing takes place slowly in spite of the low temperature, as
gaseous nitrogen insulates the tissue (The domestic phenomenon of a drop of
water surviving on a hot plate due to a cushion of steam is analogous.)
Table 11.1 reviews the practical options for freezing tissue. Direct freezing
followed by sectioning on a cryostat is the most frequently used approach. The
following procedure is recommended, partly because it is easy to perform and partly
because it has given reliable and reproducible results in numerous quantitative
cytochemical investigations also at the ultrastructurallevel (von Bülow and Hj1lyer,
1983). The arrangement shown in Fig. 11.1 is recommended.
11 Tissue Processing: 11. Freezing 167
Temperature
in cooling
Methods medium CCl Efficiency
• Suitable organic liquids for tissue immersion are pre-cooled, often in liquid nitrogen, and include
iso pentane and commercial mixtures of butane and pro pa ne in liquid form.
b n-hexane or isopentane pre-cooled in carbon dioxide ice mixed with ethanol, methanol, or acetone,
are suitable here.
C Bench- and cryostat-mounted models are available .
D
•,
,, -=-TM
.,
7
( <
, . ,,
"7 J
, , ~
'-
FM ,
L
~
~
---
absolute ethanol. The sides and the bottom of the outer vessel are surrounded
with a thick layer of insulating material, and the temperature in the n-hexane
stabilizes to between -70°C and -75°C after a few minutes.- Taking care to avoid
touching the sides of the inner vessel, a small tissue block (up to 5 x 5 x 3 mm)
is rapidly immersed in the n-hexane. After about one minute, using pre-cooled
forceps (-70°C), the block is removed quickly but carefully, and the tissue can
now either be mounted on a metal tissue holder and placed direcdy in a cryostat
chamber or kept in a freezer wrapped in aluminium foll at -70°C or lower.
There has been no detectable loss of enzyme activity in any of the systems
examined so far if the tissue is used within three days. Several of the other methods
listed (e.g. isopentane precooled with nitrogen) can give equally good results.
These procedures have, however, hitherto received minimal use in quantitative
cytochemistry. It is worth noting that, providing the freezing process occurs from
all surfaces of the block, surprisingly good morphology is achieved by freezing
with a carbon dioxide expansion cooler even when relatively large tissue blocks
are used (e.g. in qualitative diagnostic work).
This is a microtome housed in its own low temperature (usually between -16°C
and -30°C) cabinet. Very thin sections (4--6 pm) for use in histochemical and
cytochemical reactions can be made using a cryostat. In spite of freezing, structural
integrity is well preserved, at least at the light microscopic level.
Cryostat design itself is beyond the scope of this text but some of the following
points should be stressed:
1. The temperature of the cabinet should usually be between -lOoC and -30°C,
most often -17°C to -20°C. The temperature is varied in order to adjust the
hardness of the ice to the tissue hardness
2. It should be possible to cool the knife to around -70°C so that heat generated
during cutting is effectively removed from the tissue. This may be achieved by
fixing a tray containing solid CO 2 to the knife
3. The knife should have a correct1y sharpened cutting edge without curvatures,
and it should be set at the correct angle with respect to the tissue
4. The anti-roll plate should be very carefully adjusted with respect to its height
above, angle to, and distance from the knife edge
5. The cryostat should be motorized so that constant cutting speed and thereby
section thickness are ensured. It is worth noting that, even if the micrometer
screw is set for a particular pm-value, a higher speed will give thinner sections,
and vice versa (see Table 28.2). This variation also depends on the nature of
the tissue
11 Tissue Processing: ll. Freezing 169
6. The section should be transferred to slide (or cover slip) using the correct
technique
If the section is transferred to the slide with a cold brush or by placing the slide
in direct contact with the section, there is a considerable risk of mechanical injury
to the tissue. Many workers attempt to avoid this problem by cooling the slide
to cryostat temperature. Unfortunately, the supercooled water in the section forms
ice crystals as soon as room temperature is reached. Both of these disadvantages
can be avoided in the following manner: A slide held at room temperature, or
even heated to 37°C is moved in a parallel plane towards the section.When the
slide is around 2 mm from the section, the temperature gradient between the two
(50-100°C) causes the water to evaporate from the section and condense on the
knife almost instantaneously. At the same time the section, assisted by a kind of
jet-effect, adheres to the slide. The rapidly dehydrated section can now be brought
to room temperature without risk of ice crystal formation as there is no Ion ger
sufficient free water for this to occur.
11.3.2 Freeze-Drying
11.3.3 Freeze-Substitution
12.1 Definition
H. Lyon (&I.)
Theory and Strategy in Histochemistry
@ Springer Verlag 1991
172 P.E. Ht1lyer, H. Lyon, M. Mt1l11er, P. Prentt1l, B. van Deurs, E. Hasselager, A.P. Andersen
These compounds (e.g. ethanol, picric acid, chromium trioxide, mercuric chlo-
ride, platinie chloride) exert their fixative effects by denaturing protein irreversibly
(Fig. 12.1). Conformational changes open out the protein tertiary structure and ex-
pose both reactive and hydrophobie groups (Fig. 12.1). This results in further forms
of stabilization involving groups from different proteins reacting with each other
to form larger, disordered aggregates (Fig. 12.1).
In principle, coagulation can be achieved in three ways:
1. Increase in thermal motion leading to disruption of both tertiary and secondary
structure. This is achieved by heating
2. Increase in the electrostatic repulsion between different areas of the protein
(e.g. by blocking water-soluble groups). Treatment with strong acid or additive
metal salts such as mercury compounds will do this. Picric acid acts both as an
acid and as an additive
3. Removal of water from the protein and its surroundings by dehydration (e.g.
with ethanol)
In the last of these, dehydrating fixation, the aqueous phase is extracted and
substituted with the solvent. This leads to exposure of previously concealed groups
12 Tissue Processing: III. Fixation, General Aspects 173
internal crosslink
~81
82
j prote in - protein
crosslink
Fig. 12.1. Simplified view of the effects of eross-linking and eoagulant fixatives on protein
eonforrnation.
A Unfolded protein
B Native protein eonfonnation
BI Cross-linking
B2 Coagulation
whieh may then react with each other. If the process is performed at low temperature
(around -40°C or below), it is effectively a freeze-substitution (Sect.11.3.3) and
does not necessarily lead to conformational changes in proteins. In contrast, at
room temperature extensive denaturation (Sect.12.2) and coagulation take place.
The native tertiary structure of proteins is stabilized by hydrophobie/hydrophilic
interaction. In coagulant fixation, exposure of the "inner hydrophobic" areas (no-
tably glycyl, alanyl, valyl, leucyl, isoleucyl, methionyl, phenylalanyl, and tyrosyl
residues) is only slighdy reduced by interaction with similar areas in neighbouring
polypeptides (Fig. 12.1). In an aqueous solvent, this substantial residual exposure
of hydrophobie residues results in a gready increased refractive index between
tissue and water. The effect can be seen in coagulant-fixed preparations of small
transparent organisms such as fish larvae, mieroscopic crustaceans, and protozoa.
While coagulating fixation can be an advantage in light microscopy (provided
the resulting network of protein does not become too coarse), it cannot be used for
electron mieroscopy preparations. Ethanol or acetone may still be used as dehy-
drating agents but they must be preceded by additive fixation (e.g. glutaraldehyde
(+ / - formaldehyde) and osmium tetroxide). Glutaraldehyde is particularly effec-
tive and appears to work either by stabilizing proteins against the coagulant effect of
dehydrating agents, or possibly by rendering the effects of these agents reversible.
The ideal fixative and fixation process would include the following features:
a. The sampIe should be stabilized against unwanted changes due to the sub se-
quent preparative procedure
b. The re action between tissue and fixative should occur virtually simultaneously
throughout the sam pie
c. The fixation process should be fully reproducible regardless of the tissue com-
position
d. The fixative should not induce "abnormal" tissue components or extract normal
components
From the foregoing it is clear that no single fixative even approaches fulfilling
these demands. The many different fixative mixtures described in the literature
must, therefore, be regarded as an attempt to balance the desired and the un-
favourable effects of the individual components. Clearly, careful consideration of
what is to be demonstrated is essential prior to fixation.
When using any fixative one should have knowledge of its:
1. Chemistry (Sect.12.3.1)
2. Method of application (Sect.l2.3.2)
3. Reactions with different tissue components (Sect.13.1)
4. Safety and disposal problems (Sect.12.3.3)
5. Removal from the tissue (Sect.12.3.4)
6. Effects on tissue storage properties following fixation (Sect.12.3.5)
12 Tissue Processing: m. Fixation, General Aspects 175
Primary Fixatives. The reagents mentioned in Sects.12.2.1 and 12.2.2 are known
as primary fixatives. With the exception of substances that can be used as gases
(e.g. formaldehyde), fixatives are nearly always used in a defined solution compris-
ing salts and other compounds influencing pH, osmolarity, and electrode potential
(see below). The form of commercial products (e.g. degree of polymerization of
glutaraldehyde) and the presence of impurities and stabilizers (e.g. formic acid and
methanol in formalin) should be ascertained as far as possible and purification
considered. Finally, the form and extent of storage should be defined.
Fixative Mixtures. Table 12.1 surveys primary fixatives that can be used alone or
in combination with other reagents as fixative mixtures.
Table 12.1. Primary fixatives. Classification and application.
Form of application
Cross-linking
formaldehyde + + + +
glutaraldehyde + + + +
glyoxylic acid + + +
osmium tetroxide + + +
potassium dichromate + +
Coagulant
ethanol + +
picric acid + +
chromium trioxide + +
mercuric chloride + +
platinic chloride + +
trichloroacetic acid + +
Table 12.2 gives the composition of some widely used mixtures of fixatives. A
number of these (e.g. Zenker's and osmium tetroxide containing reagents) must be
prepared immediately before use as they decompose fairly rapidly.
176 P.E. H0yer, H. Lyon, M. M011er, P. Prent0, B. van Deurs, E. Hasselager, A.P. Andersen
Bouin 1.5 + + + +
Clarke + +
Helly 3.6 + + + + +
Lillie's AAF 1.5 + + + +
Zenker 2.3 + + + + +
C 2 H sOH + 0 0 + + 0 +
HgCI 2 + + + + + +
Cr0 3 + 0 + 0 0
HCHO + 0 0 +
OHC(CH 2 hCHO + 0 0
OS04 + +
K 2 Cr 2 0 7 0
AI
100 mm Hg
c
t
Fig. 12.2. Perfusion fixation of an intact animal. Inßow is through a hypodermic needle (C)
placed in the left chamber of the heart. Aorta (A). The fixative bottle is raised to give a suitable
hydrostatic pressure. Air intake (Al).
Perfusion fixation usually implies a vaseular perfusion through the heart. The
technique is popular, particularly in electron microscopy laboratories working with
smaller animals (e.g. mice and rats). In principle, the whole animal is "fixed"
although there is a pronounced tendency for the forelegs and head to become
"weH" fixed.
The major advantage of vascular perfusion is that the fixative is delivered to all
tissues at roughly the same time. Fixation takes place quicldy because the diffusion
path is reduced to half the distance between two capillaries. The denser the capillary
network for a particular organ is, the shorter the diffusion path and the greater the
potential flow of fixative and consequently the more rapid the fixation.
If only one particular organ is to be fixed then "Ioeal perfusion" is always
preferable. For example, the liver may be accessed via the portal vein and the
kidneys may be fixed by retrograde perfusion from the abdominal aorta below the
point where the renal artery branches off. Infusion pumps (e.g. Braun Melsungen)
are preferable to the bottle system for such local "miniperfusions" (Fig. 12.3.).
Examples of other hoHow cavities that may be used include the oesophagus,
intestine, renal tubules (using micropipettes), and the ventricular system in the brain.
178 P.E. Hl'lyer, H. Lyon, M. Ml'lller, P. Prentl'l, B. van Deurs, E. Hasselager, A.P. Andersen
Fig. 12.3. Arrangement for local perfusion fixation of a kidney (arterial inflow and venous outflow)
Infusion Fixation. In infusion fixation the fixative is forced into an organ through
an inlet which is then clamped. This approach can be used for fixation of the lungs
or the urinary bladder.
Fixation with Gas. This is a useful alternative in some circumstances. For example,
the respiratory system can be fixed in situ by inhalation of gaseous glutaraldehyde
(this also applies to laboratory staff!). Smears and ceIl cultures can also be fixed
with gaseous fixative in small, c10sed containers.
Aldehydes. These pose many health risks. Acrolein is particularly toxic and should
not be used unless it offers very c1ear advantages. Formaldehyde and glutaraldehyde
should always be kept in tightly sealed containers and all processing, inc1uding
cutting of blocks from fixed specimens, should take place in weIl ventilated work
areas using suitable gloves where direct handling is required. After use, aldehydes
should be collected as organic chemical waste for destruction.
Heavy Metals. Chromium, mercury, and osmium are all heavy metals constituting
potential health hazards and therefore require careful handling. Osmium, which is
used as osmium tetroxide, is strongly corrosive and its vapours are noxious to eyes
and other mucous membranes. Waste is collected as for other inorganic materials.
Protein Denaturing Compounds. These, inc1uding acetic acid, picric acid, and
ethanol, are not as hazardous as the compounds mentioned above. Concentrated
acetic acid is corrosive, while picric acid and picrates are explosive in their anhy-
drous forms, so careful storage is essential. Waste is treated as solvent waste for
the alcohols, and as organic chemical waste for the acids.
12 Tissue Processing: ill. Fixation, General Aspects 179
In general, this does not pose any serious problems (Table 12.4). The majority of
primary fixatives can be removed by thorough washing in water. An exception
is mercury chloride, which leaves 1-3 J-Lm diameter black precipitates of metallic
mercury in the tissue. These may be removed by treatment with an alcoholic so-
lution of iodine: Hg + 12 - t Hg2+ + 21- . Excess iodine will itself stain the tissue,
however, this is easily removed by washing with 70% v/v ethanol or a sodium thio-
sulphate solution (much quicker): 2S20~- + h - t S40~- + 21-. Glutaraldehyde
can only be partly removed by washing in water or aqueous buffer. The remain-
ing free aldehyde groups can, however, be blocked with glycine or reduced with
borohydride (cf. Sect.9.7.2).
Wash with:
Formaldehyde + +
Glutaraldehyde (+)
Chromium tri oxide +
Potassium dichromate +
Mercuric chloride
Osmium tetroxide + *
Ethanol + +
Acetic acid + +
Picric acid + +
+ complete rem oval; (+) partial removal; - ineffective; *50% viv
ethanol will remove osmium tetroxide.
Storage of fixed tissue allows preparation of supplementary tissue blocks and pos-
sibly the use of alternative embedding techniques (e.g. frozen sections and lipid
staining). As far as possible, the tissue should be maintained in the same state as
when the original sections were cut.
In general, fixatives are not suitable for storing tissue, and several of these
reagents should be washed out with water after fixation (Sect.12.3.4). Bouin's fluid
contains picric acid which desttoys tissue after prolonged periods and also makes
subsequent paraffin embedding and cutting of sections more difficult. It should
therefore be washed out with several changes of 70% v/v ethanol.
70% v/v ethanol is also frequently used as a storage medium, an application
that conveniently complements its use for the first step in the normal dehydration
procedure. Prolonged storage (weeks) may lead to inconvenient tissue hardening.
This effect is particularly pronounced at higher concentrations of ethanol.
180 P.E. H~yer, H. Lyon, M. M~ller, P. Prent~, B. van Deurs, E. Hasselager, A.P. Andersen
The solvent in which the fixative is dissolved is frequently referred to as the vehicle.
This is nearly always water containing buffers, electrolytes, and other compounds
(Sects.12.3.7, 12.3.8, and 12.3.9). The optimum fixative concentration for a given
application gives a balance between rapid fixation and induction of artifacts. With
the exception of dehydrating fixatives, this is not usually achieved with concentrated
fixatives as the need for control of pH, osmotic effect, etc. generally lead to the
optimum concentration being much lower.
In practice, the optimum fixative concentration is usually defined in terms of
a concentration interval within which the desired feature may be reliably demon-
strated. Thus the highest permissible concentration is often determined by the need
to retain enzyme activity (Chaps.23-25). Glutaraldehyde, which gives good resu1ts
at between 0.5 and 5% w/v, is additionally limited by its tendency to polymerize at
higher concentrations (Polymerization leads to a reduced availability of the active
monomer).
12.3.7 pH
Fixatives are generally used at between pH 1.5 and pH 7.5. Acid fixatives (Ta-
ble 12.2) dissociate iron from haemosiderin (Sect.13.7) and hydr01yze nucleic acids,
so that purines are dissociated and phosphate-ester bonds are eleaved (cf. Feulgen
reaction, Sect.9.9). The resultant extraction may, however, be avoided and good
preservation maintained if a nueleic acid precipitating compound, such as ethanol
or chromium, is added (Sect.13.3). When fixation is necessary prior to the histo-
chemical demonstration of enzyme activity, it is important to know the pH optimum
as many enzymes are particularly susceptible to denaturation at this pH.
For light microscopic examination of cells, an aldehyde fixative at pH 7.2-7.4
is normally preferred. Similarly, ultrastructural morphology is best preserved when
fixative pH is elose to 7.4.
12 Tissue Processing: m. Fixation, General Aspects 181
12.3.8 Osmolarity
Although the primary fixative has some influenee on osmolarity, the eomposition
of the vehiele is usually far more important in this regard (Table 12.5).
If a mammalian eell is subjeeted to a hypertonie environment, it will shrink,
and eonversely, in a hypotonie environment it will swell. To avoid these effects,
an isotonie vehic1e (about 300 mOsm) is generally used. Alteration of the fixative
eontent, therefore, makes the solution more or less hypertonie.
1% FA -300
I%GA - 100
0.1 molll phosphate 240
0.1 molll cacodylate 190-200
1% GA 0.1 molll phosphate 340
2% FA + 2.5% GA 0.1 mol/l cacodylate 1050
FA = formaldehyde; GA = glutaraldehyde.
Gj
.. t . . ::: :f': ....
.~ ~:
2. 3.
Fig.12.4. The spatial relationship between capillary (Ca), ground substance (dots), and a cell (Ce)
under normal physiological conditions (1.), following perfusion with fixative with low colloid
osmotic pressure (2.), or high colloid osmotic pressure (3.). Hydrostatie pressure (1), colloid
-
osmotic pressure (li). Net ttansport of fluid through the wall of the vessel (--t)
mm Hg
=t===============~=
'] I~ H I~
A v
Fig. 12.5. Hydrostatic pressure (I) and colloid osmotic pressure (li) along a capillary, resulting
in a net outfiow through the wall of the capil1ary at the arterial end (A) and a net infiow at the
venous end (V). Paracapillary circulation (P).
12.3.10 Temperature
Both penetration of the fixative and the fixation process itself (Sect.12.3.11) are
temperature dependant. This is particularly useful in perfusion fixation where the
diffusion path is already short. lf high flow rates can be ensured, room temperature
or even body temperature can be used with advantage.
Autolytic processes place an upper limit on optimum temperatures for immer-
sion fixation. lf the temperature is lowered from room temperature to 0--4°C, the
combined reduction in speed of fixation and the autolytic process gives acceptable
results. It must be stressed that, unless the specimen is very small, fixation at tem-
peratures above 4°C leads to the tissue only being suitable for a rather coarse light
microscopic diagnostic assessment. Unfortunately tissue removed at operations is
traditionally, but nonetheless detrimentally, fixed at room temperature.
Fixation should always take place at 0--4°C for histochemical work. It should
be noted, however, trat the addition of CPC (cetylpyridinium chloride) to a con-
centratiön of 0.5% w/v for improved fixation of proteoglycans (Sect.13.4) leads to
precipitates in the fixative at temperatures below O°C.
12 Tissue Processing: III. Fixation, General Aspects 183
dn = -K A de
dt d dt
@TI
l..-_ _
B 03
C _,"",-0.4---'
Fig. 12.6. Diagram of tissue block illustrating the time course of fixation by immersion at a given
point in time.
A Extent of fixative penetration
B Boundary line indicating zone outside which the fixative has achieved a concentration that
aIlows cross-linking/coagulation
C Boundary line between tissue and pure fixative
I Location as yet unexposed to fixative
2 Location exposed to suboptimal concentration of fixative
3 Location exposed to optimal fixative concentration (Time insufficient for complete fixation)
4 Location exposed to optimal fixative concentration (Time sufficient for complete fixation).
rate is considerably slower in liver tissue than in a gelatin-egg albumin gel (Baker,
1966, p.24).
Cross-linking of the superficial parts of a piece of tissue do not seem to have any
appreciable effect on fixative diffusion whereas coagulation can markedly hinder
further diffusion of the fixative into the tissue (Baker, 1966).
Medawar (1941) derived the following equation for the dependence between
diffusion path (d) and diffusion time (t) from Fick's law:
d=b!i
where k is a constant, the size of which depends on the diffusion coefficient, K d, for
the fixative in question. With d measured in mm and t in hours k-values have been
measured for a number of fixatives at different concentrations partly for diffusion
into gelatin-egg albumin gel and partly into liver tissue. Some k-values measured
in this way are given in Table 12.6. Note, that k denotes the number 0/ mm the
fixative penetrates during the first hour. If, for example k = 2, the fixative will
have penetrated 20 = 2 mm after 1 hour and 2V2 = 2.8 mm after 2 hours.
Finally, it should be mentioned that problems may arise from the vehicle
(buffer) diffusing more quickly into the tissue than the fixative.
The length of the second phase (process of immobilization) depends on the
reactivity of the fixative and its concentration. Knowledge of this is rather limited.
However, it is clear that formaldehyde cross-links very slowly in comparison to
the dialdehyde, glutaraldehyde.
The influence of temperature and concentration of a fixative in solution can
be illustrated by the following observations on the times taken for fixation of the
central parts of a 3 mm thick slice of liver tissue:
N
-
~
'"
~
lfJl
S'
<!'!
....
l=
Table 12.6. Survey of diffusion rate, immobilization reaction, and total rate of fixation for some fixatives.
~
I»
t:r.
First phase kOC' Second phase Total
Concentration diffusion gelatin-egg kOC immobilization fixation ~
Fixative % rate albumin gel liver process rate References
00
VI
-
186 P.E. H(I}yer, H. Lyon, M. M(I}l1er, P. Prent(l}, B. van Deurs, E. Hasselager, A.P. Andersen
Microwaves have during the last few years been increasingly applied not only
in fixation procedures and other steps of tissue processing but also in staining
procedures (Suurmeijer et al., 1990; Kayser and Bubenzer, 1990).
Clearly one of the chief advantages of using microwaves is the much accelerated
rate at which the p?''"'<'esses take place. In fixation a discrimination should be made
between microwaVf' '.Iilbilization without the presence of chemical fixatives and
microwave-stimulateo lixation with fixatives present (Ruijgrok et al., 1990).
It is not the intell ••on in this place or elsewhere in the book to give a detailed
survey of the field of microwave application. It should, however, be understood
(Horobin and Flemming, 1990) that a thorough understanding of the physics under-
lying the processes taking place in the microwave oven is aprerequisite for being
able to fully utilize this intriguing addition to the equipment in the laboratory.
Arecent review of the use of microwaves in tissue processing has been given
by Kok and Boon (1990). Descriptions of practical applications of microwaves may
be found in Boon and Kok (1987; 1988).
13 Tissue Processing: IV. Applied Fixation
Table 13.1. The reactions of primary fixatives with nuc1eohistones, lipids, and carbohydrates.
Aldehydes e.g. nuc1eic acids not stabilized triglycerides, cholesterol, glycogen preserved
formalde- and glycosphingosides without being fixed
hyde are preserved without
and glutar- being fixed.
aldehyde Phospholipids are gra-
dually decomposed
Chromium nuc1eic acids precipitated normal fixation time gives glycol groups oxidized
trioxide in an insoluble form very little effect. Pro- to aldehydes
longed treatment leads
to fixation
Potassium nuc1eic acids not stabilized fixed after prolonged no reaction
dichromate treatment
Mercuric no stabilization complex formation with no reaction
chloride phospholipids
Osmium extensive extraction fixation takes place by no reaction
tetroxide reaction with double
bonds
Ethanol nuc1eic acids precipitated weak solvent effect g1ycogen precipitated
without being fixed
Acetic acid nuc1eic acids precipitated, preserved without being no reaction
histones extracted fixed
Picric acid nuc1eic acids not stabil- no reaction glycogen precipitated
ized, histones precipita- without being fixed
ted
Figs. 13.1 and 13.2 show the processes that are thought to occur on exposure
to the additive fixatives formaldehyde and mercuric chloride.
H. Lyon (Ed.)
TheoJ}' and Stralegy in Histocbemisrry
@ Springer Verlag 1991
188 H. Lyon, B. van Deurs, P.E. HI/lyer, P. Prent/IJ, M. MI/lller
According to Horobin (1982), p.23, the effects of fonnaldehyde (Fig. 13.1) are
largely reversed by washing in water. Puchtler and Meloan (1985), however, claim
that only loosely bound fonnaldehyde is removed even after extensive washing
(several hours). The residual bound fonnaldehyde is not dislodged by washing
extended over weeks. Before accepting the latter view it should be remembered
that the fixation process is time dependent (Sect.12.3.11). The resistance of bound
fonnaldehyde to extraction is probably dependent on the extent of cross-linking
(R-NH-CH2-NH-R in Fig. 13.1). Clearly fonnaldehyde molecules participating
in cross-links are less easily extracted by hydrolysis than those bound by single
ester linkages (R-NH-CH20H in Fig. 13.1). Cross-links are fonned progressively
over aperiod of hours-days-weeks (Sect12.3.11). The majority of the links fonned
in the early stages of fixation are not cross-links.
~ R-NH-CH 20H
~ R-NH-CH 2 NH-R
R-CHOH-CHOH-R ~ R-CH-CH-R
I I
HCHO+ ~ /0 +H 20
CH 2
R~OH
~H'
R-CH=CH-R R-CH-CHOH-R
I
(at acid PH) CH 20H
Fig. 13.1. Reactions involving fonnaldehyde (modified from Horobin, 1982). In the presence of
excess water (washing) all the reactions leading to net production of water are ful1y reversible
(cf. Table 12.4)
R_
R-
NH --. R_
R-
N-HgCI
HgCl z + HO HgCI
R-O-CH=CH-R --. I
R-O-CH-CH-R
I
HgCI
Fig. 13.2. Reactions involving mercuric chloride (modified after Horobin, 1982). All are irre-
versible (cf. Table 12.4)
ondary structure of the protein). As weIl as being time and concentration dependent,
the protein fixing effect is also sensitive to the relative abilities of different alde-
hydes to form cross-links under the same conditions. For example, the reaction
with formaldehyde is largely reversible and cross-links tend to be hydrolyzed with
prolonged washing in water. In contrast, the cross-links formed after fixation with
glutaraldehyde are considerably more stable.
In aqueous solutions of formaldehyde, a number of equilibria are established
(Jones, 1973):
HzO
H fast H" /OH
"'-C=O ---=::... C
~
I. H / H/ "OH 11.
l H+
H'-.... /OH z
H C
H/ "'-OH
-HzO
H "'- + /OH H +
---=::... "'-C=OH
/C ~
H H/ 111.
At low pB, m, (CBz+ OB, the carbonium ion), wbich is chiefty responsible
for the fixative properties of fonnaldehyde, predominates. Thus the fixing activ-
ity in neutral, buffered fonnaldehyde is reduced. Glutaraldehyde fonns a similar
methylene glycol derivative.
The mercuric ion reacts particularly with thiol groups, but also with carboxylic
and hydroxyl groups (see Sect.9.3.2 and Fig. 13.2).
Carbodiimides, R-N=C=N-R', react with a carboxyl and an amino to give a
peptide. Tbis reaction has frequently been employed for linking small peptides to
larger proteins but seldom for cross-linking soluble proteins in a fixation procedure
(Pearse, 1980, p.107).
3.6% w/v Formaldehyde Solution. When this fixative is applied for 12-24 hours,
it offers a number of advantages for histochemical purposes. The non-histones in
chromatin are fixed while the nucleohistones are not fixed, but are preserved in situ
in the chromosomes. As the fixative is a non-electrolyte the bonds between nucleic
acids and histones are not broken and there is litde or no hydrolysis in the fixative.
Some extraction of RNA and DNA does, however, take place in the peripheral
areas of the sampie.
Clarke's Fixative. Clarke's fixative (ethanol + acetic acid) gives reasonably good
fixation of DNA as it has a strong coagulant effect on nucleoproteins and precipi-
tates DNA. Nuclear morphology is well preserved.
Carnoy's Fixative. This fixative corresponds to Clarke's fixative with the addition
of chloroform. The chloroform facilitates fixative penetration by dissolving tissue
lipids. As with Clarke's fixative the nucleic acids remain sensitive to RNase and
DNase treatment. It should be noted that both fixatives give a very coarsely coag-
ulated cytoplasm with destruction of all organelles. The ribosomes together with
their content of RNA are trapped within the protein net.
Flemming's Strong Fluid. Flemming's strong fluid (Cr03, OS04, acetic acid) is
a chromic acid containing solution (pH 1.4) and is therefore an excellent fixative
for nucleoproteins. Nucleic acids tolerate subsequent processing very well and,
even after treatment with hydrochloric acid (Feulgen), DNA is not extracted to
a significant degree. The cytoplasm coagulates in a finely meshed network and
there is no dislocation of the ribosomes. Flemming's strong fluid is well suited
for histologic investigations of DNA and RNA but its somewhat drastic effects on
other cell components preclude its use for finer histochemical investigations.
Lillie's AAF. Lillie's AAF (ethanol, acetic acid, formaldehyde) is an excellent fix-
ative for nucleoproteins. It precipitates nucleic acids but the formaldehyde ensures
that coagulation of the cytoplasm is far finer than with Clarke's fixative so that
localization of RNA is much improved. RNA is, however, easily lost during the
subsequent treatment.
Bouin's Fixative. Bouin's fixative (picric acid, formaldehyde, acetic acid) directly
dissolves nucleic acids.
192 H. Lyon, B. van Deurs, P.E. Hl1Iyer, P. Prentf6, M. Ml1Iller
Glycogen. The glycogens that occur in animal tissues show substantial differ-
ences in their degree of polymerization. The lower molecular weight varieties are
both more soluble in aqueous media and more difficult to fix. Glycogens formed
in the different glycogen storage diseases are generally of low molecular. weight
(Sect.31.10.8).
In extreme cases, poor fixation of glycogen leads to complete extraction. Slightly
improved fixation leads to pronounced artifactual locallzation -\vith the glycogen
polarized according to the diffusion path of the fixative through the tissue. This so-
called streaming artifact, is especially pronounced for fixatives containing ethanol
used at room temperature.
The mechanism in "fixing" of glycogen is unknown but physical retention in
a network of fixed protein is highly probable. With glycogens of low molecular
weight it can be advantageous to coat the sections of fixed tissue with nitrocellulose
prior to staining.
The majority of formaldehyde containing fixatives give poor retention of low
molecular weight glycogens (e.g. in muscle cells and placenta) and improved results
with bigher molecular weight deposits (e.g. in liver). Bouin's fixative (picric acid,
formaldehyde, acetic acid) at 4°C gives reasonably good preservation of glycogen
but does not avoid the streaming artifact. Good results can be acbieved using freeze
substitution with Lison's "Gendre fluid" (ethanol, picric acid, formaldehyde, acetic
acid) at -73°C. If small tissue blocks (less than 1 mm3 ) are used comparable results
can be obtained with absolute ethanol at -73°C. Finally, freeze-drying followed
by fixation with formaldehyde gas can give good results in some tissues.
Proteoglycans. Proteoglycans are very highly soluble in water and to some degree
also in 70% v/v ethanol. They can be precipitated with absolute ethanol but remain
partly soluble in water.
Proteoglycans can be fixed with lead salts (Lillie and Fullmer, 1976, pA8).
The introduction of aqueous formaldehyde solution with added cetylpyridinium
chloride (CPC) (Williams and Jackson, 1956), however, represented a considerable
advance. CPC binds to the carboxylic and sulphate groups in proteoglycans lead-
ing to a precipitation effect which is combined with the protein fixing effect of
formaldehyde. 50% v/v ethanol with added 5-amino-acridine chloride (Williams
and Jackson, 1956) also fixes proteoglycans by a similar mechanism. CPC contain-
ing fixatives should not be stored or used below O°C (cf. Sect.12.3.10).
It should, however, be noted that after a study of the effect of different fixatives
on Alcian Blue staining of glycosaminoglycans, Tuckett and Morriss-Kay (1988)
13 Tissue Processing: N. Applied Fixation 193
discouraged the use of CPC. Due to its detergent effect CPC causes cell mem-
brane disruption and vesicle formation. At the light microscopic level, stained ceH
membrane-associated glycosaminoglycans appear dissociated from the cell surface
in the extraceHular matrix. In contrast, in an electron microseopie study of extra-
ceHular matrix using high iron diamine (HID) (Sect.22.2.2), Albedi et al. (1988)
noted irnproved preservation of sulphated glycosaminoglycans at the ultrastructural
level when CPC was added to the fixative.
Sections of fresh frozen tissue should be used for lipid histochemical investigations.
The reactions should be performed without delay after cutting as auto-oxidation of
the fatty acids rapidly leads to significant changes in the physico-chemical proper-
ties of the lipids. Methods relying on oxidation of double bonds (OS04, UV-Schiff)
as weH as the plasmal reaction (and essential controls) require absolutely fresh, un-
fixed seetions. Unfixed sections are also necessary for experiments involving dif-
ferential extraction of lipids. If same-day processing is not possible then changes
in the lipids can be minimized by keeping the sections sealed against oxygen and
light in deep freeze.
Several lipid reactions necessitate a preceding stabilization procedure. This
should be performed with fonnol calcium, possibly with added cadmium chloride,
and should be as short as possible.
In certain lipid reactions fixation is an integrated part of the method. This is
especially true for chromation (Sect.19.6.7) and osmium tetroxide (Sect.19.6.5)
methods. FoHowing chromation (K2Cr207/CaCh, pH 3.5-4) the greater part of the
lipids become insoluble in organic solvents, and a chromated piece of tissue can
then be transferred to paraffin, epoxy resin, or methacrylate without loss of lipid.
In a similar way osmium tetroxide (Os04/CaCh, pH 7-7.5) gives general fixation
of lipids and permits epoxy resin or methacrylate embedding of tissue if the lipid
content is not too high. Osmium fixation renders paraffin infiltration more difficult,
but can be used in combination with other fixatives before paraffin embedding. It
should be noted that free osmium tetroxide is reduced to osmium black by ethanol.
Under controHed conditions, aH the fixatives mentioned in Table 13.2 give
excellent histological preservation. In addition, the aldehyde fixatives produce vir-
tually complete dissociation of phospholipoprotein complexes and emphasize the
negative charge of the phosphoryl group. They therefore usually increase stainabil-
ity with Sudan Black, Nile Blue, and Luxol Fast Blue.
Aldehyde fixatives for use in lipid histochemistry should be of the highest
purity to avoid undesirable effects. For preference, formaldehyde solutions should
be prepared from paraformaldehyde as technical formalin contains both formic acid
and methanol which can increase the hydrolysis and extraction of lipids during
fixation. Buffered formaldehyde should be used with added calcium or cadmium
chloride as these immobilize the phospholipids. Salt concentrations above 0.1 mol!l
194 H. Lyon, B. van Deurs, P.E. Hj1lyer, P. Prentj1l, M. Mj1l11er
As enzymes are proteins they are affected by a1l fixatives but, if activity is to be
demonstrated, then the active site (Sect.23.1.2) should remain intact. Total preser-
vation using any chemical fixative must be considered impossible and, for all
quantitative and many qualitative investigations (particularly those concerned with
oxidoreductases) cryostat sections of fresh frozen tissue are essential (Sects.ll.l
and 23.2.1). Fixation has, however, been used to avoid diffusion of soluble enzymes
(Sects.24.2 and 25.1.1).
Currently formaldehyde is widely used for the fixation of hydrolases. The alter-
natives include formol calcium and paraformaldehyde in either dilute (0.05 mol/l)
phosphate or cacodylate buffer and these should be used at 0-4°C. The choice
of buffer depends on the enzyme to be demonstrated (e.g. ß-glucuronidase is in-
hibited by cacodylate buffer). If formol calcium is used, this should be prepared
13 Tissue Processing: IV. Applied Fixation 195
As most pigments are insoluble in water there is little problem with localization. The
biogenic amines do, however, constitute an exception as they are easily diffusible.
Aqueous formaldehyde is reasonably effective at fixing serotonin in enterochromaf-
fin cells (Sect.30.2.1), and the amine remains demonstrable after dehydration and
paraffin embedding. Adrenaline and noradrenaline are not normally demonstrable
in the adrenal medullary chrom affin cells after this procedure. As a role, one needs
to use freeze drying followed by treatment with formaldehyde gas for this purpose
(Sect.30.2.1). With this method the biogenic amines can also be demonstrated in
mast cells, ganglion cells, and nerve terminals.
Noradrenaline and, to some extent, adrenaline can be fixed by sequential fixa-
tion in neutral buffered glutaraldehyde solution followed by potassium dichromate
at pH 4. The chrom affin re action (Sect.18.4.5) which consists of an oxidation
of biogenic amines to coloured, insoluble products, is usually performed with a
solution of potassium dichrornate alone or with added potassium chromate. This
produces simultaneous fixation and demonstration of biogenic amines in chrom affin
and enterochromaffin cells.
14 Tissue Processing: V. Embedding
14.2 Dehydration
H. Lyon (Ed.)
TheoJ}' and Stralegy in Histocbemisrry
@ Springer Verlag 1991
198 H. Lyon, B. van Deurs, P.E. H~yer, P. Pren~, M. M~ller
in paraffin blocks after sectioning and storage (Sect.14.5). This problem may be
avoided by ensuring a sufficient number of changes of undiluted dehydrating
agent prior to the clearing process (Sect.14.3)
2. Chemical binding 0/ water. This can be achieved with a compound such as
2,2-dimethoxypropane
Acetone. Mechanism: Water extraction. Less hardening effect than ethanol, but can
make the tissue brittle-well suited for small pieces of tissue but high inflamma-
bility makes it less suitable for use in automatic tissue processing machines using
volumes of one litte.
14.3 Clearing
The use of a clearing agent is necessary when the dehydrating agent (e.g. ethanol)
is not miscible with the embedding medium (e.g. paraffin). A clearing agent should
therefore be miscible with both the dehydrating agent and the embedding agent.
There are numerous organie liquids with appropriate properties in this context
Many of these have a refractive index corresponding to that of protein and this
may lead to the tissue becoming transparent, hence the name clearing.
If ethanol has been used as the dehydrating agent and the embedding agent is
to be paraffin, the choice of a suitable clearing agent will depend on the follow-
ing factors: viscosity, boiling point, vapour pressure, miscibility with ethanol and
paraffin, effect on seetion quality, and toxicity.
The majority of clearing agents are inflammable liquids which require consid-
erable caution, especially when automatie tissue processing machines are used. On
transfer to hot paraffin, liquids with low boiling points are usually removed more
rapidly than liquids with high boiling points and, where applicable, this factor may
be accentuated by vacuum. An important exception is chloroform which has a low
boiling point but is removed considerably more slowly than xylene. The viscosity
also influences the speed with which the clearing agent penetrates the tissue.
Table 14.1 details important properties of several clearing agents.
Toluene. Toluene has similar properties to xylene. Tissue is harmed less by pro-
longed immersion. Toluene is inflammable and toxic.
Chloroform. This acts more slowly than xylene or toluene, but causes less brittle-
ness. Thicker blocks of tissue (up to 1 cm) can be used provided sufficient time is
allowed for both clearing and replacement of chloroform with paraffin. The slightest
trace of chloroform remaining after embedding leads to difficulties in sectioning.
Chloroform vapours are not inflammable but, on heating, the toxic gas phosgene
is liberated.
Benzene. This substance has properties similar to xylene but should be avoided
due to its carcinogenic properties.
Liquid Paraffin. Good results can be achieved with liquid paraffin as the clearing
agent, although the processing time will vary according to the composition of the
mixture of hydrocarbons. It is less inflammable than the agents mentioned above
and is therefore comparatively less dangerous.
Petroleum. This can be used as a clearing agent if the tissue does not contain too
much lipid. Particular advantages are reduced fire hazard and toxicity.
Amyl Acetate. This agent is seldom used, but removes ethanol fairly quickly and is
itself removed with reasonable speed by melted paraffin. It has a rather unpleasant
smell and is expensive.
Methyl Benzoate. This clears tissue at a moderate rate and causes very litde change
in the tissue. Due to its penetrating smell it is, however, not used very frequendy,
but can be of value when making museum preparations.
Cedarwood Oil. This slowly acting agent causes litde hardening and tissue shrlnk-
age. The need for longer processing time in the melted paraffin, however, outweighs
its otherwise mild effects. Due to its rather low volatility cedarwood oil is less ef-
fectively removed by vacuum treatment than the previously mentioned clearing
agents. It is expensive.
Clove on. This has similar effects to those of cedarwood oil but is even more
expensive.
14 Tissue Processing: V. Embedding 201
Agar. An aqueous solution of agar can be valuable with very brittle tissues or with
exudates (e.g. examination of contents of lipid and oil in pulmonary exudates).
Paraffin. Although there are other embedding media that have melting points and
other properties that cause less damage to tissue, paraffin is still the most frequently
used. This is largely due to the ease with which a large number of tissue blocks
can be prepared in a relatively short time (see water soluble resins Sect.14.4.2). In
addition, sectioning and subsequently staining pose fewer difficulties than most of
the other media amongst which paraffin is virtually the cheapest.
Paraffin is a mixture of hydrocarbons with overall properties that vary consider-
ably. The melting point is in the region of 40-70°C. At any particular temperature
below the melting point, paraffin with high melting point will be harder than a
corresponding paraffin with lower melting point. In order to prepare ribbons of
sections on a rotary microtome it is necessary to choose a paraffin that is reason-
ably hard at room temperature. In general, a melting point between 54-58°C gives
reasonable results.
Prolonged heating to temperatures around 10-20°C above the melting point
leads to evaporation of some of the fractions with lower melting points. The paraf-
202 H. Lyon, B. van Deurs, P.E. Hjilyer, P. Pren~, M. Mjilller
fin develops a more soap-like consistency and the previously crystalline structure
more or less disappears. Deliberate treatment of paraffin in this way has been
recommended in order to facilitate production of ribbons. A number of different
substances have also been added in order to modify the consistency. The aim has
been partly to increase the hardness of the paraffin so that thinner sections could be
cut, partly to give the necessary support to the tissue, and partly to improve forma-
tion of ribbons. Many compounds have been suggested but mention of beeswax,
which should result in better adherence between the sections, will suffice here. Lat-
terly the addition of selected plastics to commercial paraffins in order to increase
hardness has become common.
This places special demands on the embedding medium as very thin sections (20-
60 nm) are necessary in order to benefit from the resolving power of the electron
microscope. Such thin sections are made possible partly by reducing the section
area and partly by embedding in considerably harder materials than those used
for light microscopy. The embedding media used are plastics which may be water
soluble or water insoluble.
Plastic. Plastic is a polymer, which, in the present context, is also termed aresin.
The majority of the water insoluble media fall into three main groups:
1. Acrylic
2. Polyester
3. Epoxy
while the water soluble media are derivatives of epoxy or methacrylate. The tissue
is infiltrated with a monomer which is transformed into the plastic (polymer) on the
addition of a hardener and an accelerator. The monomer should ideally have a low
molecular weight, low viscosity, be stable, and polymerize uniformly (preferably
at room temperature) with a minimum of shrinkage.
The resultant polymer must be transparent, have a very small grain size, and
adhere well to tissue so that the latter is not tom out during sectioning. Further, the
polymer should, as far as possible, be indifferent to the heating effect of the electron
beam. The monomer should not influence or alter the tissue chemically and, with
regard to possible histochemical reactions, it is advantageous for the polymer to
14 Tissue Processing: V. Embedding 203
be water soluble. Last, but not least, the embedding material should not be toxic.
Unfortunately none of the embedding materials for electron microscopy meet all
of these demands and none of the classical media used for light microscopy (see
Sect.14.4.1) are suitable.
Polyester Resins. These are fairly good embedding media for electron microscopy
but they are relatively viscous before polymerization. Acetone is preferred to
ethanol for dehydration due to the low hydrophilicity of the medium.
Epoxy Resins. The epoxy resins (Epon 812 and others) comprise a large group of
very frequently used embedding media with a wide range of properties (molecular
weight, viscosity, hydrophilicity, toxicity). They are all di-epoxides, Le. they have
two epoxy rings.
Resin Viscosity
Araldite 2500
Epon 812 (A:B = 1:1) 1400
Vestopal W 900
Vinylcyclohexene dioxide (VCD; Spurr Plastic) 60
Diglycidyl ether (DGE) 23
1,2,7,8-diepoxyoctane (DEO) 20.5
Several newer epoxy resins with low viscosity have been developed (VCD,
DGE, DEO, see Table 14.2). These resins are weIl suited for the infiltration of
larger pieces of tissue and they are also easy to cut on the ultramierotome.
Strict regulations apply to all work with epoxy materials and their waste prod-
ucts and these should be closely adhered to. The health hazard associated with
epoxy resins decreases with polymerization providing the components have been
used in weIl balanced amounts. Allergie reactions can, however, be elicited by
hardened material (try to avoid too many shavings during sectioning). The aro-
matic amines whieh comprise part of the hardener are particularly toxie.
Water Soluble Resins. Finally, the water soluble resins must be considered. These
are used to avoid extraction of certain substances, such as lipids, during dehy-
dration. They are also used when resins with hydrophilie properties which permit
histochemieal staining, particularly enzymatic reactions, are required. Examples of
such water soluble resins include glycolmethacrylate (2-hydroxyethyl methacry-
late), the epoxy products Durcupan and Aquon, and the relatively new products
JB-4 (Polysciences) and EFL-67 (Serva). It should, however, be noted that at least
some of these resins (glycolmethacrylate) are themselves very strong organie sol-
vents.
There is little doubt that the water soluble resins will be much used in the
future (plastie seetions), as, with better preservation of structure than paraffin, they
14 Tissue Processing: V. Embedding 205
Most embedding media will either crystallize or harden further on storage. The
process is accelerated by heat and light, and, in paraffin, by the presence of reagents
such as water or clearing solvents. Tissue blocks are therefore often too hard or
too brittle if sectioning is delayed. Dry and cool storage in darkifess is generally
recommended. Paraffin blocks which have been sectioned should be sealed by
dipping the cut surface in melted paraffin.
15 Tissue Processing: VI. Hard Tissues
H. Lyon
The hard tissues, which include bones and teeth, pose special p1'Qblems in bisto-
logical technique due to their content of crystals of mineral salts. Bone may be
divided into dense cortical bone and thc fine network wbich forms cancellous bone.
The organic matrix in bone tissue and in two of the dental tissues, cement and
dentine, is predominantly composed of collagen arranged in parallel bands and
lamellae. Collagen is absent from enamel, the third hard tissue of teeth, and an
incompletely characterized protein is found instead.
Crystals of sparingly soluble mineral salts are deposited on and in the collagen
fibres. Based on dry weight, the inorganic component may be estimated to comprise
60-75% in bone and 95% in enamel with around 15% of the mineral salts as cal-
cium carbonate (CaC03) and 85% as hydroxyapatite crystals (CalO(P04)6(OHh).
Within the hydroxyapatite lattice a small proportion of calcium atoms are substi-
tuted by Mg and Sr as weH as by Li, Na and K, while some of the hydroxyl groups
are substituted by F and Cl.
15.2 Fixation
lL Lyon (Ed.)
Theory and Strategy in Hi8lOchemistry
@ Springer Verlag 1991
208 H. Lyon
~
Possible disseelion with removal 01 skin, museies etc.
+
Fixation (15.2)
t
Possible sawlng out 01 seetions and X·ray assessment
t
Seleclion 01 lissue blocks (15.3)
Fig. 15.1. Initial procedure for the preparation of sections of hard tissue for microscopy (references
to sections in this chapter)
15.4 Decalcification
Preparation of satisfactory paraffin, celloidin, or frozen sections from hard tissue re-
quires prior removal of the mineral salts. This is achieved by treating with reagents
which react with calcium. At low pH the following process takes place:
CaC03 + 2H+ --+ Ca2+ + H20 + C02
Strong acids (e.g. HCI) rapidly lead to coarse artifacts, while weak acids (e.g.
formic acid or chromium (IlI) sulphate) produce far less artifactual change.
The mildest decalcification can be achieved using chelating agents (cf. Sects.7.1
and 17.5.2) such as disodium ethylenediaminetetracetate (EDTA). Calcium ions are
so strongly bound to EDTA that they cannot participate in other chemical reactions.
As the Ca-EDTA complex is water soluble the process acts as a continual removal
of Ca2+ from the solution.
15 Tissue Processing: VI. Hard Tissues 209
Celloidin
Cryostat seetions
Seelions 5 • 7 11m Seelions 15 • 20 11m 5 - 20 11m
i i
s+~
Mounting
Embedding
t
Methacrylate
Seelions 10 - 20 11m on
hard tissue microtome
I
t t
Saw 100 - 150 11m
""''1'"" t
Grind 70 11m
,
Staining
l
Mounting
X-ray examination
Polarization microscopy
Fig. 15.3. Procedure for the preparation of sections from non-decalcified hard tissue.
Table 15.1. Deca1cifying agents, concentration, relative decalcification time, and artifact formation.
Relative Artifacts
Decalcifying Concentration decalcification CO 2 -
agent Type % time formation Precipitates
++ - substantial, + - slight.
15 Tissue Processing: VI. Hard Tissues 211
Decalcifying
Purpose agent Time
Several methods for determining the end point of the decalcification process are
available:
1. Mechanical methods, in which the content of calcium salts is estimated on the
basis of the mechanical properties of the specimen (hardness, flexibility), are
not recommended. These methods are very unsatisfactory partly because they
induce artifacts in the tissue and partly because smal1, central, undecalcified
areas are easily overlooked
2. Chemical methods involve testing for the presence of Ca2+ in freshly changed
decalcifying fluid. Examples include oxalate and the fluorescent product which
arises on the addition of Calcein (Fluorescein bismethylene iminodiacetate)
212 H. Lyon
3. Radiographie (X-ray) methods provide much the quiekest and safest means of
controlling the course of decalcification. Fully decalcified specimens give a
homogeneous shadow when a suitable voltage, current, and exposure time are
used
The frequency at which the course of decalcification should be assessed depends
on the degree of mineralization and the decalcifying agent. Using nitric acid or
hydrochloric acid on a biopsy, a test after 6-18 hours, depending on the expected
amount of mineral, is usually sufficient. With a weak acid, such as formie acid, a
single daily test is appropriate, while with EDTA the interval should be extended
to a week.
Surplus acid must be removed by quiekly washing in tap water before any further
fixation or dehydration procedures are undertaken. If EDTA has been used the
sampie should be placed in a formaldehyde solution with added NaCl for 12 hours
before transfer to 70% v/v ethanol. This prevents precipitation of EDTA in the
tissue by ethanol. Careful dehydration and infiltration with embedding medium are
critical in the preparation of blocks from hard tissue.
After decalcification with EDTA most staining methods present no problems. This is
also true for acid treatment providing this is stopped promptly when decalcification
is complete. If, however, acid treatment is continued beyond the optimum time,
extraction of nucleie acids occurs (cf. Table 15.3) and modification of methods
involving the use of metal complex dyes and cationie dyes becomes necessary.
For example, a standard Haematoxylin-Eosin stain yields strong red cytoplasmic
staining but no nuclear staining. Table 15.3 surveys staining methods used in the
investigation of hard tissue.
Accepting a risk of chipping the knife, firm tissue which feels only slightly gritty
can be cut when frozen without preceding decalcification. If frozen seetions are to
be made from decalcified tissue the sampie must first be washed thoroughly in tap
water to avoid damaging the mierotome and the knife.
15 Tissue Processing: VI. Rard Tissues 213
Cancellous bone can be cut, but needs special support. On the other hand cortical
bone can only be cut with special hard tissue knives or saws. Cortical bone can
also be ground to a suitable thickness.
The procedures involved are:
a. With a very stable microtome, the new hard tissue knives of specially hardened
material can be used direcdy to cut sections of undecalcified hard tissue. It is
an advantage to use a motorized ground sledge microtome with a cutting speed
of around 2 mm per second
b. Sawing produces irregular sections with a thickness of 100-150 p.m. These
are ground between two rotating pieces of plate glass with powdered pumice
and water to a thickness of 70 p.m which is weH suited for X-ray structural
examinations
c. Grinding can be extended until the minimum achievable thickness of around
20 p.m is attained. A lot of tissue is lost with this method
The staining methods listed in Table 15.3 are all applicable with the possible
additions of Chromoxane Cyanine R (see below) and the von Kossa method for
calcium (Sect.17.7.1). It should be noted that staining times for methyl methacrylate
214 H. Lyon
sections are considerably prolonged due to the slow diffusion in the methacrylate,
for example with Haematoxylin-Eosin a 60 + 30 min procedure is required.
In this procedure the tissue is first infiltrated with methyl metliacrylate monomer.
This is then polymerized and hardened. The surface of the tissue is exposed by
sawing or grinding so as to avoid unnecessary strain on the knife. Finally, sections
are prepared with a hard tissue knife or D-knife or by sawing and grinding as
described above.
The hydrophilic monomer, hydroxyethyl methacrylate, can also be used for un-
decalcified bone tissue. The advantage is that it is possible to use the "ordinary"
staining methods without prolonging staining time or increasing the temperature.
Fluorescence marking can be used for investigations into the calcification process.
For example, tetracycline, which binds to calcium ions, is taken up in vivo into
areas of active calcium deposition. Thereafter, either on frozen sections of fresh,
non-decalcified material or on ground sections of fresh tissue or material fixed in
70% v/v ethanol, the tetracycline can be visualized in the fluorescence microscope
as golden yellow deposits on a blue autofluorescent background (collagen). Ad-
ministration of two or more doses at defined time intervals enables an estimate of
the rate of the calcification process.
16 Tissue Processing: VII. Post Treatment
H.Lyon
16.1 Introduction
For an image to be seen with an unstained section in the light microscope, different
tissue components must have different refractive indices from each other and from
the mounting medium. Pieces of tissue, which are predominantly protein, have a re-
fractive index of 1.53-1.54 when fixed with formaldehyde. If they are subsequently
infiltrated with a colourless, transparent liquid with the same refractive index as
the tissue (e.g. methylsalicylate) they become invisible and no image is seen in the
microscope. Contrast can be achieved in three different ways:
1. The tissue can be stained. The section becomes visible because light of certain
wavelengths is absorbed by the dye or dyes bound to tissue components but is
relatively unaffected by the mounting medium. A mounting medium with the
same refractive index as protein is now an advantage as this gives an optically
clear image
2. The tissue can be left unstained, but can be made visible by mounting it in a
liquid with a refractive index which differs slightly from that of proteins. The
section may then be examined by phase or differential inteiference contrast
microscopy
3. A transparent piece of tissue can be made visible by direct microscopy if it
is mounted in a liquid with a refractive index substantially different from that
of protein. This method is only applicable for low power examinations and
it should be noted that boundaries between the tissue and mounting medium
appear as sharp lines which have no in vivo correlates
After staining, histochemical reaction, blocking reaction etc., the section must
be made suitable for microscopic examination. For most procedures the require-
ments are:
1. Removal of surplus reagent(s)
2. Mounting in a medium which does not interfere with the reaction product or
with the subsequent microscopic examination
H. Lyon (Ed.)
Theory and Sb'ategy in Histochemistry
© Springer Verlag 1991
216 H. Lyon
The next stage depends on the choiee of mounting medium (Sect.16.4).1f the latter
is miscible wirh water (hydrophilie) further treatment is unnecessary, however, if
a hydrophobie medium is to be used then the section must be dehydrated. Sequen-
tial exposure to an ascending series of ethanol concentrations is frequently used
(e.g. 70% v/v, 95% v/v, and 99% v/v). It should be noted that cationie dyes, es-
pecially Thiazin dyes can be extracted by ethanol. Anionie dyes are also extracted
in water-ethanol mixtures. As paraffin sections have already shrunk as much as
they are able to, sections can be moved directly from water to absolute ethanol.
This approach, however, leads to dilution of the ethanol with water from the sec-
tions after relatively few slides have been processed. Due to the problems with
extraction of dyes in ethanol (Sects.6.1 and 6.1.6), several other compounds have
been suggested for use as dehydrating agents. Examples include 2-propanol, 1-
butanol, 2-methyl-2-propanol (tert-butyl alcohol), and acetone. Finally, dehydration
can also be performed by drying the section in air, preferably at around 40°C in a
thermostatically-controlled oven containing silica gel. A vacuum oven ean also be
used for this purpose.
to fill all spaces between the tissue components. Furthennore, it must not dissolve
or corrode any of the tissue components which are to be examined nor should it
be a suitable substrate for fungal or bacterial growth.
Solid objects that cannot shrink such as ground sections of bone (Sect.15.8),
can be mounted dry (Le. with air as the mounting medium). Mounting media can
be fluid, mixtures of several fluids, or solutions of one or more solid compounds
in a liquid solvent. Volatile mounting media can be used as the edges of the cover
slip can be sealed to the slide (e.g. with gelatin). Compound mounting media often
use the volatility of the solvent as they become increasingly viscous and finally
solidify with prolonged storage.
As stated above, mounting media can be divided into hydrophobic and hy-
drophilie and both these groups can be subdivided into adhesive (e.g. EukittR ) and
non-adhesive (e.g. methylsalicylate). The advantage of the non-adhesive mount-
ing media is that the final refractive index is known as there is no change in the
composition of the mounting medium on evaporation.
The exact refractive index for an adhesive mounting medium cannot be given
as it is not possible to control the extent of evaporation. The limits for the re-
fractive index are nonnally given (Le values for the fresh mounting medium and
after complete solvent removal). It should be appreciated that complete solvent
evaporation from a mounted section cannot be expected. Against these problems,
adhesive mounting media offer the advantage of secure binding of the cover slip
to the slide.
Hydrophilic mounting media contain or are miscible with water thereby allowing
the sections to be directly transferred from water. Hydrophilic mounting media are
used when low refractive indices are needed or when dehydration is to be avoided.
Hydrophilic mounting media are valuable in investigations conceming lipids as
these are preserved regardless of whether they are fixed. Lysochromes retain their
colour weH in these media but some dyes, such as AI- and Fe-haematein, are not
quite stable.
These media are not miscible with water, and the tissue sections must therefore be
dehydrated first (Sect.16.3). After dehydration the section is usually irnmersed in the
solvent that forms the basis for the mounting medium and from here to the medium
itself. In practice it is usually possible to achieve good results by transferring the
"wet" sections from the final dehydrating step (e.g. absolute e_thanol or I-butanol
(Lyon et al., 1983)) direcdy to the hydrophobic mounting medium.
The adhesive, hydrophobie mounting media "set" as a result of solvent evapo-
ration. The solvent is usually xylene which evaporates sufficiently slowly to allow
the cover slip to be placed as desired. The congealing process can be accelerated
by placing the mounted slides on a hotplate or in a thermostatically controlled oven
at around 40°C. Most hydrophobie mounting media have high refractive indices
quite elose to that for formaldehyde fixed protein. They therefore usually give more
transparent preparations than hydrophilic mounting media.
H H
I I
-c-c-
~©
n
P. PrentrjJ
17.1 Occurrence
Some metals are normal constituents of living tissues (calcium, magnesium, iron,
selenium, zinc, copper, sodium, and potassium); some (copper and selenium) are
only needed in small amounts, and larger concentrations are nearly always due to
industrial or environmental contamination, while the occurrence of others (lead,
mercury, thorium, cadmium, gold, etc.) is always pathological and reflects contam-
ination.
Many invertebrates accumulate relatively large amounts ofmetals. Earthworms,
for instance, accumulate zinc and tunicates accumulate vanadium. It is not known
whether this represents contamination (industrial or otherwise) or a functional ac-
tivity.
From a histochemical point of view, inorganic constituents of tissues may be
present in three states:
1. Dijfusible ions. Precise localization is difficult or impossible. Examples include
sodium, potassium, chloride, and bicarbonate. See, however, Sect.28.8.10.
2. Immobile salts are generally directly detectable and localization is usually de-
monstrable without difficulty. Examples include ferric iron in haemosiderin and
calcium and magnesium phosphates
3. Masked, inorganic constituents. These are not directly reactive but are cova-
lently or complexly bound to organic molecules. Examples include iron in
haemoglobin (only detectable after demasking with alkaline 30% H202) and
certain mercury compounds (where Hg is covalently bound to thiol groups).
Before describing the "conventional" histochemical methods for detecting met-
als, two methods based on very different principles will be outlined
17.2 Micro-Incineration
This method dates back to Raspail (1833). The organic constituents are inciner-
ated by heating to 500-700 0 C, then the ashed section is examined by polarization
microscopy or other techniques, possibly in combination with a chemical or his-
H. Lyon (Ed.)
Theory and Sb'ategy in Histochemistry
© Springer Verlag 1991
224 P. Prent~
even dehydration and embedding in paraffin should be omitted. The best starting
point for an investigation of inorganic constituents is rapid freezing at -70°C or
below, followed by freeze-substitution, freeze-drying or freeze-sectioning. Freeze-
sectioning is very weIl suited for investigation of immobile metal deposits in tissue.
Buffered formalin may sometimes be used before freezing.
Metal salts are a substantial, and by far the most heterogeneous, component of
inorganic tissue constituents. Metal ions may be detected, direct1y or indirect1y, by
use of substitution methods, or by chelation with appropriate metallochromes.
Metal-A Metal-D
I + De- --+ +A e-
tissue coloured tissue
Metal-A C-A
II
tissue
+ C+ --+
coloured tissue
+ Metal+
A = anion; D = dye; C = cation
Method I is used in the Prussian BIue reaction for iron (Sect.17.7.4). Method II is
of course not in itself ademonstration of the metal, and the result is often difficult
to interpret. Nonetheless, some of these methods, such as the silver substitution
method of von Kossa which detects certain calcium deposits, are valuable.
The potential application of these methods is very extensive as they depend on the
selective binding properties of chelating dyes known as metallochromes. The prop-
erties of these reagents (chelating ability, colour, solubility, etc.) can be modified
by pH, solvent, or other complex-forming substances such as cyanide, tartrate etc.
(cf. Sect.7.1 for chelation). The properties of some histochemically useful metal-
lochromes are compared in Table 17.1. (See also Feigl, 1958).
226 P. Prent~
17.7.1 Calcium
Metal Substitution Methods. Many metal ions (silver, cobalt, iron, copper, etc.)
can be exchanged with the calcium present in calcium salt deposits. The best known
method is von Kossa' s si/ver substitution method (Lillie and Fullmer, 1976, p.539;
Pearse, 1985, p.982). First silver is substituted for calcium then the bound silver is
reduced to form black, metallic silver; e.g.
Errors. Uric acid and urates reduce silver ions and may be mi staken for calcium
deposits. They may, however, be removed by treatment with saturated, aqueous
lithium carbonate before substitution. The problems associated with oxalates are
discussed above.
Other reducing substances in tissue may sometimes (independent of light or
post-reduction) produce a spontaneous reaction (pearse, 1985, p.982). The presence
of melanin can mimic a positive von Kossa reaction as can carbon particles. Acid
proteoglycans may bind silver ions giving a variable amount of additional staining.
Table 17.1. Chelate colour after chelation of metal ions with some commonly used metallochromes. -.1
-
Metal Dithizone Rubeanic acid Alizarin Red S Na-rhodizonate Bathophenanthroline
~
228 P. Prent~
Limit of detection
in spot test (Ilg) Microscopic
(according to limit of detection
Metal Metallochrome Feigl, 1958) pg per 11m2
1 Experimental value (Prent0, 1979). This level of calcium is easily detectable, the real
limit of detection is probably 0.01-D.03 pg.
20nly applies to Cu +, as Dithizone does not react with Cu2+.
3 Experimentally (Prent0, 1979), this level of zinc is easily detectable and the real limit of
detection is probably around 0.005 pg.
Under nonnal circumstances none of these sources of error give rise to con-
fusion and most problems can be avoided by comparing the von Kossa treated
section with an untreated control and with a section where reduction has been
omitted. Sometimes intracellular calcium deposits may be difficult to identify with
certainty. Alternative methods for calcium demonstration should then be consid-
ered. Additional evidence can be obtained by showing that there is no staining when
the calcium reaction is preceded by treatment with weak acid or 2-5% EDTA.
Interestingly, pure crystalline calcium carbonate or calcium phosphate does
not react with the von Kossa method. The amorphous nature of mineral deposits
found in biological material (probably due to the elose association with organic
materials) seems to be necessary for a positive reaction. In very dense deposits
only the outer zone is positive due to poor penetration by the silvet:-salt While the
von Kossa method generally perfonns wen on vertebrate material, it is not suitable
for invertebrates (Prent~, 1979), probabIy because their calcium deposits have a
different composition from those associated with vertebrates.
Anion Substitution Methods. By treatment with certain acids the anions of calcium
deposits may be exchanged with the anions from the acid, leading to the fonnation
of weH defined crystals. This method is specific but the crystal fonnation leads to
very coarse localization.
Application of dilute sulphuric acid leads to the fonnation of gypsum crystals
(CaS04) in and above the section, while the original calcium deposit disappears.
For chelation methods it is essential that all details in the description of the
method are strictly adhered to and that the appropriate control reactions for inter-
fering metals are done. For instance, Alizarin Red S may forffi-red chelates with
several other metals in addition to calcium (see Table 17.1).
17.7.2 Magnesium
Magnesium can only be detected by chelation methods. However, the ability of the
magnesium ion to fonn chelates is relatively poor. Chelation has 10 be perfonned at
high pH (12-13), which favours magnesium chelation and precipitates most other
metals in the fonn of non-reactive metal hydroxides. (Above pH 13 even magne-
sium fonns insoluble, non-reactive magnesium hydroxide). At best the reaction is
slow and the high alkalinity disrupts the tissue. The best chelating agents for Mg
are Thiazol Yellow G (Pearse, 1985, p.997), also called "Titan Yellow", C.I. 19540,
and Magneson, 4-(p-nitrophenylazo)-resorcinol (pearse, 1985, p.997).
Titan Yellow imparts a flaming red colour to magnesium deposits. The colour
disappears after a few hours. Detection limits are shown in Table 17.2. Magneson
stains magnesium deposits bright blue and interference from other metals may be
prevented by adding potassium cyanide.
In vertebrates copper is only found as eu2+ and, with the exception of certain
pathologie conditions, the levels are below the limits of histochemical detection.
In some cases invertebrates contain relatively large amounts of copper, which is
present in both the cuprous and the cupric fonn.
Even if copper is present in detectable amounts a large part of it may be so finnly
bound to proteins that it cannot be visualized direct1y. Demasking may be perfonned
with hydrogen peroxide or better with HCI vapour, but this is by no means 100%
effective. Copper is best detected with rubeanic acid (dithiooxamide) or the zinc
or sodium salt of diethyldithiocarbamate (DDTC). The so-called oxidation-catalyst
reactions should generally be avoided. They are sensitive and selective, but are
17 Metals and Metal Salts 231
Uzman's Rubeanic Acid Method. In this method (Uzman, 1956) rubeanic acid
is a very sensitive reagent for both Cu2+ and Cu+ (see Table 17.2 for detection
limits). The chelate is green to greenish black and cannot be mi staken for the iron,
cobalt, and lead ehelates, which are yeHow to yeHowish brown, or with the nickel
chelate, which is blue.
The rubeanic acid method is the method of ehoice for vertebrate material.
H N-C-C-NH C2 Hs _
2 11 11 2 NC-S
S S C2 Hs - 11
S
Rubeanic acid Na-diethyldithiocarbamate
17.7.4 Iron
In tissue, iron does not normally oceur as inorganic deposits but is bound to proteins
which include eytochromes, haemoglobin, myoglobin, ferritin, and haemosiderin.
Both ferritin and haemosiderin bind ferrie iron in a manner that allows direet
detection provided the pH is sufficiently low. The properties of tissue iron are
summarized in Table 17.3.
Iron deposits are stainable with freshly prepared, non-oxidized Haematoxylin,
giving a blue to bluish blaek colour. The reaetion is (cf. Sect.7.2.3)
Fe3+ + Ht.H2 -+ Fe2+ + Ht.
Fe2+ + Ht. -+ Fe-Ht. chclate
The method is sensitive, but shows poor selectivity.
The most useful methods for the detection of iron are Perls' Prussian BIue
reaction and the Bathophenanthroline method (Hukill and Putt, 1962).
Perls' Prussian Blue Reaction. This reaetion (Lillie and Fullmer, 1976, p.501;
Pearse, 1985, p.974) is an anion substitution method in which the section is exposed
to a freshly prepared, acidified solution of potassium ferrocyanide. The acid releases
232 P. Prent~
I Not recommended.
2The method demands higher H 20 2 concentration (about 0.05%) than the peroxidase reactions
proper. Benzidine is carcinogenic and the method should not be used.
3 According to Lison-Dunn (Lillie and Fullmer, 1976, p. 447). Patent Blue V, C.I. No. 42051.
ferrie ions from the tissue anions (generally the proteins apoferritin and aposiderin).
The ferrie ions on liberation immediately form a eomplex with the ferrocyanide
anion, for instanee
The resultant highly insoluble eomplex is ealled Prussian BIue. The exaet eom-
position of the eomplex depends on the proportion of ferrous to ferrie ions. The
intense eolour of the Prussian Blue eomplex is assoeiated with the fact that most
of the ferrous eleetrons derived from the ferrocyanide are very loosely bound and
in eonsequenee enhanee light absorption. In the formed eoloured eomplex it is
therefore not possible to attribute the ferrous or ferrie state to any individual Fe
atom.
There are several modifieations of Perls' Prussian BIue reaetion. Without doubt,
the most valuable is Lison's (Pearse, 1972, p.975), whieh employs equal parts
of 0.25 mol/l Hel and 2% potassium ferrocyanide, giving a pH of 1.5 whieh
is optimum for the reaetion. At higher pH not all the ferrie ions in ferritin or
haemosiderin are released, and at Iower pH the rate of ferrie ion release is too
high in relation to the rate of Prussian Blue formation. If iron is present in large
amounts, or loosely bound, it may be profitable to earry out the reaetion at a
17 Metals and Metal Salts 233
higher pH (e.g. around 2) or, even better, the section may be preincubated in a
non-acidified ferrocyanide solution and the HCI added after 5-10 min.
Selectivity and Sensitivity. The formation of Prussian Blue is specific for fer-
ric ions and very sm all amounts may be detected (Table 17.2). In order to avoid
non-specific precipitates, it is necessary to use absolutely clean glassware, deminer-
alized or distilled water, analytical grade chemicals, and to avoid exposure to meta!
objects of any kind. Dust particles (or tobacco ash or smoke) are also significant
contamination risks for sections, chemicals, and solutions.
The ferrous iron in the ferrocyanide used for the Prussian Blue reaction is
quite rapidly oxidized to ferric iron by atmospheric oxygen. If the reaction time
is extended non-specific Prussian Blue formation may occur, even if the reaction
has been performed correcdy in other respects. The ferrocyanide solution must
therefore be jreshly prepared and the re action time should not exceed half an hour,
counting from the time of preparation. If a prolonged reaction time is required, the
section should be transferred to a fresh ferrocyanide solution every half hour.
Accumulations of asbestos fibres bind ferric ions in vivo and may therefore be
Prussian BIue positive to varying degrees. Similar reactions mayaiso occur with
calcium deposits or lipofuscin.
Deposits of exogenous iron may sometimes contain ferrous ions. If these are not
too firmly bound, they may be detected by use ofpotassiumjerricyanide. Although
the resulting blue complex is sometimes called Turnbull Blue, it is the same as Prus-
sian BIue. Ferrous iron is more readily demonstrated by the bathophenanthroline
method, which is also well suited for the general demonstration of iron.
Selectivity. The formation of a red chelate is specific for ferrous ions. Copper forms
a colourless, and cobalt a light yellow chelate. Addition of 0.25 mol/l thioglycolic
acid to the reagent ensures reduction of ferric ions to ferrous ions and allows
demonstration of the entire non-haem iron content of the section without having to
use acid demasking. In contrast to the Prussian Blue reaction, Bathophenanthroline
cannot give rise to false positive reactions. However, contamination is a potential
problem due to the sensitivity of the reaction for ferrous ions (see Table 17.2).
The reaction time for Bathophenanthroline should not exceed two hours, as
ferrous iron from haemoglobin and other haem proteins gradually chelates with
Bathophenanthroline. As the iron chelate is soluble in organic solvents, the stained
seetion is mounted in an aqueous mounting medium.
234 P. Prent~
17.7.5 Zine
Dithizone Method. The best method for the detection of zinc ions is chelation
with dithizone (diphenylthiocarbazone) (Mager et al., 1953).
N-N-fc5\
/ \,{+ .\::::::!I
s=c Zn
"~~-@
Dithizone (diphenylthiocarbazone) Zinc chelate
Sensitivity. See Table 17.2. The method demonstrates the natural occurrence of zinc
in areas such as the pancreatic beta cells or cells with a high content of carbonic
anhydrase.
17.7.6 Lead
Lead does not occur naturally in organisms but it may be accumulated as a result of
environmental pollution or through an occupational hazard. Organisms that- store
salts of calcium, magnesium, or zinc are particularly prone to accumulate lead.
In tissues lead is most often present in the form of salts. Lead in bone is not
removed during decalcification, provided the decalcifying fluid contains more than
10% sulphate. This ensures that lead is precipitated as lead sulphate during the
decalcification process but concomitant delocalization is probably significant.
Lead ions may be demonstrated by anion substitution with HzS (lead sulphide
methods), or by chelation with rhodizonate.
Lead Sulphide Methods. In these methods (Lillie and Fullmer, 1976, p.547),
treatment with HzS or an ammonium sulphide solution causes lead salts to form
brown insoluble lead sulphide. The methods are very poorly selective (see Demon-
stration of heavy metals, Sect.17.7.11). The selective demonstration of lead may
be improved by bleaching the section with HzOz (which gives rise to colourless,
insoluble lead sulphate), followed by re-exposure to hydrogen sulphide gas or am-
monium sulphide solution.
Rhodizonate Method. The rhodizonate method (Lillie and Fullmer, 1976, p.547;
Pearse, 1985, p.996) using sodium or potassium rhodizonate is able to form chelates
with many different metal ions. At pH 2.5-2.8 (e.g. in dilute acetic acid) red chelates
are formed with barium, cadmium, mercury (non-covalently bound), and lead. Pre-
treatment with dilute sulphuric acid dissolves barium, cadmium, and mercury salts,
while lead remains in the tissue as insoluble lead sulphate, and therefore renders the
rhodizonate method specific for lead. In contrast to the prolonged decalcification
procedures, brief treatment with dilute sulphuric acid or with a sulphate solution
does not appreciably affect the localization of lead deposits.
The chelating ability of rhodizonate is very high; even PbS reacts. At higher pH,
6-7, lead rhodizonate changes colour to more or less blue-violet. For sensitivity,
see Table 17.2.
236 P. Prent~
000_-
o
o
o
0
2 Na-
17.7.8 Cadmium
Cadmium is a very poisonous metal and its presence in the tissues is probably
always due to contamination. Unfortunately, the demonstration of cadmium is sub-
ject to interference from a variety of competing metal ions. Sumi (1980) and Sumi
et al. (1982) have described a method with increased selectivity towards zinc and
cadmium ions using various benzothiazolylazonaphthol (BTA) derivatives. For ex-
ample, after treatment of the section with BTA-ß-naphthol, cadmium, zinc, nickel,
and manganese deposits were stained blue, while most other interfering metal ions
stained red to purple. Post-differentiation with 1% oxine (8-hydroxyquinoline) in
75% v/v ethanol turned tissue cadmium from blue to red, whereas zinc remained
blue. The colour of other interfering metals remained red or their chelates were
decolourized by the oxine solution.
17 Metals and Metal Salts 237
17.7.9 Mercury
17.7.10 Silver
Silver appears in the tissues as brown to black granules. The brown granules are
blackened by treatment with hydrogen sulphide gas or ammonium sulphide solution.
The state of tissue-bound silver is uncertain; it may be partly protein-bound and
partly in the metallic state.
Si/ver salts are dissolved by solutions of potassium cyanide, sodium thiosul-
phate, or dilute sulphuric or nitric aeid. "Metallic" silver may thus be extracted
following pretreatment with an iodine/iodide solution (Weigert's or Lugol's io-
dine). One per cent potassium ferricyanide in 20% sodium thiosulphate simulta-
neously bleaches and extracts "metallic" silver. The treatment removes all silver
from the section within 30 minutes, leaving melanin granules and carbon particles
unchanged.
There is no acceptable histochemical method for the demonstration of silver.
Heavy metals form insoluble sulphides with hydrogen sulphide gas or ammonium
sulphide solution. These sulphides are mostly brown to brownish black. The brown
colour may be blackened by substitution of the heavy metal by silver as shown
below.
17.7.12 Anions
Polyphosphates. Phosphate polymers with more than 7 phosphate units with an-
hydride linkage are termed polyphosphate and exhibit ethanol-resistant metachro-
masia with Toluidine Blue and other metachromatic dyes. In contrast to sulphated
carbohydrates, polyphosphates bind lead ions at pH 3-4, and may therefore be
demonstrated by a lead nitrate-ammonium sulphide sequence (Ebel et al., 1958).
The occurrence ofpolyphosphate granules ("volutin" granules) is doubtful in higher
animals and plants, but they are relatively common in bacteria, fungi and protozoa
(Sect.2.3).
18 Pigments
H.Lyon
Pigments are defined and classified in Sect.2.1.7. Metals and metal salts are dis-
cussed in Chap.17.
This group of pigments is derived from the haem molecule which is characterized
by the heteroaromatic pyrrole ring:
I[)
N
H
(also cf. tryptophan and its compounds, e.g. serotonin, Sect.18.1.4). The major-
ity of haem pigments contain four pyrrole rings connected by methene groups
and are called tetrapyrrolic pigments. One pigment, Dubin-Johnson pigment, may
only comprise two rings and is therefore dipyrrolic. The basic structure in the
tetrapyrrolic pigments is the porphin molecule:
H. Lyon (Ed.)
TheoIy and Sttategy in Histochemistry
@ Springer ~lag 1991
240 H. Lyon
a pigment moiety, haem. Haem comprises a ferrous iron (FeH) complexed with
porphyrin (a substituted porphin moleeule):
Porphyrins. These are derivatives of porphin which do not contain iron. In addition,
the 8 H-atoms are substituted as shown above in haem. Porphyrins may be found
in tissues in both congenital and acquired porphyria.
spleen, liver, bone marrow, and lymph nodes. An identical pigment is encountered
in schistosomiasis.
Bile Pigments. These are formed as part of the normal degradation of haemoglobin
from red cells. The haem part is dissociated from the globin and an oxidative
opening of the ring system then leads to the formation of the iron containing product
verdohaemin. The iron is transferred to apoferritin and the remaining product is
biliverdin. This is produced mainly in the phagocytes of bone marrow and spleen
and is transported via the blood to the liver, where, under normal conditions, it is
reduced to bilirubin in hepatocytes. Bilirubin then forms a mono- or diglucuronide
with one or both of the propionic acid residues to Cl-hydroxyl in glucuronic acid.
The conjugated bilirubin is secreted into the bile capillaries and passed on to the
intestine where further reduction leads to urobilinogen (stercobilinogen). This is
subsequently oxidized to urobilin (stercobilin), the yellowish-brown pigment in
stool and urine.
biliverdin
bilirubin
242 H. Lyon
HOO~C
00H
OH
HO
OH
glucuronic acid
urobilinogen
urobilin
18.1.2 Lipofuscin
Lipofuscins are yellowish-brown pigments which are derived from the degradation
of fatty acids. Their somewhat varying histochemical properties are attrlbuted to
varying degrees of polymerization and oxidation. Thus, ceroid, originally described
in Kupffer cells in experimental cirrhosis, is a partieularly eoarse granular fonn of
an early lipofuscin. In addition, lipofuscins are particularly prominent in:
• liver cells
• cardiac musele cells, especially around nuclei. Large arnounts are seen in brown
atrophy
• zona retieulosa of the adrenal cortex
• testis, especially Leydig cells
18 Pigments 243
18.1.3 Melanin
Melanins are complex polymer substances derived from tyrOsine. They are found
bound to protein in the form of granules. Melanins comprise the black to brown
pigment encountered in epidermis, hair follicles, hair, naevi, and cutaneous malig-
nant melanomas. Melanins further comprise the ocular pigments in the iris, ciliary
body, choroid and here and there in the meninges. Malignant melanomas from the
eye seem to develop from the choroid. The pigments in the substantia nigra, locus
coeruleus and nucleus dorsalis nervi vagi in the brain differ histochemically from
the above described melanins and are termed neuromelanins.
The biosynthesis of melanin in the skin and probably in the eye takes place
in melanocytes, where, in the presence of oxygen, the enzyme tyrosinase slowly
catalyzes the oxidation of tyrosine to dihydroxyphenylalanine (DOPA) and more
rapidly the further transformation of DOPA to DOPA quinone.
HO
~ COOH NH 2 _
(0)
t.
H0XQJt
O
HO
NHCOOH
2
_ (0)
t.
O~COOH
~ NH 2
o
t. =tyrosinase
The subsequent steps occur spontaneously with oxidation and polymerization
of several moleeules and finally conjugation to protein.
derived from these eells. Adrenaline and noradrenaline are most abundant in the
adrenal medullary eells and in phaeoehromoeytomas, the eorresponding tumour.
18.1.5 Carotenes
These are exogenic pigments eomposed of hydroearbons with long ehains and
eonjugated double bonds.
18.1.6 Carbon
Acid Haematins. These pigments (Sects.18.1.1 and 31.6.4) are birefringent and
are soluble in alcoholic picrie acid.
Bile Pigments (Seets.18.1.1 and 31.6.5) give several rather unreliable reactions
(see Sects.18.4.6, 18.4.7, 18.4.8, and 18.4.11).
18.2.2 Lipofuscins
These pigments (Sects.18.1.2 and 31.6.6) are autoftuorescent (Sect.30.1) and re-
ducing (positive ferric ferricyanide reaction, Sect.8.3.1). In addition, they show
18 Pigments 245
acid fast basophilia (Sect.18.4.1), resistance to bleaching with H202, and varying
degrees of sudanophilia (Sect.19.6.1). Although they are derived from fatty acids,
lipofuscins are resistant to lipid extraction (Sect.19.3) and are retained in paraffin
sections.
18.2.3 Melanins
These reducing pigments (Sects.18.1.3 and 31.6.7) are somewhat more easily
bleached by HzOz than lipofuscins (Sect.18.4.2), but this is an unreliable reac-
tion. If present in larger amounts the most selective method appears to be the
ferrous ion uptake method (Sect.18.4.9). The DOPA reaction (Sect.18.4.13) may
sometimes be of value.
These compounds (Sects.l8.1.4 and 31.6.8, and 31.6.9) are most selectively demon-
strated by induced fluorescence techniques (Sect.30.2) but may also, if present in
sufficient amounts, be demonstrated by their reducing properties (Sect.8.3.1) and
their ability to azo-couple with diazonium salts (Sect.18.4.10).
From Table 18.1 it is apparent that many of the reactions are common to several
pigments. It may therefore be difficult to select the first reaction. Table 18.2 pro-
vides an algorithm to make this task somewhat easier. The identity of a pigment
determined using this key may be confirmed by selecting an alternative reaction
according to Table 18.1.
Several important pigment reactions are discussed below. Some other reactions have
been considered in more appropriate sections (e.g. ferric-ferricyanide (Schmorl's,
Sect.8.3.1) and argentaffin reaction, Sect.8.3.2, PAS-reaction, Sect.9.2.1, and su-
danophilia, Sect.19 .6.1).
Table 18.1. Reactions for pigments and similar substances.
~
Dubin- cate-
haemo- por- haemo- acid' bile Johnson Iipo- sero- chol- caro- ascorbic
globin phyrin siderin haematin pigment pigment fuscin melanin tonin amines tenes acid carbon
Fluorescence l 0 + 0 0 0 + + 0 + + + 0 0
Specific light absorption + 0 0 0 0 0 0 0 0 0 0 0 0
Birefringence 0 0 0 + 0 0 0 0 0 0 0 0 0
Basophilia 0 0 0 0 0 0 + + + + 0 0 0
Acid fast basophilia 0 0 0 0 0 (+) + 0 0 0 0 0 0
Acidophilia + 0 0 0 + 0 0 0 0 0 0 0 0
Sudanophilia 0 0 (+) 0 0 (+) + 0 0 0 0 0 0
9-phenyl-2,3,7-trihy- + 0 0 0 0 0 0 0 0 0 0 0 0
droxy-6-fluorone
Bathophenanthroline 0 0 + 0 0 0 0 0 0 0 0 0 0
Bleached with H 2 0 2 0 0 0 0 0 0 + + + + 0 0 0
H 2 0 2 -Perls' reaction + 0 + 0 0 0 0 0 0 0 0 0 0
Perls' reaction 0 0 + 0 0 0 0 0 0 0 0 0 0
Ferric ferricyanide 0 0 0 0 + 0 + + + + 0 + 0
Argentaffin 0 0 0 (+) + + + + + + 0 + 0
Chromaffin 0 0 0 0 0 0 0 0 + + 0 0 0
Gmelin 0 + 0 0 + 0 0 0 0 0 0 0 0
Aldehyde-oxidation 0 0 0 0 + 0 0 0 0 0 0 0 0
PAS 0 0 + 0 + (+) + 0 0 0 0 0 0
Ferrous ion uptake 0 0 0 0 0 0 0 + 0 0 0 0 0
Azocoupling + 0 0 0 + 0 0 (+) + (+) 0 0 0
Soluble in alcoholic 0 0 0 + 0 0 0 0 0 0 0 0 0
picric acid
DOPA reaction 2 0 0 0 0 0 0 0 + 0 0 0 0 0
This is a modification of the Ziehl-Neelsen method for acid fast bacteria. Even the
faintest red staining is considered positive for lipofuscin or possibly Dubin-Johnson
pigment. The colour of the pigment itself can sometimes make the reaction difficult
to assess. This is particuIarly true of highIy polymerized lipofuscin. A doubtful
reaction can be verified using a Iysochrome reaction (best with Sudan Black B),
as lipofuscin is not extracted from the tissue during the preparation of paraffin
sections. According to Lillie and Fullmer (1976), p.516, it is an advantage to use
Night Blue (C.!. 44085) instead of Basic Fuchsin.
248 H. Lyon
See Seet.17.7.4.
The iron in haemoglobin ean be demonstrated after demasking with alkaline H202,
this dissociates it from the haem moleeule and eauses oxidation to the ferrie eon-
dition. This method, however, frequently fails to produee the required result. The
H202-benzidine methods of Lepehne and Pickworth are alternatives; N.B. benzi-
dine is earcinogenic (Pearse, 1985, p.888). Lillie and Fullmer (1976), p.447 and
p.487 suggest the use of either the leueo-Patent Blue-H202 method or reaetion with
9-phenyl-2,3,7 -trihydroxy-6-f1uorone.
In this reaetion potassium diehromate is used as the fixative. This oxidizes the
monoamines to brown pigments of unknown ehemieal eomposition. In Hillarp and
Hökfelt's iodate method (1953) potassium iodate is used as an oxidizing agent.
This makes it possible to differentiate between adrenaline and noradrenaline as
adrenaline takes longer to reaet. The method ean be used in the diagnosis of
phaeoehromoeytomas, but it is not nearly as sensitive as the f1uoreseenee methods
(Seet.30.2).
18 Pigments 249
A few drops of the reagent (equal parts of concentrated nitric acid and ethanol)
are placed on the section which is observed under the microscope. Bile pigments
change colour according to the sequence yellow, green, blue, violet-red. The method
is capricious and the colour unstable.
The majority of aldehydes can oxidize bilirubin to biliverdin. This also applies
to formaldehyde and the characteristic green colour is seen with formaldehyde
fixed material on unstained sections or sections stained with reagents such as van
Gieson's Picrofuchsin.
The yellowish-brown colour which bile pigments acquire in Al-haematein Eosin
stained sections is due to mixing of the green colour of the pigment with the red
colour of Eosin.
Ferrous ions are bound complexly to melanin and can be demonstrated with potas-
sium ferricyanide, giving rise to Prussian blue ("Tumbull BIue", Sect.17.7.4). Due
to its intrinsic colour, melanin appears dark green to green-black, while the back-
ground can be stained bluish, especially if washing in water after ferrous sulphate
treatment has been inadequate. For instance goblet cells are frequently stained rather
strongly blue. The method is, however, one of the most selective for melanin (green
to green-black), but the sensitivity is only moderate and is surpassed by far by the
ferric ferricyanide method (Sect.8.3.1).
18.4.10 Azo-Coupling
This reaction is best performed on frozen seetions with diazonium salts freshly
prepared from either ethyl anthranilate, (Raja, 1965; Desmet et a/., 1968; Lillie
and Pizzolato, 1970), Safranin 0 or Methylene Violet (Lillie and Pizzolato, 1969).
When reacting with the diazonium salt bilirubin is hydrolytically cleaved giving rise
to unsubstituted a-carbon atoms in two of the pyrrole rings to which the diazonium
salt is coupled fonning two isomer azobilirubins:
HOOC COOH
1 1
CH 2 CH 2 CH 2 CH 2
11 1 1 11
H 3C CH H 3 C CH 2 CH 2 CH 3 CH 3 CH
HOJ:5=CHJ:5- CH 2J:5-CH=O-OH
H H
~ +H 20
COOH
1
CH 2 CH 2
11 1
H3 C eH H 3 C CH 2
H O C C H - v . - C H 20H
N N
H
~+HP
COOH
1
CH 2 CH 2
H3C
11
CH
1
CH 3 CH 2 t
HOCCH)j
N N
+
H
18.4.13 DOPA-Reaction
The DOPA-reaetion (Lillie, 1956) is aetually areaction to deteet the enzyme ty-
rosinase which, as shown in Seet.18.1.3, eatalyzes the first steps in the synthesis
of melanin. Cells eontaining tyrosinase (melanoeytes) are able to oxidize DOPA
(dihydroxyphenylalanine) to DOPA quinone after which the synthesis eontinues
spontaneously to melanin. A positive DOPA-reaetion therefore, shows formation
of melanin, and it is the development of the brown-blaek pigment in eells which
is registered. The reaetion ean be performed on frozen sections of fresh tissue but
this does allow some diffusion. Formaldehyde fixation of small pieces of tissue for
1-4 h gives the best results. Usable results may still be obtained even after fixation
in cold formaldehyde solution for up to a week.
Interpretation of the DOPA-reaetion is rather diffieult as:
1. Preformed melanin eannot be distinguished from that synthesized during the
reaetion
2. DOPA ean be autooxidized in both alkaline and acid solution. The reaetion
should therefore be performed at pH 6.8-7.4
3. Other oxidases ean oxidize DOPA and thereby initiate melanin synthesis. Leuko-
eytes and red eells also give false positive reaetions due respectively to their
peroxidase eontent and the peroxidative aetivity of haemoglobin. Cytoehromes
may also have this effeet, so that there is a theoretieal possibility for reaetion
in non-melanin producing tumours
19 Lipids
P. Prentr/J
19.2 Pretreatment
H. Lyon (Ed.)
Theory and Strategy in Histochemistry
© Springer Verlag 1991
254 P. Prent~
Table 19.1. Histochemical methods for demonstration of lipids based on chemical and physical
characteristics with notes on pretreatment.
Pretreat-
Method Reactive group Lipid dass Staining result ment'
, Pretreatment:
A: cryostat section of unfixed tissue, no postfixation.
B: eryostat seetion of unfixed tissue, post fixation I h in formol-calcium in refrigerator.
C: cryostat section of unfixed tissue, postfixation for at least I week in formol-calcium in
refrigerator.
confluence of the droplets. If fixation is essential then periodic acid (0.5%, 5°C,
5 min) can sometimes be effective.
Table 19.1 outlines recommended pretreatment for the different histochemical
reactions for lipids (see Sect.11.3.1 regarding cryostat sections).
Triglyceride C (C) C C C
Cholesterol C (H) C C C
Cholesterol ester C (H) C C C
Waxes o or H 0 o or H C or H C
Phosphoglycerides 0 o or C o or H C C
Lecithin 0 C H C C
Cephalins, phosphoinositides 0 0 H C C
Plasmalogens 0 o or C o or C o or C C
Phosphosphingosides 0 o or (C) H C C
Sphingomyelin 0 C H C C
Sphingocephalins 0 0 H C C
Glycosphingosides Oor H 0 0 o or H C
Cerebrosides H 0 0 H C
Gangliosides I 0 0 0 0 C
Fattyacids
Free fatty acids C C orH C C C
Ca- and Mg-soaps 0 0 0 o or (C) C
Isoprenoids C C C C C
Failure to Extract Soluble Lipids. Soluble lipids may fall to be extracted due
to binding to insoluble cell components. This may involve other lipids (particu-
larly membrane lipids), proteins (notably phospholipids), or lipid-meta! complexes
(e.g. calcium and magnesium salts of cephalins, and fatty acids). The protein-
bound lipids can be made soluble by treatment with protease (Tuqon and Adams,
1961; Adams and Bayliss, 1962) or by denaturing protein (absolute ethanol, picric
acid, trichloroacetic acid, HCl etc.); disruption of metal-Iipid binding can often
be achieved with 0.1 mol/l HCl or 0.5% Na2EDTA. EDTA treatment leads to
Na-saponification and loss of free fatty acids.
Differential lipid extraction is not a routine procedure. Some general guidelines
may, however, be given. As the solubility of lipids is increased with increasing
temperature, solvents should not be used above 0-5°C, unless total lipid extraction
is desired. Qnly solvents that are acceptable for use in biochemicallipid investiga-
tions should be used. Where there is doubt, the extracted lipids should be identified
by chemical analysis or chromatography.
Extraction of homophasic lipids can usually be carried out without difficulty on
sections fixed for short periods in aldehydes providing the material is fresh. For all
other lipids fixation must not be performed before extraction. For differential lipid
extraction, only fresh unfixed frozen sections should be used for starting material.
Lipid solvents, unless otherwise stated, should be completely anhydrous. Ace-
tone, methanol, and ethanol may be dried over anhydrous CaCh. Frozen sections
must be completely dry before they are placed in the solvents.
The procedure set out in Table 19.3 forms the basis for a rational differen-
tial lipid extraction. To counteract extraction and dislocation of phospholipids the
frozen sections are first treated with a 1% CdCh or CaCh solution pH 7-7.5; see
Table 19.3.
If a single step, total lipid extraction is desired then the water in the chlo-
roform/methanol/water solvent should be replaced with 1 mol/l HCl. This makes
calcium salts more readily extractable. Total lipid extraction may require preceding
treatment with protease. The presence of water in chloroform/methanol aids the
extraction of the hydrophilic lipids.
19 Lipids 257
Procedure Extraction of
take place very quickly and in a way corresponds to a natural demasking of storage
lipid (Dixon, 1970; 1973).
A negative lipid reaction is only of value if the correct performance of the procedure
has been confirmed using a positive control section. The following control material
can be recommended: rat or mouse brain (all types of phospholipids, cholesterol);
fatty liver (triglycerides, free fatty acids); rat adrenal gland (cholesterol, cholesterol
esters).
All three tissues may be frozen into one common block and cut simultaneously.
The cryostat sections may be stored in a deep-freeze, preferably at -70°C or below,
for several months.
Table 19.4 is not exhaustive, but is sufficient for most histochemical investigations.
As several types of lipids often occur together, it must be appreciated that a positive
reaction for one type does not exclude the occurrence of others.
19 Lipids 259
1. Lipid
Lipid (ineluding hydroearbons and higher aleohols) - - - - - - - - - - - - - - - - ,
b Osmium tetrexide
c a and b after total lipid extraetion with ehlorolorm:methanol:HCI
2.
2. Hydrophobie or hydrophilie lipid
Hydrophilie lipid
Hydrophobie lipid
(ineluding hydroearbons and high er aleohols) - - - - - - - - - ,
3. 4.
3. Hyd~ophobie lipid (storage lipid)
Lipofusein or phospholipids - - - - - - - - - - - - - - - - - - - - - . ,
Triglyeerides --------------~..,
a Calcium lipase + o
bPAN + o
c Digitonin-PAN +
d Nife Blue pH 1 + o
e Copper(II)aeetate-rubeanie acid + (t) 0
f b-e following HCI, 1 mol/I, 1 h - acetone, 4 'C, 1 h o o o o +
260 P. Prent!<l
Phospholipids
a Nile Blue, pH 2 +
b PAS +
c a and b following HCI . acetone extraction + (+)
d a and b following chloroform:methanol:HCI extraction 0 0
5.
5. Phospholipids
Sphingomyelin
Cephalins
Lecithin
Acetal phosphatides (plasmalogens)
Phosphoglycerides
Lysochromes differ from dyes proper in that they lack ionized groups. A lysochrome
is practically insoluble in water, but readily soluble in alcohols, acetone, and liquid
lipids. Lysochromes are soluble in alcohol/water mixtures but the saturation point
is relatively low. The distribution of a lysochrome proceeds towards an equilibrium
between the alcohoIlwater phase in the stain bath and the lipid phase in a seetion
containing triglyceride droplets according to its relative solubility in the two phases.
The solubility of the lysochrome in the lipid phase is always high compared to its
solubility in the solvent (high partition coefficient), so the lysochrome is concen-
trated in the lipid provided the lipid phase and the solvent phase are immiscible.
Following staining (10-20 min), the section is rinsed, "differentiated", by washing
quickly with the lysochrome solvent. Critical factors in lysochrome staining are:
1. The concentration of the dye in the lysochrome solution
2. The lipid extractive or dislocative effect of the solvent
3. The staining temperature
4. The post-rinsing 'or differentiation step
These points are discussed individually below.
1). For ethanol-water mixtures, the ethanol content should never be below 70% v/v
to ensure a sufficient concentration of the lysochrome dye (see Oil Red 0). Below
60% v/v ethanol the solubility of the lysochrome is extremely low.
2). The ethanol content should not exceed 80% v/v as this leads to pronounced
dislocation, or even extraction, of lipid, even with short staining times.
3). The staining temperature should always be kept as low as possible in order
to minimize lipid extraction and dislocation. Except for tissue from mammals and
birds, staining may generally be performed at 5-15°C, and even for mammals it
is seldom necessary to use above 20-25°C, unless cholesterol or cholesterol esters
are to be stained.
4). In order to permit differentiation without excessive loss of dye from the lipid
droplets, neither lysochrome solution nor rinsing solvent should exceed an ethanol
concentration of 80% v/v.
The organic solvents most often used for lysochrome solutions are ethanol
(optimally 75% v/v), 2-propanol (60% v/v), and triethylphosphate (60% v/v). Of
these triethylphosphate seems to be the most gentIe with respect to lipid extraction
and dislocation.
Visible accumulation of lysochrome implies that a substantial volume of lipid
is present. If the lipid is present in a disperse form as opposed to droplets, the
lysochrome cannot achieve a high enough concentration to give a visible result.
Furthermore, if droplets are smalI, much of the lysochrome is washed out by the
262 P. Prent\'!
rinsing solvent. The possibility for a ja/se negative reaction is therefore always
present when using lysochrome reactions.
If the tissue section is pretreated with an acid protein coagulating agent such as
picric acid or HCI, with urea, or with a protease the disperse lipid will be direcdy
exposed to the aqueous phase and will sometimes coalesce into larger or smaller
droplets. The thus "demasked" lipid can now be demonstrated with lysochrome
methods. This approach does, however, give rise to artifacts of localization. Prior
to lysochrome staining, the section may be pretreated for 5 min with cold 0.5%
periodic acid. This stabilizes the section and improves the final morphology. At
the same time it reduces the eoaleseence of the lipid droplets.
The most frequently used lysochromes are Oil Red 0 and Sudan Black B.
Sudan Black B. This lysochrome stains all lipids and hydrocarbons if they are
liquid. The staining of hydrophilie lipids is probably related to Sudan black being
considerably more polar than the other lysochromes and the resultant possibility for
binding in parallel to the phospholipid moleeules in cell membranes. This organized
disposition can be utilized by examining the section in polarized light where the
Sudan Black B stained phospholipids show varying shades of red.
Sudan Blaek B contains two seeondary amino groups which are responsible for
the heterophasie properties of the compound. Unfortunately, the amino groups also
make it possible for the molecule to be ionized thereby offering the properties of
a eationic dye and possibilities for binding to non-lipids. Usually, this is of little
importance but, if required, the lysochrome ean, in theory, be made specific for
lipids by esterifying the amino groups in Sudan Blaek B with acetic acid anhydride
(Casselman, 1954).
Sudan Blaek Bis sold as a mixture eontaining two main eomponents (the ortho-
and para-isomers of the same eompound) and 8-11 coloured impurities (Marshall,
19 Lipids 263
1977; PfüHer et al" 1977). Sudan Black Bis unstable, and older solutions (includ-
ing stock solutions) contain degradation products which stain nucleic acids and
proteins, but not lipids (Frederiks, 1977). It is therefore essential for reproducible
and interpretable results, that the solution used, including any stock solution, is
freshly prepared (not more than one week old). Even small differences in the con-
centration of organic solvent in the dye solution are of considerable importance for
the concentration of the lysochrome. Ethanol should be used as 70% v/v and 2-
propanol and triethylphosphate as 60% v/v solutions, and an araeometer should be
used during the preparation. More lysochrome can be dissolved in both 2-propanol
and triethylphosphate than ethanol, and this allows the staining time to be kept
down to 5-10 min and minimizes the extraction of lipid.
In addition, triethylphosphate gives rise to considerably less conJluence of lipid
droplets than the other solvents and gives more stable solutions of lysochromes.
Regardless of the solvent, staining should normally take place at room temperature
as the extraction of lipid is considerably increased with increasing temperature.
According to Dunnigan (1968a) Nile Blue contains two dyes, a blue oxazine and
a red oxazone, where the latter is an oxidation product of the oxazine. Nile Blue
has a positive charge, but is otherwise a lysochrome that binds to phospholipids
and free fatty acids including calcium soaps. The red oxazone is uncharged and
264 P. Prentlil
The plasmal reaction (Hayes, 1949; Cain, 1949a, 1949b) demonstrates plasmalo-
gens ("acetal" phosphatides).
19 Lipids 265
Table 19.5. Reactivity with copper(II) acetate-rubeanic acid following extraction with
hydrochloric acid and/or acetone.
The reactive group is an a,ß-unsaturated ether bond between the a fatty acid
chain and glycerol. This bond is most unstable and is hydrolyzed with mercuric
chloride forming a fat-aldehyde-mercury addition product.
The fatty aldehyde can be demonstrated with Schiff' s reagent whieh binds to
the aldehyde group (Sect.9.7.1) or alternatively with reagents such as diphenylcar-
bazone whieh react with mercury. This latter approach, however, also demonstrates
Hg bound non-specifically to alkene groups. Schiff's reagent is therefore preferred
for the demonstration of fatty aldehydes.
The alkene groups in unsaturated lipids are relatively easily oxidized by at-
mospheric oxygen giving rise to Schiff-reactive aldehyde groups, "pseudoplasmal"
reaction (Cain, 1949a). In fixed tissue and in older frozen sections a strong pseu-
doplasmal reaction can "drown" the plasmal reaction. For this reason unfixed,
completely fresh, frozen sections should always be used.
With vertebrate material a control section whieh only has been treated with
Schiff's reagent should be colourless apart from any lipofuscin present (Sect.18.2.2).
In the case of invertebrate material a positive pseudoplasmal reaction may some-
times be natural.
Osmium tetroxide (Adams, 1965, p.34; Adams et al., 1967) selectively demon-
strates alkene groups in unsaturated fatty acids, both free and in esters. OS04 should
always be used in aqueous solution together with CaCl2 or CdClz. As OS04 vapour
is extremely toxie, any handling of the reagent should take place in a fume-hood.
The process of addition of OS04 to the alkene group is described in Sect.9.1.1. The
unsaturated storage lipids are oxidized to relatively insoluble products. The OS02
formed, "osmium black", remains in the lipid phase. Black, spherical accumulations
are therefore a specific indicator for the presence of liquid storage lipids (triglyc-
erides, cholesterol esters). Triglyceride accumulations containing only saturated,
fatty acids do not react with OS04. This sometimes makes it necessary to perform
a control with a lysochrome staining at elevated temperature. In heterophasie lipids
266 P. Prentf6
the OS02 formed makes cellular membranes electron den se, giving the classieal
trilaminar appearance.
In unfixed frozen seetions only alkene groups can reduce OS04 to OS02. The
formation of "osmium black" is therefore a speeifie indieator for the presenee of
unsaturated lipid.
Marchi Methods. In the Marehi methods (Adams, 1958; 1960; 1965, p.36) it
is possible by using OS04 together with another oxidant to distinguish between
hydrophobie and hydrophilie lipids. Usually, potassium chlorate (KCI03)is used
as the other oxidant. OS04 then aets as previously deseribed, however, when the
lipid present is hydrophilie, KCl03 penetrates and hinders the eomplete reduetion
to "osmium blaek". Instead, uneoloured addition eompounds eontaining hexavalent
osmium are formed. The result is that only hydrophobie lipids, Le. fat droplets
and possibly disperse triglycerides, are stained. (See however Sect.19.7, myelin
methods). OS04 and Os04/KCI03 are exeellent methods for the demonstration of
lipid as long as one remembers that any saturated short ehain fatty acids present do
not reaet, and that disperse hydrophobie lipid sometimes remains unstained with
the Marehi method.
Seetions stained with Marchi methods should be evaluated and any photomi-
crographs taken immediately after mounting as hexavalent osmium is gradually
reduced with the formation of more "osmium black". This readily occurs with
alcohol containing mountants.
OTAN Method. The OTAN method (Adams, 1959; 1965, pAO) is a further de-
velopment of the Os04/KCI03 method where the double oxidation is followed by
treatment with a-naphthylamine (osmium tetroxide-alpha-naphthylamine); where
hydrophilie lipids are present, a-naphthylamine forms a complex with hexavalent
osmium giving rise to red to brownish red a-naphthylamine-Os04-alkene produets
of addition. The staining results are, however, very variable and furthermore, a-
naphthylamine is very carcinogenie. A comparison of seetions stained with OS04
alone, Os04/KCI03, and Sudan Black gives far more information than OTA;~.
This method (Adams and Davison, 1959; Adams et al., 1963) is specific for phos-
phoglycerides. As noted above the carboxyl ester bonds in phosphoglycerides are
labile to alkali. In the presence of hydroxylamine a hydroxamic acid, or more
correctly, a hydroxamate is formed during hydrolysis, which can form coloured
complexes with different metals.
19 Lipids 267
o o
11 11
R- 0- C - hydrocarbon .. R-OH + HON-C-hydrocarbon
I
H
hydroxamic acid
If an ammmoniacal silver solution is used, the silver ions are reduced to free
silver, and the phosphoglycerides are stained black. It is usual to follow this by
toning with gold chloride.
As the reagents are insoluble in the lipid phase, triglycerides and cholesterol
esters do not react. Sphingolipids do not contain carboxyl ester bonds, but the more
alkali-resistant amide bond. The method is therefore specific for phosphoglycerides.
Treatment with NaOH makes it necessary to fix the seetion. This must not be
too prolonged as lecithin forms lysolecithin during formol-calcium treatment.
CH 2 00C-R(1)
I
CHOH + HOOC-R(2)
In this method (Baker, 1946; Elftman, 1954; Sinapius, 1961) the dichromate anion,
Cr2072-, oxidizes alkene groups in hydrophilie lipids relatively quiekly, while hy-
drophobie lipids, due to the insolubility of the dichromate anion in the lipid phase,
are oxidized slowly. During oxidation, cross-links are formed between the fatty
acid chains and the hydrophilie lipids are rendered insoluble in organie solvents.
Chromated tissue can therefore be embedded in paraffin and cut essentially without
loss of phospholipids. During oxidation divalent chromium is formed, possibly as
CrG, and this chelates with aminophosphate in the choline-containing phospho-
lipids, lecithin and sphingomyelin.
On treatment of the chromated seetion with an acid Haematein solution, Hae-
matein is bound to the divalent chromium whieh is concentrated in areas where
diehromate has reacted with lecithin anel/or sphingomyelin. Deposits of these lipids
are thus stained black to dark blue. Contrary to other double-bond reactions chrom-
268 P. Prent{3
In the UV -Schiff reaction (Belt and Hayes, 1956) ultraviolet irradiation for 3-
4 h transforms alkene groups in unsaturated lipids to aldehyde groups and these
can subsequently be demonstrated with Schiff's reagent. This oxidation method is
considerably more specific, easier to perform, and more gentle than performic acid
or peracetic acid oxidation. A high-pressure mercury lamp is a suitable source of
light. When used on unfixed seetions the method is specific for unsaturated lipids.
Larger fat droplets will usually only be stained on their circumference. After similar
irradiation, polyunsaturated fatty acids are fragmented and this sometimes results in
diffusion of the aldehyde products of hydrophilie lipids. Microscopic examination
should therefore take place immediately after the reaction.
A control seetion (no irradiation) is essential to rule out any possible "pseudo-
plasmal" reaction (see Sect.l9.6.4).
19 Lipids 269
There are no specific direct methods for the demonstration of the glycosphingo-
side group as such. However, the PAS (periodic acid-Schiff's reagent) method
(Adams and Bayliss, 1963) gives a positive reaction with the 1,2-glycol groups in
the hexoses in cerebrosides and in protein-bound gangliosides. The majority of the
gangliosides are not protein-bound and are extracted by water. As both polysaccha-
rides and glycolipids can give a positive PAS reaction, the presence of glycolipid
must be confirmed by lipid extraction and various blocking reactions. It is usually
sufficient to perform a total lipid extraction and a "pseudoplasmal" reaction (see
Sect.19.6.4) as a control for the demonstration of cerebroside. It is important not to
oxidize for more than 5 min in periodic acid as this makes alkene-groups reactive.
Recently Buk and Bayliss High (1986) have described a borohydride-periodate-
Schiff sequence as a reliable method for the specific demonstration of sialic acid
containing lipids (protein-bound gangliosides).
There is no satisfactory method for the demonstration of non-protein bound
gangliosides.
Triglycerides lack unique reactive chemical groups and can therefore only be iden-
tified unequivocally using enzymes. Adams et al. (1966) have developed a calcium
lipase method. The triglycerides are c1eaved into glycerol and fatty acids by pan-
creatic lipase. In the presence of calcium ions, insoluble calcium soaps are formed
which are transformed into lead soaps on the addition of lead nitrate. Bound lead is
visualized by reaction with ammonium sulphide which gives rise to the formation
of brown to black sulphide.
Calcium salts and free fatty acids and calcium soaps present in the section prior
to treatment also react with ammonium sulphide (NRthS. A control section, where
treatment with lipase has been omitted, is therefore necessary.
As the enzymatic degradation of the lipid and soap formation takes place at the
water-lipid boundary, larger lipid droplets will only be stained on their circumfer-
ence.
Crystalline cholesterol is nonnally stained blue with the reaction, while disperse
or lipid dissolved cholesterol will be stained red. The colour is stable for several
hours, and the method is relatively gentle to the tissue. For this reason the PAN
method is much preferred to the previously used Schultz' methods, which use 5-8
mol/l H2S04.
Carbohydrate deposits may sometimes give a weak red staining with PAN and
an acetone extracted control section should therefore always be included.
Some of the most important methods used for demonstrating nucleic acids are
enumerated in Table 20.1, while Table 20.2 outlines a strategy for their identification
using the classical histochemical reactions.
Demonstrates Quantitative
Method DNA RNA Selectivity method Reference
Under well-defined conditions, cationic dyes such as Toluidine Blue can bind
selectively to nucleic acids. One of the conditions is that pH should be sufficiently
low for proteins to exhibit a positive net charge, i.e. pH not over 3 (cf. basophilia
Sect.20.2).
Another approach is to use specific probes for the detection of the nucleic acids.
These affinity probe techniques can be divided into:
H. Lyon (Ed.)
Theory and Sttalegy in Histochemistty
@ Springer Verlag 1991
272 P.E. Hl'lyer, A.K.N. Iversen, E. Schulte, H. Lyon
Table 20.2. Identif1cation of nucleic acids in a tissue component which exhibits basophilia.
20.2 Basophilia
The use of Toluidine Blue for demonstrating nucleic acids is diseussed in Seet.6.1.1,
while the Methyl Green-Pyronin method is treated in Seets.6.1.5 and 28.8.4. The
use of Cuprolinic Blue and Chromium-galloeyanin ean also be included under this
heading.
Chromium-galloeyanin. This metal complex dye (cf. Sect.7.3) binds to the phos-
phate groups of the nucleic acids.
The Feulgen nucleal re action is the best method for in situ quantitation of DNA
in nuclei. The Feulgen reaction in itself does not give any information as to the
functional condition of the nucleus (celI).
In situ investigations on the localization and relative amounts of RNA were
performed by Brachet (1940a; 1940b; 1942; 1953) and formed the basis for the
first hypothesis on the exchange of information and transport:
nucleus (DNA -t RNA) -t cytoplasm (RNA -t protein).
Brachet used the Methyl Green-Pyronin reaction coupled with RNase for the spe-
cific demonstration of RNA. This method is fast, reliable, and gives a reasonably
good histological picture. For cytological details it is less weIl suited than Tolu-
idine BIue or Cresyl Violet Acetate for example. The majority of cationic dyes
are metachromatic and therefore not weIl-suited for quantitative studies. For this
purpose Cr-gallocyanin is preferable.
Brachet's investigations showed that, in most cases, there was a functional
connection between the amount of RNA in the cytoplasm, the size of the nucleolus
and the content of non-histones in the nucleus for one cell type. Consideration of
the level of protein synthetic activity should include these measurements as weIl
as the ratio between the volumes of nucleus and cytoplasm (Sect.2.2.1). In normal
circumstances, the following cell types contain high amounts of RNA and large
nucleoli (usually):
• fibroblasts, chondroblasts and osteoblasts,
• plasma cells, both free and in active lymph nodes,
• nerve cells,
• exocrine protein secreting cells,
• epithelial cells continually being replaced (e.g. in the stratum germinativum of
the skin, and in the intestines),
• embryonic cells ulldergoing rapid divisions,
• many tumours.
autoradiography and also that results may be obtained in much sharter times. Sen-
sitivity can be increased by exposing cells to be labelIed with BrdU simultaneously
to 5-fluoro-2'-deoxyuridine, an inhibitor of thymidilate synthetase, thus increasing
BrdU incorporation by decreasing competition from endogenous thymidine (Ell-
wart and Dörmer, 1985). It is necessary to denature cellular DNA to allow access
of BrdU. This can be achieved by nuclease digestion simultaneously with the an-
tibody incubation (Gonchoroff et al., 1985). Nuclease treatment can be substituted
using microwave irradiation (van de Kant et al. 1990).
Another approach was described by Apte and Puddle (1990). Their use of
sodium ethoxide to remove plastic from tissue sections made aseparate denaturation
step of DNA unnecessary. The effect of different fixatives was compared. Best
results were obtained with formaldehyde or Bouin's fluid.
Several authors (Hamada, 1985; Morstyn et al., 1986; Apte and Puddle, 1990)
have found a high correlation between the incorporation of 3H-thymidine in DNA
as detected by autoradiography and data based on the BrdU technique.
Frederiks et al. (1990) using BrdU-immunocytochemistry on isolated hepato-
cytes after partial hepatectomy in rats found the same labelling index of binuclear
diploid, mononuclear tetraploid, and binuclear tetraploid cells. There did not appear
a special role for mononuclear diploid cells in proliferation.
Visualization. This can be achieved by labelling the probe in one of the following
ways:
1. RISH, radioactive in situ hybridization using suitable radioactive elements such
as 125 1 or 3H followed by autoradiography (cf. Chap.29)
20 Nucleic Acids 275
In the first type of in si tu hybridization the target nucleic acid sequence can
be either nucleic DNA or nucleic/cellular RNA. Conceptually, the hybridization
to the two types of nucleic acid is quite similar, but the technical details differ
significantly. In both cases, in order to control hybridization conditions the labelIed
strand of nucleic acid, the probe, must have a known length. Whether the sequence
of the probe has to be known in detail depends primarilyon the problem. The
stringency calculations only call for a knowledge of the approximate content of
guanine-cytosine base pairs (cf. specificity).
In the second type of in situ hybridization called primed in situ labelling
(PRINS, Kock et al., 1989), the target can be either DNA or RNA. The oligonu-
cleotide primer is usually not longer than 25 bases as it is difficult to calculate
the optimum temperature when the primer is longer. The DNA polymerase used
varies with the nature of the template. When this is DNA, the Klenow fragment
of E. coli, DNA polymerase I, is most frequently used, while with RNA as the
template reverse transcriptase is usually employed.
Hybridization can only take place when the probe and target sequences are
single stranded (following denaturation). The conditions used should be adjusted
so that only complementary sequences can hybridize. Depending on the chosen
probe and target, the resulting double strand can be DNA:DNA, DNA:RNA, or
RNA:RNA. DNA:RNA hybrids have the advantage of having a higher optimum
hybridization temperature.
Several reagents are usually added to the hybridization solution to ensure as
specific areaction as possible. Polyethylene glycol or polymers of dextran sulphate
form networks in the solution from which the probe is excluded thereby increasing
its relative concentration and the rate of reaction of the hybridization process.
EDTA (Sect.15.4) blocks the activity of contaminating nucleases and the addition of
form amide up to 50% effectively reduces the optimal temperature of hybridization
(for DNA the temperature is reduced 0.65°C for each per cent form amide while the
equivalent reduction is 0.38°C for RNA). In addition, Larsson and Hougaard (1990)
found that hybridization temperatures between 40-45°C produce the best signal to
noise ratio. Moreover, the inclusion of 50% form amide produces an enhanced signal
to noise ratio in spite of higher background staining (Larsson and Hougaard, 1990).
Finally, denatured heterologous nucleic acid sequences, usually derived from
salmon sperm DNA, are included. These form electrostatic bonds with positively
charged components in the sampie thereby reducing the opportunity for involve-
ment of the probe in similar interactions. The effect of these blocking factors is
maximized by introducing a prehybridization step using a solution composed as the
hybridization solution without the addition of the probe. Although hybridization is
usually complete in 3-5 hours, overnight incubation is commonly employed for
convenience. The hybridization time should, however, be empirically tested as too
long a hybridization time might reduce the sensitivity of the assay.
higher cation concentrations (low stringency) give greater opportunity for partial
homologous annealing as mismatches are stabilized In practice, with conditions
of low stringency, one in situ hybridization procedure can demonstrate several
related sequences such as those in HPV subtypes (human papilloma virus), while
high stringency makes it possible to distinguish between different subtypes of HPV
which differ only slightly in sequence.
Washing. Following the hybridization step was hing is performed usually first as a
low stringency wash (high concentration of monovalent cations and/or low temper-
ature, and/or with or without low concentrations of formamide ) to remove unböund
or loosely-hybridized probe together with the other components of the hybridization
solution. This is followed by a high stringency bath (low concentration of mono-
valent cations, and/or high concentrations of formamide, and/or high temperature)
to "fine-tune" the specificity of probe hybridization.
ControIs. Control reactions are necessary to ensure the specificity of the in situ
hybridization reaction. These include:
1. Nucleases for removing DNA or RNA. This determines the class of nucleic
acid to which the probe has hybridized. As RNase usually is contaminated with
DNase and vice versa the unwanted enzyme must first be removed (cf. Sect.3.4)
2. When probes are carried on a vector such as a plasmid, it is essential to include
a labelled vector (without probe) control to determine the extent to which direct
vector hybridization is contributing to the result
278 P.E. H~yer, A.K.N. Iversen, E. Schulte, H. Lyon
It should be noted that even though PCR can be used for demonstrating specific
nucleotide in single cells, cell spreads, or even paraffin sections, the reaction is not
cytochemical in the strict sense since the amplified sequence is not demonstrated
in situ in the cello The aim of this section is to provide an introduction to PCR
in recognition of its potential as an adjunct to histochemical studies. For com-
prehensive reviews of PCR readers are referred to: Innis et al. (1989) and Erlich
(1989).
PCR technology has an extensive range of actual and potential applications.
As well as producing more target DNA for hybridization studies, the amplified
nucleotide sequence can also be used to study DNA-protein interactions and for
nucleotide sequence determination. Different primer sets can be used; these can
enable differentiation of more subtle genetic differences such as ,yirus subtypes
present in tissue sampies. Gene expression can be studied by detecting and even
quantitating RNA transcripts after first making copy DNA (cDNA) using reverse
transcriptase (Krawetz et al., 1989; Gilliland et al., 1989). Finally, PCR can itself
be used for the production of probe DNA for in situ hybridization.
Mechanism. The polymerase chain reaction can be described as a three step cycling
process: Denaturation-Annealing-Extension (Fig. 20.1).
The first two steps are, in principle, the same as those used in in situ hy-
bridization with annealing taking place between the two primers and the flanking
nucleotide sequences on either side of the target DNA fragment to be amplified
Both these steps take about 20 to 60 seconds. In the third step, extension, the tem-
perature is raised to 70-75°C, the temperature optimum of TAq DNA polymerase,
and the annealed primers are extended in the 5' to 3' direction.
The time necessary to secure full synthesis of the desired fragment, depends
on the length of the fragment being amplified. Under standard conditions the rec-
ommended time is 1 min per kB. Depending on the nature o{ the DNA template,
incorporation can approach 150 nucleotides/sec/enzyme molecule at 75-80°C (Innis
et al., 1989). The cycle can then be repeated, allowing re-annealing of the primers
and further rounds of DNA synthesis. Between 25 and 35 cycles are employed for
most purposes. Each cycle is achieved by stepwise changes in temperature without
the need for addition of new reagents. The so-called "PCR machines" are therefore
programmable heating/cooling systems that achieve the appropriate temperature-
time profile required for the proposed reaction.
Although each cycle should theoretically double the amount of target DNA,
in practice around ten-fold less than the expected quantity is produced. After an
initial exponential amplification, a linear increase in DNA occurs with each cycle
and the process becomes less efficient
~iiiiiiiilllllliiilllllliiii""'iiiillll"'iiilllllliiiilll"'iiiillll~
& 1"qc1dWilllrMC '
Unamplilied DNA
o
o Primer extension
illllllliiiillllliiiillll"iiilllllllliiii~
Cycle 2
~~-----------------o--~
__~o~mm~~__......o
o~""----wmm=mro~~
Denature and anneal primers
~~o~mm~~ __________~~
~~IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII~IIIIIII
0 Y'.A
o l1li11111l1li1 0
Primer extension
o 1"111l1li1111 0
o IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII~
I
Cyclc 3
~r------------,.m= .. ~o~
__~mmmmL-__~. .~O
o
o Denature and anneal primers
o
o
o
~"'IIIIIIII"'"IIIIIIII"'IIIIIII"IIIIII o
o 111"''''11''' o
o 'iiiiliiii!lll
o Primer extension
o liiiillliliii' o
o i!iiililli liii
o
o iiililllIlii!l 0 ~
o IIIIIIIIIIIIII"'IIIIIIIIIIIIIIIIIIIIIIIII~
Cyclcs 4 · 25
Fig. 20.1. Polymerase chain reaction technique (PCR) based on repeated cycles of high tempera-
ture template denaturation, oligonucleotide primer annea1ing, and polymerase mediated extension.
Used with permission of Perkin-Elmer Cetus.
20 Nucleic Acids 281
Source of Target Sequences. Target nucleic acid is extracted from the source
material prior to amplification. The original source can be fresh frozen tissue/cells
(DNNRNA-cDNA) or formaldehyde fixed paraffin embedded tissue/cells (DNA).
Unbuffered formaldehyde soltltions or fixatives containing picric acid (Bouin) or
mercuric chloride (Zenker, B5) cause severe degradation of DNA (Nuovo and
Silverstein, 1988; Dubeau et al., 1986).
If fresh tissue!cells are used, and if the fragment to be amplified is present as
one copy per cell, only a few cells are necessary, and no particularly efficient DNA
extraction is called for. If, however, more than 600 cells are necessary in order to
get enough template sequences, DNA extraction must be performed (Innis et al.,
1989; Maniatis et al., 1989).
If the source of DNA templates is formaldehyde fixed paraffin embedded tissue,
three different ways to extract the DNA have been published:
1. Using xylene/ethanol (Dubeau et al. 1986; Shibata et al., 1988)
2. By boiling the section(s) in a chelating agent (Chelex) (Singer-Sam et al., 1989)
3. Using proteinase K/phenoVchloroform (Impraim et al., 1987)
The yield of DNA is often only about 60-70% of what a similar amount of
fresh tissue would give (Dubeau et al., 1986).
Nucleic acids may be extracted from the tissue either non-enzymatically or en-
zymatically. For non-enzymatic extraction trichloroacetic acid is usually preferred.
This treatment primarily leads to the extraction of RNA but, depending on the tem-
perature and the length of the treatment, DNA is also extracted to a greater or lesser
degree. Hot trichloroacetic acid removes nucleic acids completely, after which the
histones may be demonstrated with an anionic dye (Sects.6.2.5 and 21.4.2).
Perchloric acid and hydrochloric acid extract histones simultaneously with the
nucleic acids. As conditions for the extraction of RNA vary from cell type to cell
type and from fixative to fixative it is best to treat the tissue with the enzyme
282 P.E. H~yer, A.K.N. Iversen, E. Schulte, H. Lyon
ribonuclease (RNase), which is specific for RNA. The purity of the enzyme is
critical (Sect.3.4.3). Dichromate containing fixatives render RNA very resistant
towards RNase. Aqueous formaldehyde solutions render RNA more resistant to
RNase, while alcoholic formaldehyde solutions on the other hand seem to promote
the degradation of RNA by RNase. It should also be noted that many salts (buffers)
may have a direct extractive effect on RNA, especially at higher temperature.
In general, the same principles apply to deoxyribonuclease (DNase) as those
outlined for RNase.
21 ProteiDS
21.1 Introduction
Some proteins, such as the fibre proteins elastin and collagen, can be identified
direcdy by their tissue location and their physico-chemical properties. Similarly,
actin and myosin can be recognized in striated museie by virtue of their organization
into actomyosin, which appears cross-striated.
In general, selective demonstration of proteins according to amino acid compo-
sition is only possible for proteins which are present in a very high concentration
at certain locations (Table 21.1). This approach depends on the protein concemed
having an exceptionally high content of one or more of the amino acids that can be
demonstrated histochemically (Cys, Trp, Tyr, Arg, etc.). Normally, individual pro-
teins occur at low concentrations mixed with many other proteins. In most cases,
selective demonstration of a given protein can therefore only be accomplished by
using its biological properties such as enzyme activity (Chaps.23-25) or specific
antigenicity (Chap.26).
21.2 Fixation
21.3 Demonstration
IL Lyon (Ed.)
TheoJy and SIralegy in Histochemisuy
© Springer Verlag 1991
284 H. Lyon, P.E. H~yer, P. Prent~
Frequency
(mean number
Amino acid per 1(0) Occurrence Methods
~H]-2
neurosecretion. serum albumin RSR, DDD, ferricferricyanide,
maleimide
exocrine and endocrine granu- reduction-RSR, etc. oxidation-
les in pancreas cationic dye
Cys-S
Notes:
*: Basic amino acid residues. Can be demonstrated with reactions for acidophilia.
**: Acid amino acid residues. Can be demonstrated with reactions for basophilia.
R
I
©O
H2C-C~
0;; H
NH-R
+ HCHO -+
Reactions for Protein Bound Cysteine and Cystine. Reliable histochemical dis-
tinction between -SH and -S-S- is not possible. A reasonable evaluation can, how-
ever, be achieved using trichloroacetic acid fixed material as follows:
Table 21.3. Staining of proteins with anionic dye at high pH and varying pretreatments.
pH 8 8 10.5
pretreatment none TCA none
Cytoplasm stained
Certain globins (myoglobin in muscle cells and haemoglobin in red + + 0
cells)
Ribosomal basic pro teins 0 + 0
Dissociated ribonucleoprotein from ribosomes in certain patholo- + (+) 0
gical conditions
Proteins (lysozyme) in eosinophils and Paneth cells + + +
Nuclei stained
Dissociated nucleoprotein (histones) in certain pathological condi- + (+) (+)
tions. Mitotic chromatin and spermatozoa also frequently react
Histones and proteins after DNA extraction 0 + +
21 Proteins 287
Masked BasophiIia; Cationic Dyes at Low pH. With dyes capable of staining
metachromatically, the reaction is termed masked metachromasia. This approach
has been developed into a procedure for demonstrating polypeptide hormones in
cells belonging to the APUD system (Solcia et al., 1968). These include calcitonin
(C-cells in the thyroid), glucagon (A-cells in the pancreas), insulin (B-cells in the
pancreas), and gastrin (G-cells in the stomach), all of which show a high content
of glutamie and aspartic acids (acidie amino acids) or their amides. All are found
in 0.2-0.4 11m diameter secretory granules. Treatment with HCI removes RNA and
simuItaneously converts the amides glutamine and asparagine to the corresponding
acids. These can subsequently be demonstrated with a cationic dye at pH 4.
The procedure has been modified for use on Epon embedded tissue. This en-
ables the identification of positive areas by light mieroscopy, and these can then
be selected for ultrathin sectioning and electron microscopic examination (Pabst,
1985).
Table 21.4 shows a comparison between collagen and elastin. The higher pI of
collagen is due to a higher content of arginine, histidine, and lysine, while the
principal non-polar amino acids in elastin are tyrosine, leucine, and valine.
anionic or cationic dyes to elastin should therefore be possible providing the ability
of these dyes to form pure ionic bonds is suppressed. This approach is used in the
methods mentioned under point 3 in Table 21.5.
I. Autoftuorescence (30.1)
2. Diazotization coupling reaction for tyrosine (9.4.3)
3. Use of dyes:
Selected anionic dyes at high pH, e.g. Biebrich Scarlet
Selected cationic dyes at low pH, e.g. Victoria Blue 4R
Selected metal complex dyes, e.g. Verhoeff's Iron-haematoxylin
Aldehyde Fuchsin without preceding oxidation
Orcein
Resorcin Fuchsin
Goldstein (1962) has shown that binding to elastin by both anionic and cationic
dyes, including Aldehyde Fuchsin, Orcein, and Resorcin Fuchsin, is not hindered to
any appreciable degree by the addition of sodium chloride to the staining solution.
This finding makes it highly unlikely that ionic bonds are important in the staining
of elastin by these dyes.
The importance of van der Waal attractions has been stressed by Horobin and
Bennion (1973). One consequence of this is that most of the dyes involved can
be dissolved in organie solvents without losing their ability to stain elastin. It
should however be re-emphasized (cf. Sect6.3) that. depending on the prevailing
conditions, a dye binds to its substrate by al1 the bonding mechanisms available to
it. Proctor and Horobin (1988) have presented further work on chemical structure
and staining mechanisms of elastic fibre stains supporting this view.
Among the different methods for demonstrating elastin the Orcein methods can
be singled out for their high selectivity and the technical convenience. In contrast,
Verhoeff's Iron-haematoxylin is not recommended on account of its low selectivity,
poor reproducibility, and its difficult differentiation step.
Van Gieson Methods. The van Gieson methods probably give the most selective
staining of collagen fibres. The 25-50-fold excess of Pieric Acid relative to the
collagen dye in a van Gieson solution renders the cytoplasmic background "Picric
Acid yellow". Puchtler et al. (1988) investigated a number of anionic dyes for
their ability to stain collagen selectively when used together with Picric Acid, and
concluded that dispersion forces (Sect.4.5.4) and hydrophobie bonding (Sect.4.5.5)
were decisive, whereas ionic forces did not impart affinity. For the collagen dye it
is advantageous to use long, planar molecules such as Sirius Red 4B (C.!. 28160)
and Sirius Red F3B (C.!. 35780) as these become oriented parallel to the collagen
fibre structure and increase birefringence. Red staining with "Piero-Sirius F3B"
accompanied by a pronounced increase in birefringence is an unequivocal indication
of the presence of collagen and reticulin fibres (Junqueira et a/., 1979; Montes et a/.,
1980) providing the presence of amyloid can be ruled out (see Sect.21.7). In contrast
to the above, the classieal "Piero-Fuchsin" method does not demonstrate the thinnest
collagen fibres and therefore excludes both retieulin fibres and basal membranes.
Acid Fuchsin does not lead to an increase in birefringence associated with collagen
fibres. Wolman and Kasten (1986) have produced a detailed discussion of the use
of polarized light mieroscopy for studying the molecular structure of collagen and
retieulin.
21. 7 Amyloid
c. Amyloid in multiple myelomas and other diseases of the immune system with
the same distribution as (a)
d. Tumour associated amyloid is particularly associated with apudomas, Le. tu-
mours derived from cells in the APUD system
Ultrastructurally amyloid consists of irregularly deposited fibrils, each consist-
ing of two filaments. X-ray diffraction analysis has shown that amyloid filaments
possess a ß-pleated sheet conformation.
Table 21.6 sets out the results of chemical analyses performed on amyloid.
Three different kinds of proteins have been found.
Congo Red, Sirius Red F3B, Toluidine Blue 0, and Thioftavine TCN are bound
to the ß-pleated sheet eonformation of amyloid by hydrophobie bonds. In addition,
it is known that the planar and linear dye moleeules are orientated parallel to the
amyloid fibrils (Glenner, 1981). Non-specifie bonding of the dyes by ionie bonds
is kept at a minimum by inereasing pH of the staining solutions and by adding
inorganie salts, e.g. NaCl, and organie solvents, e.g. ethanol. Some indieation of
whether an amyloid deposit is primary or seeondary ean be obtained by pretreat-
ment with trypsin or (more effeetively) an acid potassium permanganate solution
(Wright et al., 1977). Primary amyloid is resistant to these treatments and eontin-
ues to give a positive reaetion with Congo Red. Amyloid associated with multiple
myeloma, despite the similarity of the protein to that found in primary amyloidosis,
is frequently degraded. The meehanism involved in staining with Crystal Violet is
probably similar but the reason for the reddish eolour is uneertain.
The preferred method is staining with Congo Red. It should be noted that to
aehieve the optimum "apple green" eolour seen by polarization microseopy, it is
neeessary to use tissue seetions whieh are at least 10 p,m thiek.
22 Carbohydrates
22.1 Introduction
22.2 Demonstration
Fixation has been covered in Sect.13.4. The most frequently used methods for
demonstrating carbohydrates are given in Tables 22.1 and 22.2. The periodic acid-
Schiff reaction (PAS) is discussed in detail in Sect.9.2.1, while metachromasia
and the Alcian Blue reaction are discussed in Sect.6.1.1 respectively Sect.6.1.2. In
the following, modifications of these methods and supplementary methods will be
discussed, while the enzymatic extraction methods are expounded in Sect.22.3.3.
Note that selective identification of acid carbohydrates using the AB/MgClz and
the PAS-reaction is only possible if the PAS- reaction is negative. For example,
a mixture of a sulphate containing proteoglycan and a neutral glycoprotein will
give the same staining result as sulphomucin. The histological localization and
morphology of the stained material is therefore essential for the interpretation.
This method (Reid et al., 1973; Culling et al., 1974; 1976) depends on three pro-
cesses:
1. Reduction of aldehyde groups induced by periodic acid oxidation (or in other
ways) to primary alcohol groups with borohydride, BH4 , thereby blocking any
subsequent reaction with Schiffs reagent
H. Lyon (Ed.)
Theory and Strategy m Hlstochemistry
© Springer Verlag 1991
N
Table 22.1. ReactlOns of histochemically demonstrable carbohydrates m formaldehyde fixed, paraffin embedded tissue. 'f
HOMOGL YCANS PROTEOGL YCANS GL YCOPROTEINS
polycarboxy-
polycarboxylates sulphates polysulphates slalomucms sulphomucms
PAS + + 0 0 0 + + +
Am-PAS 0 + 0 0 0 + + +
MCpH 2 0 0 0 + + 0 0 +
MCpH 5 0 0 + + + 0 + +
AB 0 0 0 + + + 0 + +
AB 0.2 0 0 0 + + 0 0 +
AB 0.5 0 0 0 01+ + 0 0 +
AB 0.7 0 0 0 01+ + 0 0 01+
AB 0.9 0 0 0 0 01+ 0 0 0
AB pH 1 0 0 0 + + 0 0 +
AB pH 3 0 0 + + 01+ 0 + 01+
AF 0 0 0 + + 0 0 +
LID 0 0 + + + 0 + +
HID 0 0 0 + + 0 0 +
+: positive reaction; 0: no reactlon; Am-PAS: amylase-PAS; MC pH 2: Toluidine Blue metachromasla at pH 2; MC pH 5: Toluidine
Blue metachromasia at pH 5, non-dehydrated section; AB 0: Alclan Blue 0 moljl MgCI 2 ; AB 0.1: Alcian Blue 0.1 moljl MgCI 2 ; AB 0.2: ;:c
Alcian Blue 0.2 moljl MgCI 2 ; AB 0.5: Alcian Blue 0.5 moljl MgCI 2 ; AB 0.7: Alcian Blue 0.7 mol/l MgCI 2 ; AB 0.9: Alclan Blue 0.9 mol/l
MgCI 2 ; AB pHI: Alcian Blue pH 1, AB pH3: Alcian Blue pH 3; AF: Aldehyde Fuchsin pH 1-2; LID: Low lron diamme; HID: HIgh Iron j
diamine. "1::1
rn
g:
'<
~
:-0
I
22 Carbohydrates 295
J
a: PAS +--+2
b: PAS 0--+ 3 or non-carbohydrate
2. Amylase - PAS «(X-homoglycans):
a: PAS 0 = (X-homoglycans (glycogen, amylose, amylopectin)
ß-homoglycans (e.g. cellulose)
other homoglycans (e.g. dextrans)
b: PAS + { ------~)
1-. 3
glycolipids (in frozen sections)
glycoproteins
3. Alcian Blue pH 3 without MgCI 2 :
PAS + = acid g1ycoproteins
a: AB+ { JI----~)4
b: ABO [
PAS 0 = acid proteoglycans
PAS 0 =
J
neutral proteoglycans* or non-carbohydrate
x
-C
4. Alcian Blue pH 5.7 with 0.2 moljl MgCI 2 :
PAS + = sulphate containing mucins ]
a: AB + 1-------+) 6
PAS 0 = sulphate containing proteoglycans
~
PAS + = sialomucins (possible control with neuraminidase)~5
b: ABO
PAS 0 = polycarboxylate proteoglycan (e.g. hyaluronic
acid (possible control with hyaluronidase)) or
nort-carbohydrate
5. Periodic acid-Thionin/KOH/PAS method (PAT/PAS):
a: PAT + (blue) = non-acylated sialomucins
b: PAS + (red) = O-acylated sialomucins
-C
6. AIcian Blue pH 5.7 with 0.7 mol/I MgCI 2 :
PAS + = sulphomucins ~
~AB+ 7
PAS 0 = probably polysulphate proteoglycans
PAS + = sulphomucins
b: ABO [
PAS 0 = carboxysulphate proteoglycans (e.g. chondroitin
sulphates, heparin, heparan sulphate)
296 H. Lyon, P.E. H\'!yer, P. Prent\'!
PAS + = sulphomucins
b: ABO [
PAS 0 = carboxysulphate proteoglycans
non-acylated SM O-acylated SM
Procedures 1 and 2 are identical, except for the colour of the final product.
Procedures 2 (1) and 3 can be performed on consecutive sections, or if the two
kinds of sialomucins are present in approximately equal amounts in the section, it
may be satisfactory to only perform procedure 4. In humans O-acylated sialomucins
are not present in the gastric mucosa but appear in the intestines in increasing
amounts towards the anus.
22 Carbohydrates 297
Volz et al. (1986) note that the use of the PAS reaction for the histochemical
identification of sialic acids is complicated by oxidation of vicinal diols on other
carbohydrate residues. One solution to this problem is to selectively oxidize with
dilute periodie acid ("mild" periodie acid) (Weber et al., 1975). Another solution is
to use the periodie acid-phenylhydrazine Schiff method (PAPS) (Spicer, 1961). The
mechanism of the latter method is that the aldehyde groups engendered by periodie
acid are blocked with phenylhydrazine. Subsequent treatment with Schiff's reagent
reverses the blocking of sialic acid monoaldehydes (Reid et al., 1984). Volz et al.
(1987b) showed that the use of 0.4 mmol!l in 1 mol!l HCl for one hour at 4°C leads
to the selective visualization of sialic acids in the PAS procedure. This selectivity
is a result of an increase in the rate of the oxidation of the sialic acid residues
together with a decrease in the rate of oxidation of neutral sugars (vicinal diols
located on hexose, 6-deoxyhexose, or N-acetyl-hexosamine residues of sialo- and
sialosulphoglycoproteins).
At this juncture it is worth pointing out that an exhaustive range of additional
sequential treatments have been proposed on the basis of the investigations of
Volz, Reid, Park, and coworkers. To assist in identifying these new sequences, we
have chosen to number them consecutively and to list the abbreviations used in
alphabetic order.
AB 1.0 0.1 % w/v Alcian BIue in 0.1 mol!l HCI for 30 min at room temperature
Az Azure A Schiff reagent for 6 hours at room temperature
Bh 0.1 % w/v sodium borohydride in 1% w/v dibasic sodium phosphate
(anhydrous) for 20 min at room temperature
DNPH saturated solution of 2,4-dinitrophenylhydrazine in 1 mol!l HCI for 2
hours at 4°C
KOH saponification with 0.5% w/v potassium hydroxide in 70% v/v ethanol
for 15 min at room temperature
PA(2) 1% w/v (44 mmol!l) aqueous periodie acid for 2 hours at room temper-
ature
PA(1) 1% w/v (44 mmol!l) aqueous periodie acid for 1 hour at room temper-
ature
PA(1/2) 1% w/v (44 mmol!l) aqueous periodie acid for 30 minutes at room
temperature
PA* 0.4 mmol!l periodie acid in 1 mol/l HCI for 1 hour at 4°C
S Pararosanilin Schiff reagent for 1 hour at room temperature
T Thionin Schiff reagent for 2 hours at room temperature
Reid et al. (1988) have presented a new general method for the specific histo-
chemical identification of O-acyl sugars in any epithelial glycoprotein. The term
O-acyl sugar indicates the presence of 8- (or 9-) O-acyl sialic acids and an ester
substituent(s) on all the potential vicinal diols of the hexose, 6-deoxyhexose, and
N-acetylhexosamine residues (for nomenclature, see Sect.2.1.5).
10. PA-Bh-KOH-PA*-Bh-PAS
The initial Pa-Bh treatment renders vicinal diols located on either sialic acid
or neutral sugars PAS unreactive. In the subsequent steps ester substituents are
removed by the PA*-Bh sequence, and O-acyl sugars are stained with the PAS
technique. With this method it has been demonstrated that O-acyl sugars occur in
the epithelial goblet cell glycoproteins of adult human colon (Reid et al., 1988).
The staining results expected from the ten new variants cited above are sum-
marized in Table 22.4.
Table 22.4. Staining results of material containing epithelial glycoproteins using the more recent
developments of Culling's methods.
1 M 0 M 0 T T
2 M 0 M T T T
3 0 0 M T T 0
4 M B M 0 0 0
5 M 0 M 0 0 0
6 Y 0 Y A A A
7 0 0 y A A A
8 Y 0 Y 0 A A
9 M 0 Y A A M
10 M 0 0 0 0 0
In addition to its use for staining RNA (Sect.20.2), Cuprolinie Blue has also been
used for the staining of acid glycoproteins and proteoglycans (Scott, 1980; Van Kup-
pevelt et al., 1984a; 1984b). With Cuprolinie Blue applied to semithin epoxyresin
embedded material, Juarranz et al. (1987) have obtained metachromatic staining of
goblet cell mucin, mast cell granuIes, and cartilage matrix (Sect.6.1.1). Electron
microscopy of ultrathin sections of the same material showed an eIectron dense
reaction in the same structures. Using Cuprolinie Blue combined with enzymatic
digestion, Hussein et al. (1988) have demonstrated that heparan sulphate is the
main glycosaminoglycan in the basement membrane of human gall bladder.
300 H. Lyon, P.E. Hf/lyer, P. Prentf/l
22.3.1 Acetylation
As stated in Sect.9.2.1 the first step in the PAS-reaction can be blocked by acylating
the vicinal hydroxyl or hydroxyl and amino groups. The ease with which the groups
are blocked, however, varies considerably as shown in Table 22.5.
Table 22.5. The infiuence of the length of acetylation for the blocking of the PAS-reaction of
different tissue components.
Tissue component 2 6 16 20
Collagen, basal membranes, reticulin, lipofuscin, colonic "melanin", 0 0 0 0
and brush border in tubules of the kidney
Intestinal mucins, cornea, corpus vitreum, and capsule of the lens + 0 0 0
Glycogen, starch, cellulose, mucin in the stomach, and chromaffin + + 0 0
in the adrenal glands
Cartilage matrix + + + 0
These methods can be used for differentiating between different acid carbohydrates.
The principle has been discussed in Sect.9.8.3 and a summary is given in Table
22.6.
22 Carbohydrates 301
Table 22.6. The effect of mild (MM) and drastic methylation (DM) combined with saponification (S)
on the binding of cationic dyes (C) to acid carbohydrates.
polycarboxy-
Reactions polycarboxylates sulphates polysulphates sialomucins sulphomucins
C (+) + + (+) +
MM/C 0 0 0 o 0
MM/S/C (+) + + (+) +
DM/S/C (+) (+) 0 (+) 0
The value of these reactions is, however, limited in two respects. First, the
reactive groups show very great individual variation in their willingness to react
and second, the saponification reaction is very harsh on the tissue. Broadly, block-
ing is achieved quickest for sulphate in proteoglycans and sulphomucins followed
by the carboxylate groups in proteoglycans, sialomucins and proteins. A longer
methylation step is required to block the basophilia of lipofuscin and even longer
for melanin.
22.4 Lectins
Lectins are proteins with a molecular weight between 200 and 300 kDa which can
be extracted from various plants. They bind specifically to certain carbohydrates in
a manner that often corresponds to the hydrogen and hydrophobic bonds formed
between antigens and antibodies. It is possible to label lectins in exact1y the same
way as antibodies (see Sect.26.2). Examples of labelling include Fluorescein Isoth-
iocyanate (FITC), ferritin, and horse radish peroxidase (HRP). It is also possible
to label with 3H-acetal groups (autoradiography, Table 29.1).
The specificity of the lectin must be determined by control stainings. The most
important control is the inhibition of the staining by addition of high concentrations
of carbohydrates which bind the lectin in question. For instance, concanavalin
A is specific towards a-O-mannose and a-O-glucose. Oye binding can also be
hindered by chemical blocking; for example, acetylation hinders the binding of
concanavalin A.
The effects of different fixation protocols on the subsequent binding of different
lectins have been evaluated by Allison (1987). There is no universally appropriate
fixative, but the best results are usually obtained with 95% ethanol or fixatives
containing mercuric chloride. Formol-saline is, however, adequate for the study of
routine paraffin-processed tissue in many instances. There are nonetheless situations
where frozen sections may be preferable.
Jeffrey et al. (1987) have stressed the importance of using purified enzyme
preparations (notably trypsin) in the pretreatment of sections from formalin fixed
paraffin embedded material for lectin histochemistry.
The use of lectins is thus still associated with many practical problems but is
expected to assist in a more precise classification of tissue carbohydrates.
Part 5
Enzyme Histochemistry
23 Enzyme Histochemistry I: General Considerations
23.1.1 Definition
Enzymes are proteins or glycoproteins with selective, often specific, catalytic ef-
fects. They characteristically increase the rates of reactions by a factor of at least
10'.
23.1.2 Specificity
The specificity of an enzyme results from the manner in which the substrate is
bound 10 the enzyme.
The structurallocation at which the substrate is bound is called the active site
of the enzyme. It is only a tiny portion of the entire enzyme but results from a
complicated three-dimensional structure consisting of groups from several points
in the linear amino acid sequence. This complex structure limits form and size of
molecules that can be bound and thereby confers substrate specificity.
In many cells aseries of enzymes function together in a multienzyme system.
A compound is processed through a number of steps where the product from one
process is the substrate for the next.
H. Lyon (Ed.)
Theoty and Strategy in Histochemistry
© Springer \\lrlag 1991
306 A.P. Andersen, PE. H~yer, H. Lyon, P. Prent~
C + D
+
Fig. 23.1. The activation energy for the process A +
B +=t C D
I non-catalyzed reaction
n catalyzed reaction
Energy (E).
Compounds A and B can be transformed to C and D, if the energy level of A + B is higher than
that of C + D and E = activation energy is present.
Table 23.1. Comparison of KM values for selected enzymes obtained by histochemical and biochemi-
cal assays (modified after van Noorden and Butcher, 1990).
Cytochemical Biochemical
assay assay
Enzyme Technique KM (mmoljl) KM (mmoljl) References
PV A = polyvinyl alcohol.
Note that when using PVA in cytochemical assays, KM is often considerably larger than that
determined by biochemistry. This is probably due to a decrease in diffusion of the substrate in the
viscous medium.
Vrnax is the rate of reaction when the enzyme is fully saturated with substrate.
It is an expression of the molecular activity of the enzyme = "turnover number" =
number of substrate moleeules turned over in one minute per enzyme moleeule.
Examples of maximum molecular activities are given in Table 23.2, and graph-
ically depicted in Fig. 23.2.
Table 23.2. Number of substrate molecules turned over per minute per
enzyme molecule (molecular activity) by different enzymes (determined
biochemically).
Acetylcholinesterase
Chymotrypsin
Kinases
Peroxidase
[S]
V = Vrnax [S] + KM
For low [S] and with [S] ~ KM applies KM + [S] ~ KM
[S]
V = Vrnax KM = k[S]
308 A.P. Andersen, PE. Hjilyer, H. Lyon, P. Prentfil
vmax ---------------------
1/2 Vmax
[5]
For high values of [S] the reaction is of zero-order (rate of reaction is constant
and independent of the substrate concentration).
KM and Vmax can be determined by measuring the rate of reaction (initial
velocity) at different [S] values.
Reaction Rate and pH. If the enzyme activity for a process is determined at
different pR values a resuIt similar to that shown in Fig. 23.3 will often be found.
The shape of the curve is determined by the following factors:
1. At extreme (high and low) pR values the enzyme is denatured
2. At the optimum pR there is optimum structure and charge of both enzyme and
substrate
At the pR optimum for the process E + S ~ ES -+ E + P the formation of
ES complexes is at a maximum. If pR is varied from the optimum the charge dis-
tributions and consequentially the structures of enzyme and substrate are changed.
This lowers the affinity of E and S for each other so that fewer ES complexes
form thereby reducing the rate of product formation.
Reaction Rate and Temperature. Within certain limits, reaction rate increases
by a factor of 2 for each lOoe increase in temperature (QlO); however, enzyme
23 Enzyme Histochemistry I: General Considerations 309
pH
PHoptimum
denaturation also increases with temperature. The net substrate conversion therefore
depends on the balance between reaction rate and denaturation. At temperatures >
t max the denaturing is greatest; t max is not a constant value for the individual enzyme,
as it also depends on time. With longer incubation times denaturation becomes
more significant. The practical consequence of this is that for short incubation times
higher temperatures can be used but for longer incubation times temperatures should
be below the threshold for denaturation of the enzyme or be carefully monitored.
Reaction Rate and lonic Strength. Low ionic strength does not inhibit enzyme
activity and may even be activating. Certain enzymes need particular ions as co-
factors. Greater ionic strength may inhibit activity either by blocking the active site
or by denaturation.
Reaction Rate and Oxidants. Many enzymes contain thiol groups which are es-
sential for their activity. Oxidation to disulphide bonds therefore results in reduced
activity.
Hydrogen transferring
Nicotinamide-adenine dinucleotide NAD+ Hydrogen
Nicotinamide-adenine dinucleotide-phosphate NADP+
Flavine adenine-dinucleotide FAD
Flavine mononucleotide FMN
Coenzyme-Q CoQ
Lipoic acid Lip. Hydrogen + acyl groups
Group transferring
Adenosine triphosphate ATP Phosphoric acid and AMP
Uridine diphosphate UDP Uronic acid and glucose
Pyridoxal phosphate PLP Amino groups
Tetrahydrofolie acid FHY Formyl groups
Biotin Carboxyl groups
Coenzyme-A CoA Acyl groups
Thiamine pyrophosphate TPP C2-aldehyde groups
Bu-vitamin Alkyl groups
These are enzymes that differ in structure but catalyze the same process. They
can show different patterns of activity and this is important for their biological
function. One of the best examined isoenzyme systems is lactate dehydrogenase
(Sect25.4.1 ).
23.1.5 Proenzymes
A number of enzymes, including several of the proteolytic enzymes from the al-
imentary tract, are secreted in an inactive form called proenzymes. For example,
pepsin is secreted as pepsinogen. At low pH pepsinogen dissociates into several
peptides and is thereby transformed into active pepsin. In addition, pepsin itself
catalyzes the transformation of pepsinogen to pepsin (autocatalysis).
Enzymes are classified according to the reactions they catalyze. In 1955 the In-
ternational Union of Biochemistry (IUB) appointed the Commission on Enzymes
(Dixon and Webb, 1979).
According to IUB (1984) enzymes are classified in the following chief groups:
1. Oxidoreductases
2. Transferases
3. Hydrolases
4. Lyases
5. Isomerases
6. Ligases
Each main group is subdivided into sub- and sub-subgroups. The groups are
numbered according to a system that allows for further additions.
The numerical classification system can be summarized as follows:
1. Each enzyme is given an EC number consisting of four digits
2. The first digit indicates which main group the enzyme belongs to
312 A.P. Andersen, P.E. H!,!yer, H. Lyon, P. Prent!'!
Many enzymes are completely or partly localized to certain cell organelles, e.g.
acid phosphatase in lysosomes and succinate dehydrogenase in mitochondria (Sect.
31.11.1). Enzyme histochemical methods can therefore often be used as markers
for certain organelles. Furthermore, to a certain extent these methods also enable
cell identification as the relative occurrence of different organelles depends on the
function of the cell concerned. This approach is applied in haematology where
cytochemical methods allow distinction between cells of the myeloid and erythroid
series (Sect.31.11. 7).
Many pathological conditions give rise to histochemically demonstrable changes
in enzyme distribution and activity in cells. Such changes may be due to organelle
damage (e.g. increased permeability of lysosomes during hypoxia) and the effects
of the biochemicallesion itself on enzyme activity. Examples of the latter include:
the inhibitory effects of lead on ö-aminolaevulinate synthetase with resultant de-
crease in porphyrin synthesis and barbiturate induced production of the enzymes
comprising the cytochrome P450 detoxification system.
At an experimental level a considerable number of publications report attempts
to determine whether a cell has undergone a malignant change on the basis of
altered enzyme activity (e.g. many malignantly transformed cells show increased
succinate dehydrogenase activity).
While all the standard principles of histochemical reactions apply (Sect.1.2) the
following aspects may give rise to particular difficulty with enzyme reactions:
1. Immobilization of the enzyme in its original location
2. Deposition of the final reaction product corresponding to the enzyme under
investigation
3. Adequate preservation of cell morphology
Where quantitation is desired a further requirement is that no activity should
be lost during processing (Sect.28.1).
A very fine balance between fixation and retention of activity is necessary since
many enzymes diffuse easily and nearly all oxidoreductases and several hydrolases
are inhibited by even brief exposure to aldehydes.
The first steps in achieving reproducible results for the demonstration of any
enzyme activity should be the rapid freezing of the fresh, unfixed tissue followed
23 Enzyme Histochemistry I: General Considerations 313
Protection of Tissue during Incubation. Some enzymes are tightly bound to mem-
branes or other cell components (bound enzymes) while others are more or less
freely diffusible (soluble enzymes). The latter may be found free in the cytoplasm,
loosely bound to membranes, or enclosed in organelles. Membrane enclosed en-
zymes are not very diffusible under optimum conditions, but membranes may be
damaged during processing and incubation.
Diffusion of the enzyme can be avoided in three different ways:
1. Preceding fixation of the section. Clearly this results in varying degrees of
enzyme inhibition
2. Use of certain semipermeable films, usually dialysis membranes (McMillan,
1967; Meijer, 1980). The membrane is placed between the section and the
incubation medium. This technique has been successfully used for the demon-
stration of certain hydrolases (e.g. acid phosphatase, ß-glucuronidase, leucine
aminopeptidase, E600-resistant esterase) and a few oxidoreductases. Recently,
MiMly et al. (1988) have used this approach for the demonstration of carbonate
dehydratase (carbonic anhydrase). Agarose gel films containing the complete re-
action medium have also been used. The reaction is simply initiated by placing
the film over the tissue section (Pette and Brandau, 1962; Pette and Wunmer,
1980)
3. Addition of certain polymer compounds to the incubation medium. These in-
clude agarose, gelatin, polyethylene glycol, polyvinylpyrrolidone (PVP), poly-
peptides, carbohydrate polymers (e.g. Ficoll 400 and Dextran 10), ~d various
kinds of polyvinylalcohol (PVA). A quantitative study of the effects of dif-
ferent grades of PVA on the activities of certain enzymes was carried out by
Henderson et al. (1978). Reviews on the use of tissue protectants in enzyme
histochemistry have been published by Altman (1980b) and van Noorden and
Vogels (1989a)
Cofactors. As opposed to FMN and FAD, which are tightly bound prosthetic
groups, NAD+ and NADP+ are diffusible and must therefore be added to the
314 A.P. Andersen, P.E. H~yer, H. Lyon, P. Prentl'l
Activators. These are usually metal ions. In some cases they are bound to the
substrate, in other cases to the enzyme molecule. Certain buffer anions can form a
complex with the activating metal ion and block the activation.
Buffer. The pH chosen for the histochemical reaction is often a compromise be-
tween the pH-optimum for the enzymatic catalysis and the pH-optimum of the
capture reaction. As the enzymatic process nearly always changes the pH of the
incubation medium it is necessary to maintain pH with the help of a buffer. Clancy
(1987) has presented a computer program which enables the formulation of buffers
and media of known pH and ionic strength, and the calculation of the thermody-
namic acid dissociation constant of buffer substances.
Capture Reagent. Requirements for a good capture reagent or trapping agent are
that it should react easily and quickly with the products of the enzymatic reaction
and that this leads to precipitation of an amorphous coloured product at the location
of the enzyme under investigation. At too low a concentration of the capture reagent
coupling is difficult, while high concentrations frequently inhibit the enzyme.
The enzyme reaction is stopped by washing the section in distilled water, buffer, or
fixative. If desired, the material may then be subjected to postfixation (e.g. formol
calcium at 4°C for 15 min). The section is mounted in a hydrophilie mountant as
the end products are usually soluble in organie solvents.
This may be very difficult to achieve in tissue sections. Enzymes may be either sub-
strate or group specific. Thus, glucose-6-phosphatase specifically catalyzes the hy-
drolysis of glucose-6-phosphate while acid phosphatase catalyzes the hydrolysis of
most phosphate-monoesters including glucose-6-phosphate. Glucose-6-phosphatase
activity cannot therefore be equated with the presence of the enzyme glucose-6-
phosphatase. Identification must be attempted by:
1. Determination of activity with different substrates
2. The effect of specific inhibitors and activators
3. Biochemical analysis of the tissue, possibly after tissue fractionation
4. Immunohistochemical analysis with specific antibody directed at the enzyme in
question
24 Enzyme Histochemistry ll: Hydrolases
AB+H20~HA+B-OH
For the purposes of this account it is sufficient to note that the plasma membrane
and lysosomes are prominent locations for hydrolytic enzymes.
Histochemical studies aim to determine both the location and the activity of an
enzyme. In some cases, notably with lysosomal enzymes, assessments of the acces-
sibility of the enzyme in its normal cytologicallocation (membrane permeability)
and metabolic context are of interest. Under physiological conditions lysosome
membranes are relatively impermeable. When lysosomal enzymes are examined
histochemically there may be a latent period after the incubation has been initi-
ated before a drastic increase in lysosomal membrane permeability occurs. The
length of this period may be a direct expression of the physiological condition of
the membrane and the cello In formal biochemical terms this lag is often referred
to as latency. Investigations of latency may be carried out by comparing aseries
of sections incubated in media with or without added detergent (Fig. 24.1). An-
other approach has been described by Bitensky and Chayen (1977) in a study on
leucyl-ß-naphthylamidase in lysosomes in follicles of the guinea pig thyroid. The
substrate used for the reaction can only sparingly penetrate the intact lysosomal
membrane. The incubation medium is buffered to pH 5.0 and after a certain time
an inflexion point is noted This is taken to indicate that the acid medium ren-
ders the membrane more permeable for the substrate. The conditions pertaining to
the permeability of the lysosom al membranes of thyroid follicle cells have formed
the basis for developing a sensitive cytochemical bioassay for circulating levels of
thyrotrophin (Bitensky and Chayen, 1977; cf. also 28.8.11).
H. Lyon (Ed.)
TbeoJy and Strategy in Histochemistry
@ Springer Verlag 1991
318 H. Lyon, P.E. H!Ilyer, P. Prent!/l
t
Fig. 24.1. The relationship between hydrolase activity (extinetion) and incubation time.
a without detergent
b with detergent
Extinction (E), time (t).
OH
e~
©6 H ~ H 0
Although the method has been used sparingly for the demonstration of alka-
line phosphatase, reasonably good localization has been achieved with the calcium
salt of 5-bromoindoxyl phosphate in the presence of a ferrocyanide-ferricyanide
oxidation catalyst The precipitated 5,5' -dibromoindigo is microcrystalline.
Problems with the method include the rather slow spontaneous oxidation of
the leuco compound to indigo by action of atmospheric oxygen and the inhibitory
effect of the ferrocyanide-ferricyanide reagent on a number of enzymes.
The indoxyl ester methods are often used to demonstrate carboxyl esterase
activity and are very suitable for the quantitative demonstration of ß-galactosidase
(Lund-Hansen et al., 1984).
With all indoxyl ester methods the liberated indoxyl couples very effectively
with diazonium salts such as hexazotized Pararosanilin (simultaneous coupling
24.1.3). However, Fast Blue VB is the diazonium salt of choice (Lojda et aZ.,
1979, p.64). It yields the highest amount of azoindoxyl dye, allows precise local-
ization, and does not decompose even after prolonged incubation. Moreover, the
coupling velocity of the azoindoxyl procedure using Fast Blue VB is higher than
that obtained in the simultaneous azocoupling methods.
The indoxyl-tetrazolium salt method for the localization of alkaline phosphatase
activity was first described by McGadey (1970). In this method the substrate 5-
bromo-4-chloro-3-indolyl-phosphate is hydrolyzed at alkaline pH with a concomi-
tant release of hydrogen which can reduce tetrazolium salts. Tetranitro BT is the
tetrazolium salt of choice, as it yields an amorphous or microcrystalline formazan of
relatively high stability (Gossrau, 1978). The method has recently been optimized
by van Noorden and Jonges (1987) who developed a specific and valid quantita-
tive proeedure. They showed that addition of polyvinyl alcohol to the incubation
medium greatly improved the localization of the final reaction product in cryostat
seetions. The electron transfer from the substrate to tetranitro BT was significantly
enhanced by I-methoxyphenazine methosulphate which seems to indicate that elee-
trons may be transferred by endogenous electron transport systems in the absence
of I-methoxyphenazine methosulphate. Addition of azide to the incubation medium
enhaneed the formazan production considerably by blocking the interference from
oxygen. The specifie reaction obeyed the Beer-Lambert law. A KM of 0.05 mmol/l
for aqueous media and a KM of 0.55 mmol/l for polyvinyl alcohol-containing media
were obtained. These values indicate that the indoxyl-tetranitro BT method is con-
siderably more sensitive than any metal salt or diazonium salt method developed
so far.
These methods rely on the primary reaetion produet being sufficiently insoluble to
remain at its primary loeation without diffusion during both the initial incubation
and the subsequent performance of the visualizing reaction.
All post-coupling methods use a diazonium salt, and the final produets are
therefore azo dyes. These methods are particularly appropriate when long ineu-
320 H. Lyon, P.E. H~yer, P. Prent~
bation times are necessary, and when the pH optimum of the enzyme is in the
acid range. As distinct from the simultaneous coupling methods (Sect.24.1.3), a
compromise between the pH optima for the enzyme and forthe coupling reaction
between diazonium salt and naphthol or naphthylamine is not necessary. Similarly,
the enzyme reaction and the coupling reaction can both be performed at their re-
spective optimum temperatures (e.g. 37°C and 4°C). Finally, any inhibitory effect
of the diazonium salt on enzyme activity is avoided.
Despite these advantages and the fact that they are suitable for quantitative
investigations (end-point measurements, cf. Sect.28.8.7), post-coupling methods
are seldom used. The main reason for this is that it is necessary to use substituted
naphthol compounds to achieve reasonable localization. These compounds have
such a pronounced affinity for tissue proteins that coupling is inhibited and this
may give rise to false negative reactions. In some cases, however, the approach
may still offer significant advantages for particular enzymes (e.g. demonstration of
N-acetyl-ß-glucosaminidase, Robertson, 1980).
Diffusion of the primary reaction product before it reacts with the diazonium
salt is an important problem that may prevent localization of the enzyme activity.
Key factors inftuencing the final result are:
1. Specificity of substrate
2. Rate of substrate hydrolysis (naphthol compound)
3. Diffusion coefficient of the primary product (the naphthol)
4. Rate of the coupling reaction
5. Stability of the final reaction product
Simultaneous coupling methods are always a compromise between these fac-
tors. All diazonium salts lead to a greater or lesser degree of inhibition of enzyme
activity. Unfortunately, a relatively high concentration of diazonium salt is neces-
sary to achieve an effective coupling.
24 Enzyme Histochemistry II: Hydrolases 321
oor
00
Naphthol-AS
OH
CO-NH-@
These were the earliest methods developed for demonstrating enzyme activity. They
were introduced for the demonstration of alkaline phosphatase by Gomori (1939)
and Takamatsu (1939). The substrate is sodium ß-glycerophosphate or another
phosphomonoester and the incubation solution contains a high concentration of
calcium ions which act as the capture reagent. Pb2+ , Cu2+, and C02+ are alternative
capture reagents. Magnesium ions are added as an activator. The enzyme catalyzed
reaction is:
ß-glycerophosphate -+ ß-glycerol + phosphate
and the capture reaction with Ca2+ is:
3Ca2+ + 2P04 3- -+ Ca3(P04h
322 H. Lyon, P.E. H~yer, P. Prent~
The reactions taking place in the four methods described have been summarized
below in order to clearly indicate the number of steps before the final, insoluble,
coloured product is formed. (Precipitates underlined).
Indoxyl ester method:
AB -+ A + B; primary reaction product insoluble and stained
Simultaneous coupling method:
AB -+ A + B; the primary reaction product is soluble
B + C -+ BC; and unstained; the secondary reaction
product is insoluble and coloured
Post-coupling method:
AB -+ A +B; the primary reaction product is
B + D -+ BD; insoluble and colourless; the secondary reaction product is
insoluble and coloured
Metal salt method:
AB -+ A + B; the primary reaction product is soluble
B + C -+ BC; and colourless; the secondary reaction
BC + D -+ BD + C; product is insoluble, but colourless;
the tertiary reaction product is
insoluble and coloured
The choice of technique depends on the enzyme to be examined and on whether
observation is to be by light or electron microscopy. If there are several possibilities,
quality of localization is often the dominant influence in making the final choice.
Of the four approaches, it is only with the meta! salt method that natural sub-
strates can be used.
24 Enzyme Histochemistry II: Hydrolases 323
24.2 Pretreatment
Many hydrolases are diffusible and precise localization therefote requires the use
of semipermeable membranes, colloid stabilizers, or fixation (see Sect.23.2.2). If
fixation is chosen this should take place under carefully controlled conditions since
enzymes are often either denatured or their substrate requirements altered or ren-
dered less specific.
The ability of the different hydrolases to tolerate fixation is highly variable and
an individual approach developed for each enzyme should be closely adhered to.
In no circumstances should the recommended fixation time be exceeded.
24.3 Incubation
24.3.1 Substrate
Regardless of whether this is a diazonium salt or a metal salt the following re-
quirements apply:
1. Ready solubility with a high diffusion coefficient
2. Easy and rapid coupling under the conditions of the reaction (pH, temperature,
etc.)
3. Maximum substantivity of the reaction product formed
4. Preferably colourless
5. Unaltered by light
6. Stable under the various reaction conditions imposed (pH and temperature)
7. Minimal inhibitory effect on the enzyme catalyzed reaction
324 H. Lyon, PE. HfIlyer, P. Prent(ll
Metal ions are often necessary for, or increase, the activity of certain enzymes. For
ea
example, ATPase associated with myosin is activated by z+ and the cel1 mem-
brane "sodium pump" by Na+ and K+ . In some cases a meta! ions can be exchanged
for another provided the charge is unchanged. For instance, in the histochemical
demonstration of alkaline phosphatase both Mg2+ and Mn2+ are activators, while
the natural cofactor is Zn2+ . It is worth noting that many of the stabilized diazonium
salts contain Zn2+, even though this is in suboptimal amounts.
24.3.4 Inhibitors
Metal ions can act as inhibitors (e.g. Bez+ for alkaline phosphatase). Chemical
compounds which block -SH, -NHz, or -COOH groups frequently act as in-
hibitors, probably because these groups are often part of the active sites of the
enzymes.
Where possible specific inhibitors are used but, failing this, heat inactivated sec-
tions. A further control is the omission of substrate.
24.5 Quantitation
In theory this can be achieved by determining Amax for the final reaction product
and measuring the mean integrated extinction (MIE) at this wavelength.
Clearly prolonged chemical fixation (e.g. 3-4 hours in formalin) would com-
pletely invalidate quantitative results, as would storage of the sections in gum
acacia sucrose (cf. Sect.13.6 and Chap.28).
24 Enzyme Histochemistry II: Hydrolases 325
Included in this group of enzymes are carboxylic ester hydrolases, EC 3.1.1, and
phosphoric monoester hydrolases, phosphatases, EC 3.1.3.
It must, however, be stressed that the majority of the inhibitors are extremely
toxic. The use of natural substrates together with metal ions as the capture reagent
raises some further possibilities for distinguishing between the different esterases.
The term cholinesterases has unfortunately been used by histochemists to cover
both acetylcholinesterases, EC 3.1.1.7, or true cholinesterases and cholinesterases,
EC 3.1.1.8, or pseudocholinesterases. Both types are sensitive to organophospho-
rus compounds and are also inhibited by physostigmine (eserine) of which the
latter does not inhibit the non-specific esterases. Both acetylcholinesterase and
cholinesterase can be selectively demonstrated by use of acetylthiocholine which
is not attacked by other carboxylesterases.
Acetylcholinesterase activity decreases with the chain length of the acyl group
of the substrate and has the natural substrate acetylcholine. The enzyme can often be
preferentially demonstrated by use of the histochemical substrate acetylthiocholine,
but only at high substrate concentration which inhibits cholinesterase.
Cholinesterases show increased activity with increased chain length of the sub-
strate acyl group, and, as stated above, are also inhibited tQ some extent when the
concentration of acetylcholine or acetylthiocholine is high.
Lipases are able to attack long-chain (n more than 8) fatty acid esters. Most
esterases show little activity towards long chain substrates. Bile salts are, as-under
in vivo conditions, required to disperse the lipase substrate (fats) so that it is
accessible to the enzyme. In the absence of bile salts true lipases, or pancreatic-like
lipases, show little or no activity. Severallipases are, however, known ("lipoprotein
lipases") which hydrolyze long-chain esters in the absence of emulsifiers and many
non-specific esterases attack long-chain fatty esters, although they do so more
slowly than short-chain esters. Formerly, lipase activity was demonstrated using
sorbitan esters ("Tween") but this method shows low selectivity and has been
superceded by the naphthol-AS-nonanate-taurocholate method of Abe et al. (1964).
The latter method permits selective demonstration of pancreatic-like lipases. As
yet there are no published methods for the selective demonstration of lipoprotein
lipases.
Proteases generally hydrolyze peptide bonds but several proteases also show
significant activity against ester links. In fact, bromoindoxyl acetate and naphthol-
AS-ß-chloropropionate are useful substrates for the demonstration of cathepsin C
and for chymottypsin, as are naphthol-AS phenyl acetate, butyrate, and propionate.
When demonstrating esterases, it may therefore often be wise to bear in mind the
possibility that some, most, or all the enzyme activity may be due to a protease.
In this connection it should be noted that a particular aryl esterase, the naph-
thol AS-D chloroacetate esterase, shows several similarities with a-chymottypsin
(secreted by pancreatic exocrine tissue). Naphthol-AS-D chloroacetate esterase is
found in neutrophils and myelocytes as weIl as in mast cells. It is also noteworthy
that the enzyme activity is preserved after both formaldehyde and methanol fixation
and following paraffin embedding (see also Sect.3l.11.7).
ions as activator. Ca2+ are used as the trapping agent followed by substitution with
Co2+ and visualization with sulphide as described in Sect24.1.4. The reaction
is preferably carried out on unfixed cryostat sections. The pH optimum of the
enzyme is between 7.5 and 8.5. The reaction can with advantage be carried out
at pH 8.3 minimizing activity due to acid phosphatase while alkaline phosphatase
activity is inhibited by adding L-p-bromotetramisole to the incubation medium.
5'-Nucleotidase is itself inhibited by Zn2+ and Ni2+ ions.
Addition of PVA to the incubation medium results in better preservation of
tissue morphology (Henderson et al., 1980; Frederiks and Marx, 1988). A disad-
vantage with PVA is that total 5'-nucleotidase activity is significantly lower than
the activity in unfixed sections incubated with an aqueous medium (Frederiks and
Marx, 1988). The authors suggest that the activity of the enzyme is kept in the
latent state by PVA.
The enzyme is bound to plasma membranes. An increased activity is observed
in synoviocytes in rheumatoid arthritis (Henderson et al., 1980).
Natural substrates are glycogen and maltose. The enzyme is thus able to hy-
drolyze (a-l,4) and (a-l,6) glycoside linkages and therefore completely hydrolyzes
glycogen to glucose. Hydrolysis of the a-l,6 linkage is rate-limiting.
Deficiency of the enzyme is present in the storage diseases type 11 glycogenosis
or Pompe's disease.
The nomencIature of this group of enzymes is confusing and has probably not yet
been finally settled Proteases are divided into peptidases and proteinases.
24 Enzyme Histochemistty 11: Hydrolases 331
Peptidases (exopeptidases, Table 24.2) only attack peptide bonds at or near the
ends of peptide chains. Subclassification and designation is made according to their
specificity rather than their mechanism of action (McDonald, 1985).
Aminopeptidases 3.4.11 single amino acids from the N-terminus of the peptide chain
Dipeptidases 3.4.13 dipeptides
Dipeptidyl peptidases 3.4.14 dipeptide units from the N-terminus
Peptidyl peptidases 3.4.15 dipeptide units from the C-terminus
Under this heading only enzymes acting on diphosphate bonds in compounds such
as di- and triphosphates (EC 3.6.1) will be mentioned.
;r:
~
ß
'"d
i:I1
~
~
:-0
i
24 Enzyme Histochemistty ll: Hydrolases 335
In general, the ATPases are very sensitive to fixation and fresh frozen cryostat
sections should therefore be used for histochemical demonstration. As ATP is the
only substrate, metal salt methods must be used.
Ca2+ -Dependent ATPase, EC 3.6.1.3. In the presence of Ca2+ this enzyme gives
rise to Ca3(P04h that is only sparingly soluble at alkaline pH. Calcium is ex-
changed with cobalt by treatment with a solution of CO(N03h. For light microscopy
the section is then treated with (~hS which leads to the formation of black COS.
NAD
Formazan
(NNAAgp~~DHH)
SUbstrate]
(NADP) ] [
DH [
NADH
Product (NADPH) NitroBT
IL-___.._ PMS t
Fig. 25.2. Demonstration of NAD or NADP dependent dehydrogenases with tetrazolium salts as
the capture reagent.
Acceptors
Blocking or Redox potential Intermediate Histochemical,
inhibiting (mY at pH 7, 25°C) redox potential
compounds (mY' at pH 7, 25°C)
malonate
SDH·FAD
W-
I
·220 PMS ,
I L etc .
K
------'
CoQ
BPST
TNBT
_ _ _ _ _ _ _ _ _-----...J NBT(50)
cytochrome b
antimycin A ..
cytochrome C1 - - - - - - - - - - ,
220
cytochrome c NT (170)
~ 250
cytochrome a _ _ _ _ _ _ _ _ _--..J
cyanide
azide
J~----<
...
• 290
cytochrome a3
carbon monoxide ] - 820
N2
•
O2
Fig. 25.1. The respiratory chain.
SDH-FAD succinate dehydrogenase-flavine adenine dinucleotide
CoQ coenzyme Q or ubiquinone
PMS phenazine methosulphate
BPST 2-(2-benzothiazolyl)-3-(4-phthalhydrazidyl)-5-styryl-2H-tetrawlium
TNB T tetranitro blue tetrazolium
NBT nitro blue tetrawlium
NT neotetrawlium
T
Fig. 25.3. "Perspex-chamber" used for the demonstration of enzyme activity.
S = cover slip T = tissue section
P = ring of Perspex G = slide
In practice, both monotetrazolium salts (e.g. BPST) and ditetrazolium salts (e.g.
NBT, TNBT, NT) are used. In the latter compounds the two tetrazolium rings are
linked at a position corresponding to the two quatemary nitrogen atoms:
The tetrazolium salts used most frequently in current practice are shown in
Table 25.1.
342 PE. HfiSyer, H. Lyon
Table 25.1. Survey of the tetrazoliwn salts commonly used in enzyme histochemistry
=
MW molecular weight; SN =systematic narne; E' =redox potential (mv at pH 7, 25°C; TN =
trivial narne
N-N/\J§J
0
~
CH =CH-ef
\.
I 0
CI
N=N©O~
o
MW: 501.96; SN: 2-(2-benzothiazolyl)-3-(4-phthalhydrazidyl)-5-styryl-2H"tetrazoliwn chloride;
TN:BPST
lei
2 CI
~O, o~
.i-V
02NJc5V-I~
~N=N 0
( S \ N02
0 N=l\::::lr
2CI-
CHJO OCH J
MW: 907.63; SN: 3,3'-dimethoxy-4,4'-biphenylene)-bis-(2,5-p-nitrophenyl-2H-tetrazoliwn chlo-
=
ride); TN: TNBT tetranitro blue tetrazoliwn
25 Enzyme Histochemistry ill: Oxidoreductases 343
CI
H R2 R2 H
\ / \ I
N-N N-N
5 5
R -/( 3 3 _}- R
N-N-R-R-N-N
1). Most tetrazolium salts are sufficiently soluble in water to allow maximum reac-
tion rate and precise localization of formazans. They are colourless to pale yellow.
Slow decomposition always occurs on exposure to light, so these compounds should
always be stored in the dark. Moreover, according to van Noorden and Tas (1982a),
incubation in darkness is aprerequisite as tetrazolium salts are photochemically re-
duced whether or not an exogenous electron carrier is present. 1t is particularly
important to know the degree of purity as impurities may have a similar colour
to that of the monoformazans. There is a considerable difference in the quality
of reagents obtained from different suppliers. Buying from the right source, it is
always possible to acquire pure BPST, NBT, TNBT, and NT. If, for some reason,
impure tetrazolium salts must be used, it is necessary to purify to an acceptable
level judged by chromatography (see Sect.3.3.8).
344 P.E. H~yer, H. Lyon
2). BPST, TNBT, and NBT are easily reduced in the test tube, while the NT re-
duction rate is much lower even under anaerobic conditions (van Noorden, 1988b).
The relative order is as folIows:
BPST> TNBT 2: NBT » > NT
In tissue sections, similar reaction rates for glucose-6-phosphate dehydrogenase
activity were found with BPST, TNBT, NBT, and NT (Butcher and van Noorden,
1985). This seems to indicate that NT reduction in sections is mediated by tissue
components (see also Sect.25.1.3).
3). The redox potential should allow the tetrazolium salt to capture the reducing
equivalents from a position in the respiratory chain appropriate to the investi-
gation. In consequence, NT (high redox potential) is sometimes the only tetra-
zolium suitable for certain investigations, e.g. determination of so-called Type I
and Type 11 Hydrogen (Altman and Chayen, 1970; Chayen et al., 1973a). Type I
Hydrogen is believed to be involved in hydroxylating reactions (e.g. in steroido-
genesis), whereas Type 11 Hydrogen is believed to participate in other types of
biosynthetic reactions (e.g. fatty acid production). Moreover, for optimal activity
NT must be used as the final acceptor of reducing equivalents when investigating
NADPH-ferrihaemoprotein reductase (van Noorden and Butcher, 1986a).
4). Tetrazolium salts are potential enzyme poisons, and some may be carcinogenic.
It should be noted that they can be absorbed through the skin. The working concen-
tration should be as high as solubility allows without causing inhibition. In this way
as precise a localization as possible is ensured, and the reducing equivalents are
captured as effectively as possible. The inhibition caused by high concentrations
of tetrazolium salt may often be diminished by inclusion of a polymer stabilizer
such as PVA in the incubation J11edium (Fig. 25.4).
E
11
" - - - - - - - -_ _ _ _ _ _ _ _ _ [TSJ
Fig. 25.4. Theoretical curve showing the effect of tetrazolium salt concentration ([TS]) on enzyme
activity in a traditional aqueous medium (I) and in a medium containing an optimal amount of
PVA (polyvinyl alcohol)(II). Extinction (E).
25 Enzyme Histochemistry ill: Oxidoreductases 345
Although ditetrazolium salts possess two positive charges which are probably
responsible for their relatively pronounced protein substantivity (Sect.1.2.3), this
property does not seem to be the cause of the enzyme inhibition which sometimes
occurs.
5). The relatively large planar ditetrazolium salts with two positive charges and po-
lar groups often do not readily diffuse through intact membranes. To overcome this
problem, passage through mitochondrial membranes is facilitated by brief fixation
and the addition of DMSO. The use of a phosphate buffer, in place of others such
as glycylglycine, has the same effect. Normally, however, the membranes are pen-
etrated so quickly that the local intracellular concentration is not rate-determiIiing.
Formazans should:
1. be poorly soluble and show a high substantivity for proteins.
2. not be soluble in lipids
3. be bound as very small and homogeneous particles
4. have distinct colours for which the absorption maxima in tissue are known
5. be stable to light
1). Substantivity for tissue proteins should be as high as possible. First, the for-
mazan must almost instantly bind to proteins elose to the enzyme molecule if exact
localization is to be achieved Second, low substantivity will increase the tendency
to recrystallization and lead to formazan being removed when the incubation is
stopped in distilled water. Diformazans show a relatively high substantivity. The
large, planar molecules and hydrogen bonds are believed to be of importance in this
respect. In general, monoformazans (from mono- or ditetrazolium salts) show less
substantivity; thus BPST-formazan binding may be minimal in some cells. Under
some circumstances, notably in certain tumour cells and in necrotic areas, even the
diformazans can show decreased binding.
2). Solubility in lipids can lead to formazan diffusion from the site of formation
to lipid droplets where unsaturated fatty acids can cause a non-specific reduction
of tetrazolium salts. Of the tetrazolium salts in Table 25.1 only NT is significantly
soluble in lipids. The formazans of all four salts have a certain affinity for lipid-
water interfaces, and this may lead to accumulation in locations such as nuelear
membranes and the surface of lipid droplets. This tendency is most pronounced for
NT:
NT > > NBT > TNBT > BPST
The more lipid-soluble a formazan, the more pronounced is its tendency for both re-
crystallization and changes in its absorption maxima. In the case ofBPST-formazan
these problems can be completely avoided by chelating with Ni2+ or Co2+ (Altman
et al., 1979).
3). The sm aller the formazan particles the more precise the localization iso In liver
sections incubated to produce heavy formazan deposits, Altman (1976) found that
the formazans of TNBT, NBT, BPST, and NT were granular in shape and had
diameters of approximately 0.2, 0.5, 0.8, and 1.7 pm, respectively.
346 P.E. H~yer, H. Lyon
4). Tbe individual fonnazans each have characteristic absorption maxima in tissue.
BPST-fonnazan can be measured direcdy at its maximum (530 nm). However,
when NBT, TNBT, and NT (ditetrazolium salts) are used, it is necessary to take
into account the fonnation of mono- and difonnazans (see Sect25.1.4 and Chap.28).
5). Slow decomposition and recrystallization of fonnazans occur when the sections
are stored in light. They should therefore be kept in the dark and preferably be
examined shordy after mounting in a hydrophilic, non-extracting medium.
In conclusion, TNBT and NBT are generally preferred as they fill most of the
requirements cited regarding both the tetrazolium salts and the resultant fonnazans.
If the tissue is lipid-rich, TNBT should be used in place of NBT in order to achieve
as precise a localization as possible. As shown by Horobin (1982b) it is possible by
theoretical considerations of structural parameters to arrive at the same conclusions.
Tbe parameters of interest, which may be derived from the structural fonnulae of
the tetrazolium salts, are the conjugated bond numbers and Hansch 7f values of
molecular fragments (Horobin, 1982a, pp.237-246).
Seidler (1980), in a review on new nitro-monotetrazolium salts, concluded that
these salts are particularly suited for quantitative histochernistry due to the uni-
fonnity of their amorphous reaction products. Tbe substitution of nitro groups in
tetrazolium salts (and the corresponding fonnazans) gives advantages of increased
substantivity and reducibility along with irnproved localization and decreased sol-
ubility of the fonnazans in lipids. Of the monotetrazolium salts tested, 2,3-di(p-
nitrophenyl)-5-phenyl-2H-tetrawlium chloride (2,3-p-DNTTC) was recommended
for quantitative work as its fonnazan always appears in a blue fonn. Most of the
other tetrazolium salts tested gave red fonnazans in nonnal tissue whereas varying
amounts of blue fonnazans occurred in ischaemic lesions. Full details for the syn-
thesis of 2,3-p-DNTTC are given by Seidler (1980). Unfortunately, this tetrazolium
salt does not appear to be commercially available.
L
SDH
J+ Cu"
CU2Fe(CN)6, xH 20
(Hatchett's Brown)
25 Enzyme Histochemistry m: Oxidoreductases 347
[coenzyme]
Fig. 25.5. Theoretical curve showing the dependence of enzyme activity on coenzyme concentra-
tion ([coenzyme]).
Usually high coenzyme concentrations result in zero-order kinetics (I) but inhibition may be seen
with some enzyme-coenzyme pairs (II). Extinction (E).
348 P.E. Hlilyer, H. Lyon
Coenzyme. In quantitative work zero-order kinetics must be ensured (Fig. 25.5 and
Chap.28). This usually means that the concentration of coenzyme should be of the
order 1-2.5 mg/mI.
Occasionally, high concentrations of NAD+ or NADP+ can be inhibitory
(H~yer, 1988). NADH and NADPH can still diffuse in gel media and, as noted in
Sect.25.1.2 (Protection of tissue), this may result in the final product not having
the same distribution as the enzyme.
If the dehydrogenase to be examined is dependent on coenzyme Q in the cy-
tochemical system, it is important to determine whether the mitochondria have a
rate-determining low concentration of this coenzyme (Andersen and H~yer, 1974).
Henderson and Loveridge (1981), using a PVA technique, found Meldola Blue to
be of little value. Stellmach and Severin (1987) found that the amount of fonnazan
obtained after incubating living cells with Meldola Blue as the electron carrier
was greater than that obtained with a number of the other intennediate acceptors
including PMS, Methylene Blue, and menadione.
Butcher and Evans (1984) compared the effects of several intennediate electron
acceptors using incubation media containing polyvinyl alcohol. Menadione was the
least effective both in transferrlng reducing equivalents from the primary dehydro-
genase 10 the tetrazolium salt and in preventing diffusion. PMS, Methylene Blue,
and Thionin were an more efficient.
Considerable diffusion occurred when PMS was used, whereas diffusion was
minimal with either of the thiazin dyes. Meldola Blue had litt1e or no effect on
fonnazan deposition. Kugler (1982b) suggested that the inefficiency of Meldola
Blue in incubation media containing polyvinyl alcohol might be due to this dye
being almost completely bound to the colloid stabilizer.Butcher and Evans (1984)
agree with this view.
Van Noorden and Tas (1982a) compared the applicability of mPMS with PMS,
Meldola Blue, and menadione using TNBT and polyacrylamide films into which
either purified glucose-6-phosphate dehydrogenase or intact liver cells where in-
corporated. Neither intennediate compounds nor formazans diffuse in gel films of
this kind. Consequently, it was not necessary to include PVA during incubation.
In films containing purified enzyme, equimolar concentrations of PMS, mPMS,
and Meldola Blue enhanced the transfer of electrons from NADPH to TNBT to
a comparable degree. However, no enhancement of electron transfer was noted
when menadione was used. They concluded that when menadione is used for the
cytochemical assay of NADP+ - or NAD+ -dependent dehydrogenases in situ, it is
not the primary enzymatic activity that is demonstrated but one of the compounds
present in the cellular electron transport systems. This is explained by the fact
that no electron exchange takes place between NADPH and menadione, whereas
reduced menadione reduces TNBT (van Noorden and Tas, 1982a). The authors
therefore suggest that menadione may be useful for the detection of succinate de-
hydrogenase as FAD/FADH has a higher redox potential than NAD(P)+ INAD(P)H.
In films containing intact liver cells both PMS and mPMS enhanced fonnazan
production. The "nothing" dehydrogenase activity (Sect.25.1.3) was less than 10%
of the actual enzyme activity when mPMS was used. The background staining was
somewhat higher with PMS, probably due to non-specific fonnazan production and
a light pink staining of cells by PMS itself. With both menadione and Meldola Blue
the non-enzymatic fonnazan production and background staining of the cells was
quite high, and these substances were deemed "useless" in this context
The background staining with Meldola Blue is partly due to binding of the
intensely blue coloured compound to lipids and DNA (van Noorden and Tas,
1982b). Furthennore, mPMS is light-stable and photochemical reduction of tetra-
zolium salts is minimal. The authors therefore conclude that mPMS is the carrier of
choice when extramitochondrial dehydrogenases are assayed (van Noorden and Tas,
1981a; 1982b). In the investigation of mitochondrial dehydrogenases PMS should,
350 P.E. Hl'lyer, H. Lyon
Buffer. In principle the pH optimum must be detennined and used. The buffer
should have reasonable capacity at this pH. For example, it is unacceptable to use
phosphate buffer at pH 8. For further information on this and the ionic strength of
buffers, see 23.2.2 and Clancy (1987). Buffers may have a direct effect on some
enzymes or substrates. For example, Tris buffer can oxidize some steroid substrates
(Ferguson and MacPhee, 1975) and maleate in Tris-maleate buffer inhibits urate
oxidase (Angermüller, 1989). If a metal ion functions as an activator for the de-
hydrogenase a glycylglycine buffer may not be suitable due to its ability to form
complex bonds. In general the ionic strength should be kept as low as possible,
although it must be possible to keep pH constant throughout incubation. At pH > 9
spontaneous reduction of tetrazolium salts may occur. Both NAD+ and NADP+
are most stable at pH below 7, while the reduced versions are most stable at pH
above 7.
Incubation Time. For quantitative work this must be chosen so that the amount
of final reaction product due to the dehydrogenase under investigation varies lin-
early with time (Fig. 25.6). As shown by several authors (e.g. van Noorden and
Butcher, 1986a) both the test reaction (plus substrate) and the control reaction (mi-
nus substrate) sometimes show levelling off of the reaction rate from the very start
of the incubation. When, however, the control reaction is subtracted from the test
re action the resulting specific reaction is usually linear for a considerable period
of time. One exception from this general rule has been described by van Noorden
and Vogels (1989b).
Even in the linear zone, longer incubation times allow more extensive diffusion
of NADPH and reduced carrier substances.
Fig. 25.6. Theoretical curve showing the relationship between activity and incubation time. Ex-
tinction (E), incubation time (t).
25 Enzyme Histochemistry III: Oxidoreductases 351
The Atmosphere. When using NT either the terminal part of the mitochondrial
respiratory chain must be blocked (see Fig. 25.1) or a nitrogen atmosphere must
be used. In the latter case the medium must be saturated by bubbling through with
nitrogen.
Butcher (1978) has shown that reduction of NT in vitro only occurs under
strict anaerobic conditions. Apparently, oxygen reduction is energetically far more
favourable than NT reduction, both in the test tube and in non-carcinomatous
tissue. NT reduction in nitrogen gave the localization pattern of the NADPH de-
hydrogenase, both in the presence and absence of intermediate electron carriers.
This suggests that the transfer of reducing equivalents from the exogenous electron
carrier to NT proceeds via cellular electron transport systems (van Noorden and
Butcher, 1984).
In some cases it may be of diagnostic value to use aseries of different oxygen
tensions when using NT as the tetrazolium salt. Thus, the re action for glucose-6-
phosphate dehydrogenase in normal bronchial epithelial cells shows pronounced
inhibition with high oxygen tension, while the same reaction in tumour cells from
a bronchial carcinoma shows considerably less or no inhibition (Butcher, 1979).
Similarly, in studies of human stornach and colon (Ibrahim et al., 1983) and of
human breast tissue (Petersen et al., 1985) it was found that when incubation took
place in an oxygen atmosphere, glucose-6-phosphate dehydrogenase activity could
not be demonstrated in normal tissue whereas activity was retained in carcinoma-
tous tissue. The components in carcinomatous tissue, which are responsible for this
oxygen insensitivity, are not known at present.
Van Noorden (1988a) studied the direct effect of oxygen on the tetrazolium
salt reduction mediated by NAD(P)H and PMS. The conc1usions of this paper are
given below. The first five relate to spectrophotometry using cuvettes containing
cell and tissue free media with or without addition of pure enzyme, while the sixth
conc1usion was derived from work carried out on rat liver sections. The conc1usions
were:
1. Oxygen may interfere in the NBT and TNBT reduction, not only via flavine-
containing enzyme systems (inhibited by azide), but also direct1y via tetrazolinyl
radical intermediates. In the latter case oxygen inhibited competitively
2. In atmospheric air, 5 mmol/1 NBT or TNBT is sufficient to exc1ude the oxygen
effect via radical intermediates (cf. point 6)
3. Macromolecules (e.g. albumin, PVA, and tissue macromolecules) dirninish the
oxygen effect on NBT and TNBT reduction, probably by keeping the local
concentration of tetrazolium salt high
4. BPST reduction appeared to be unaffected by oxygen, probably due to for-
mazan production directly via divalent electron transfer and not via a radical
intermediate
5. The inhibitory effect of oxygen on the reduction of NT was not significantly
affected by increasing the concentration of tetrazolium salt
6. The maximum reaction rate and optimum localization of glucose-6-phosphate
dehydrogenase in rat liver sections were obtained by using 5 mmol/l TNBT or
NBT and 20% Sigma PVA. From the above it becomes apparent that when using
352 PE. H!6yer, H. Lyon
TNBT or NBT, any inhibition due to oxygen in atmospheric air in the absence
of PMS can be overcome by increasing the concentration of the tetrazolium sah
to 5 mmol/l and by adding macromolecules (e.g. PVA). If PMS is employed,
it is necessary to block oxygen interference via flavoproteins by adding azide
or cyanide (Butcher, 1983).
Temperature. Incubation is usually performed at 37°C as this is optimal for the
majority of anaerobic dehydrogenases. If the reaction rate is extremely high, it may,
however, be of advantage also to perform the incubation at lower temperature, e.g.
25°C. (See Sect.23.2.2).
Other Methods. In the final analysis, however, comparison of the results of enzyme
histochemical analysis with those obtained by other methods, such as biochemical
and immunocytochemical procedures, is always valuable.
Another valuable approach is the use of model systems (Horobin, 1982a, pp.
229-234). Different materials have been proposed as the matrix for experiments of
this kind (van Duijn, 1976; van Duijn and van der Ploeg, 1980), the most frequently
used being polyacrylamide films (van der Ploeg and van Duijn, 1964).
Usually soluble fractions of isolated cells or pure preparations of macromole-
cules have been incorporated in the films, but van Noorden and Tas (1981b) have
described a method for incorporating intact cells or pure enzymes in the polyacry-
lamide film. The films can be assayed biochemically as well as cytochemically.
With model films linearity may be measured between increase of absorbance on
the one hand and incubation time and film thickness on the other hand. Furthermore,
absorbance per thickness of film (per unit of time) and the amount of enzyme in-
corporated can be studied (Beer-Lambert's law, cf. Sect.3.3.2). The polyacrylamide
film technique is, however, not without problems. In the paper by van Noorden and
Tas (1981b) on glucose-6-phosphate dehydrogenase these are shown to include:
1. A certain degree of enzyme inactivation must be expected during polymeriza-
tion
2. Whereas NBT and TNBT can be used as final acceptors, the use of BPST and
NT does not lead to formation of any formazan, possibly due to free radical
residues, generated in the acrylamide during polymerization, reacting with the
intermediary free radicals arising during tetrazolium salt reduction. Apparently
the nitro groups in NBT and TNBT protect the intermediary radicals arising
from these tetrazolium salts against interference from the free radicals in the
polyacrylamide
354 P.E. Hl'lyer, H. Lyon
3. TNBT measurements are most suitably carried out at 535 nm although TNBT
is a ditetrazolium salt with an isobestic point (Sect.25.1.4) at 557 nm in tissue
sections
The above conclusions emphasize the importance of optimizing matrix model
systems. Other enzymes, apart from glucose-6-phosphate dehydrogenase, that have
been studied in polyacrylamide films include peroxidases (van der Ploeg and van
Duijn, 1964), alkaline phosphatase (van der Ploeg and van Duijn, 1968), acid phos-
phatase (Lojda et al., 1967) aspartate aminotransferase isoenzymes (papadimitriou
and Van Duijn, 1970), esterases (Cabrini et al., 1977), and acetylcholinesterase
(Andrä and van Duijn, 1985).
25.1.4 Quantitation
Many of the peroxidases are iron containing haemoproteins, which catalyze the
process:
25 Enzyme Histochemistry ill: Oxidoreductases 355
where in histochemistry DHz can be amines, phenols, or the so-called leuco fonn
of certain dyes.
Graham and Karnovsky (1966) introduced 3,3'-diaminobenzidine (DAß) as a
substrate for ultrastructural studies. This is now the substrate of choice due to
the insolubility in water and ethanol of the polymerized reaction product. A further
advantage is its electron density rendering DAß suitable for electron microscopical
studies. With added H202 as acceptor of reducing equivalents DAß is oxidized and
polymerized fonning indamine and/or phenazine coupling products (Seligman et
al., 1968). The precipitate may be amplified with OS04 fonning Osmium Black.
25.4.1 Dehydrogenases
LDH can withstand brief fixation with formaldehyde but normally cryostat
sections of freshly frozen tissue are preferred for histochemical demonstration. Van
Noorden and Vogels (1989b) determined the reaction rate of the enzyme using a
quantitative assay. They were able to show that the addition of polymer compounds
(Sect.23.2.2) to the incubation medium improves localization considerably . A
method for simultaneous quantitative histochemical assay of lactate and succinate
dehydrogenases in the same cell has been described by Stoward and Nakae (1988).
NADH Dehydrogenase, EC 1.6.5.3 and 1.6.99.3. The first reaction in the mito-
chondrial respiratory chain is the transfer of electrons from NADH to ubiquinone-
10 (coenzyme Q). This reaction is catalyzed by complex 1 (NADH dehydrogenase
(ubiquinone) (EC 1.6.5.3) acting via ftavoprotein (FMN) and iron-sulphur protein
(FeS)).
Complex 1 can be degraded by various treatments to different types of NADH
dehydrogenases which may use acceptors other than ubiquinone-lO. The Enzyme
358 P.E. H~yer, H. Lyon
Commission (IUB, 1984) recommends that these types should be called NADH
dehydrogenase (EC 1.6.5.3) without specification of an acceptor in brackets.
At present it is not clear which of the multiple NADH dehydrogenases are
demonstrated in cells with cytochemical methods, and for this reason several syn-
onyms have been widely used in the histochemical literature including NADH-
diaphorase, NADH-ferricyanide oxidoreductase, NADH tetrazolium salt reductase,
and NADH oxidoreductase. These terms are now considered inappropriate and
NADH-dehydrogenase has been used throughout this text as a collective term for
the oxidoreductase systems which catalyze the oxidation of NADH in cytochemical
assays.
A substantial amount of NADH dehydrogenase activity is retained after brief
formaldehyde fixation. Nonetheless, the use of cryostat sections of freshly frozen
tissue is strongly recommended.
NADH may reduce tetrazolium salt non-enzymatically. This non-specific re-
action becomes more pronounced above pH 8 and may therefore be reduced by
lowering pH to 7.2-7.5.
It should be noted that when measuring the activity of an NAD+ -dependent
dehydrogenase, a test, as to whether NADH dehydrogenase activity is rate-limiting
for the overall reaction, should be made by performing the reaction at the pH
optimum of the NAD+ -dependent enzyme (cf. Sect.25.1.2, Buffer). Furthermore,
when tetrazolium salts with low redox potentials (e.g. NBT or TNBT) are employed,
it may be necessary to use a relatively low concentration of NADH (about 0.5
mmol/l) in order to avoid too high non-specific formazan production in the medium.
With NT an optimum concentration of NADH may, however, be used.
25.4.2 Peroxidases
Peroxidases catalyze the oxidation of many different compounds (e.g. fatty acids
and amino acids) using hydrogen peroxide as the hydrogen acceptor.
According to the Enzyme Commission (IUB, 1984) enzymes belonging to
EC 1.11 act on hydrogen peroxide as acceptor. There is only one sub-subclass,
EC 1.11.1, the peroxidases, that comprise several different enzymes of both animal
and plant origin. From a histochemical viewpoint it may be convenient to subdivide
peroxidases into endogenous and exogenous.
Exogenous Peroxidases. This term covers peroxidases used as probe tracer sub-
stances. Peroxidase extracted from horseradish (HRP) is of particular interest. HRP
is widely used as a tracer in permeability investigations (Sect.27.2.3) and in im-
munohistochemistry (Sect.26.2.2). Histochemical demonstration using benzidine
and its derivatives is described in Sects.26.2.2, 27.2.3, and 28.8.8. It should be
noted that cytochrome-c oxidase, endogenous catalase, and haemoglobin may cause
false positive reactions.
The peroxidase-like activity shown by haemoglobin may be suppressed by
preliminary blocking with either H202 or by treatment with periodic acid (pearse,
1980, p.235).
Cytochrome-c oxidase can be inactivated by fixation with formaldehyde, which
HRP is relatively resistant to or inhibited by azide or cyanide. Catalase activity
localized to peroxisomes can be suppressed if the hydrogen peroxide concentration
of the incubation medium is kept under 3 x 10-3 mol/l. If quantitative determination
of exogenous peroxidase (HRP) activity is to be made in probe assays, the plateau
absorbance principle may be used (Sect.28.8.8).
In the majority of qualitative studies on "peroxidase" either glutaraldehyde or
formaldehyde fixation has been used even though fixation must cause more or less
pronounced inactivation.
In attempts to differentiate between the different peroxidases several different
factors have been investigated, including: different fixatives, the influence of fix-
ation temperature and time, concentrations of DAB and H202, and pH. See for
instance Fahimi (1975), LeHir et al. (1979), and van Bogaert et al. (1980).
If quantitative demonstration of the activity of a specific endogenous peroxidase
is intended, unfixed cryostat sections and optimal kinetic conditions for the enzyme
in question must be used (Chap.28). Only a few endogenous peroxidases have been
extensively studied cytochemically. An example, iodide peroxidase, is given below.
was achieved after one day (Ealey et al., 1984). Perrild et al. (1988) investigated the
acute effects of low doses of TSH on the metabolism of guinea-pig thyroid segments
maintained in non-proliferative organ culture. A sustained rise in peroxidase activity
reaching 129% over control was observed after 30 min.
25.4.3 Oxidases
the corresponding imino acids. The imino acid is then spontaneously hydrolyzed to
the keto acid and NH3. The enzyme is found in liver cells and cells of the proximal
tubules in the kidney. Pipecolic acid or thiazolidine-2-carboxylic acid can be used
as substrates for the cytochemical localization of the enzyme. Inhibitors are kojic
acid and sodium benzoate (AngermüIler, 1989).
26.1 Definition
26.1.2 Antibodies
H. Lyon (Ed.}
Theory and Strategy in Histochemistty
© Springer Verlag 1991
368 P.P. Clausen, M. Mßller, B. van Deurs, O.W. Petersen
ANTIGEN
SELECTION
Produclion 01 antibody
Fig. 26.1. Scheme showing three lymphocytes (I, 2, and n) of which (2) is "triggered" to produce
antibodies. The triggered lymphocytes have receptors for the specific antigen on their surface.
Protein
Albumin
e Electrophoretic mobility@
Fig. 26.2. Serum electrophoresis showing separation of serum albwnin and different classes of
globulin (a . ß. ')').
Heavyehaln
Fab
Disulfide
bond
Fe
of light ehains are found: K and A. Aeeording to the five immunoglobulin classes,
the heavy ehains are named / (IgG), a (IgA), 6 (IgD), e (IgE), and jJ, (IgM).
The enzymes papain and pepsin split immunoglobulin molecules into several
eomponents with different biological properties. From Fig. 26.4 it ean be seen that
digestion of IgG with papain results in the formation of two fragments termed Fab
(fragment antigen binding) and one termed Fe (fragment erystalline). The binding
between antigen and antibody takes plaee at one end of the Fab fragment, whereas
the Fe fragment has other biological properties.
370 P.P. Clausen, M. M!1!lIer, B. van Deurs, O.W. Petersen
~: I-- F ab-l
Fig. 26.4. Diagram illustrating the etIects of the proteolytic enzymes pepsin and papain on the
IgG molecule. Papain will produce two Fab fragments and pepsin will produce a single F(ab)z
fragment.
Within both light and heavy chains there are different regions called variable and
constant regions. The variable region contains the amino acid sequence and steric
confonnation responsible for the specific binding of the antibody to the antigen.
Epitope
Epitope
Fig. 26.S. Diagram illustrating the sttucture of a folded polypeptide. Five different epitopes are
shown.
amino acids in the primary sequenee brought together by the sterie eonformation
of the molecule (Fig. 26.5).
The larger the antigenie site, the stronger is the binding between antigen and an-
tibody. The binding is non-eovalent, based on eleetrostatie forces, hydrogen bonds,
van der Waal forces, and hydrophobie interaction. The reaction between antigen
and antibody is reversible: AB + AG ~ ABAG, and the reaction ean be deseribed
by the law of mass action:
K
AB+AG~ABAG
If K is large, the equilibrium is displaeed to the right and the antibodies bind
strongly to the antigen and the antibody displays a high affinity. In praetiee we
deal with multivalent antibodies, and in this ease the term avidity is used instead
of affinity of the antibody. In practieal immunohistochemieal work it is important
to be aware of the fact that dilution of the antibody, for instance during prolonged
washing between ineubation steps, will shift the equilibrium to the left and thereby
eause a weaker staining.
The classical way of producing antibodies is to inject the purified antigen subcu-
taneously into an animal. Most often rabbits, swine, goats, sheep, or guinea pigs
are used. The animals are immunized several times with intervals of 2 or 3 weeks
in order to obtain a high level of antibodies. When the serum from an immunized
animal is isolated, it contains a mixture of different antibody molecules produced
by different plasma cell clones, each of which reacts with only one antigenic de-
terminant (Fig. 26.6).
An antiserum comprising multiple different antibodies towards the same anti-
gen is referred to as a polyclonal antiserum. The emde antiserum is characterized
372 P.P. Clausen. M. M~ller. B. van Deurs. O.W. Petersen
IMMUNE RESPONSE
IN RABBIT
tnJection of an antigen
bearing fourdeterminants
----Q 1.."."" tu
2 3 4
~(I
Antibodles I. 2. 3. 4.
enter the blood stream
by the presence of both non-specific immunoglobulins of natural origin and the ar-
tificially induced-specific immunoglobulins. For immunohistochemical staining the
IgG-containing immunoglobulin fraction is often isolated. Even the purified im-
munoglobulin fraction will be a mixture of specific and non-specific immunoglob-
ulins.
Further purification using antigen-affinity chromatography results in the isola-
tion of specific antibodies. Theoretically, the use of affinity purified antibodies in
immunohistochemistry should yield lower background staining, however, this is
not always the case as some of the specific antibodies may be removed during
purification.
26 lnununohistochemistry 373
IMMUNE RESPONSE
IN MOUSE
2 3 4
@@@@
Nonsecretory mouse Antigen reactive Iymphocytes
myeloma culture
Fusion
Hybrid cells
Briefty, mice are immunized with an antigen (not necessarily pure) and im-
mune sensitized lymphocytes harvested from the spleen. These cells are fused with
myeloma cells that are unable to secrete immunoglobulins themselves. Following
this, only fused (hybrid) cells can grow in the specially fonnulated culture medium
and some of these secrete immunoglobulins. The specificity of the antibody is de-
termined by the original activated lymphocyte that took part in the fusion. The
hybridoma cells are cloned and screened for antibody production. After test reac-
tions with the purified antigen, the cell clones of interest are further expanded. As
antibodies produced this way originate from a single cell clone, they are referred
to as monoclonal antibodies.
Having raised and isolated an antiserum, it is important to test its specificity, i.e.
the degree to which it reacts with antigens other than the one used for immuniza-
tion. Lack of specificity may result from presence of contaminants in the original
immunizing antigen preparation. These may give rise to their own specific antibody
response. It is therefore important to use highly purified antigens as immunizing
material when polyclonal antisera are being made. Traces of antibodies to contam-
inant antigens can be removed by absorption procedures.
Another source of non-specific immunoreactivity is antibody present in the
serum of the animal before immunization. This problem may be overcome by
using affihity purified antibodies. Finally, genuine serological cross-reactivity may
also cause non-specific reactions. If amino acid or sugar sequences are shared by
both the test and other antigens, antibodies directed at such sequences bind to both
the specific and non-specific antigens. This sort of false positive reaction cannot
be circumvented.
Testing for specificity is primarily performed by means of gel techniques such
as crossed immunoelectrophoresis, immunoblotting (Western blots), and immuno-
precipitation. In addition, radioimmunoassay (RIA) and the ELISA (enzyme-linked
immunosorbent assay) technique are often used.
Even if the antiserum appears monospecific by these methods, its specificity
still has to be confirmed at the immunocytochemicallevel before interpretation is
valid. This is necessary because the specificity of the antiserum applied to sections
may vary from that obtained in test tube reactions. Reasons for this include fixation
and other aspects of tissue processing as well as other factors that may influence
antigen conformation. It is helpful to test antibodies in a simple model system
that reproduces many of these features while allowing many different antigens
to be tested at the same time. Droplets of the antigens to be tested are applied
to filter paper or nitrocellulose strips and immobilized by vapour fixation. The
immunoreaction can then be perfonned in circumstances similar to the proposed
tissue reaction and potential cross-reactions with a very large panel of different
antigens evaluated.
26 Immunohistochemistry 375
Several enzymes have been used for labelling antibodies (e.g. horseradish peroxi-
dase (HRP), glucose oxidase, alkaline phosphatase, and ß-galactosidase). Among
these horseradish peroxidase (HRP) has been most widely used. It is a very sta-
ble enzyme, the enzymatic activity of which is hardly influenced by the coupling
procedure. HRP has a molecular weight of 40 kDa and is coupled chemically to
antibodies using glutaraldehyde or periodate.
The enzyme catalyzes the transformation of hydrogen peroxide to water if a
hydrogen donor is available; 3,3' -diaminobenzidine tetrahydrochloride (DAB) is
often used as hydrogen donor. During the catalytic process, DAB is transformed
to a water-insoluble, brown product. As the reaction product is osmiophilic, DAB-
labelled antibodies can be used for both light and electron microscopical investi-
gations (Sect.27.3).
In the peroxidase-an ti-peroxidase technique (pAP-technique, see Sect.26.3.3)
three PAP molecules are bound to two IgG molecules in an immune complex. This
approach increases the efficiency of the immunohistochemical detection system.
Ferritin is an iron containing protein with a molecular weight of 500 kDa. Due to the
high content of iron (each molecule possesses up to 4000 iron atoms) this molecule
is electron dense and thereby visible in the electron microscope. (Sects.27.2.3,
27.3). The moleeule is 10 nm in diameter. Using antibodies coupled to ferritin
it is possible to quantify the number of antigens revealed in the section, as each
"dot" seen in the electron microscope corresponds to one antigen molecule. A
major disadvantage with ferritin is its high molecular weight. This leads to serious
problems for antibody penetration into seetions.
26 Irnmunohistochemistry 377
This section covers the theoretical aspects of staining methods using antibodies and
conjugates on tissue seetions.
}==4
3
radish peroxidase (Fig. 26.10). Alkaline phosphatase complexes have also been used
(APAAP-complexes).
In recent years, new marker systems have been developed which have increased the
sensitivity of the immunohistochemical procedures. Biotin is a small water-soluble
vitamin (molecular weight 0.244 kDa). This molecule binds strongly to avidin, a
large glycoprotein (molecular weight 67 kDa) which is abundant in egg white. Each
avidin molecule possesses four binding sites for biotin.
Antibodies conjugated to biotin in combination with avidin coupled to fluo-
rochromes, enzymes, or colloida! gold represent a very sensitive detection system
(Fig. 26.11).
380 P.P. Clausen, M. M~l1er, B. van Deurs, O.W. Petersen
CJ BIOTIN $ AVIDIN
Even higher sensitivity can be obtained by using biotinylated IgG and unconju-
gated avidin followed by a biotinylated enzyme, e.g. HRP or alkaline phosphatase
(the ABC-technique) (Fig. 26.12).
Streptavidin (molecular weight 60 kDa), a protein extracted from the bacterium
Streptomyces avidinii, is often used instead of avidin due to its lower non-specific
binding at neutral pH. Boon et al. (1990) have demonstrated a considerable re-
duction in incubation times with the ABC technique when applying microwaves
(cf.Sect.12.3.12).
26 Immunohistochemistry 381
Fig. 26.14. The PAP-technique with protein A as an intennediate layer. 1 = tissue seetion; 2 =
antigen in tissue section; 3 = specific antibody against the antigen; 4 = protein A; 5 = PAP-
complex.
P = horse radish peroxidase.
are eluted in glycine/HCI buffer, or in acid potassium permanganate, and this is fol-
lowed by incubation with a new primary antibody. It is essential that the antibody
used for detection of this second primary antibody carries a different label from
that used in the detection of the first antigen. FITC, TMRITC, alkaline phosphatase,
and HRP have all been used in this sort of procedure.
In a modified version of the "sequential incubation method", the primary an-
tibody is visualized by the PAP-method using DAß as substrate. By a further
incubation in cobalt chloride the brown DAß reaction product is changed into a
black substance and the enzymatic activity of HRP irreversibly inhibited. This ap-
proach enables the use of DAB again in the second round of incubation without
confusing the results of the first reaction. Clearly, this method cannot be used to
26 hnmunohistochemistry 383
demonstrate different antigens within the same cell because the brown reaction
product related to the second cell antigen would be obscured by the black reaction
product from the first round of incubation.
In the "multiple immunoenzymatic staining method" , elution of the first round
of antibodies is not necessary. If the primary antibodies have been raised in different
species (e.g. rabbit, goat, sheep, or swine), or if monoclonal antibodies of different
subclasses are used, the first incubation is performed in a solution containing both
primary antibodies at the same time. The localization of the primary antibodies
is then achieved using secondary antibodies directed against the different species
or subclasses. This approach also requires separate labels for the two secondary
antibodies (ß-galactosidase has become increasingly popular as a marker in this
area).
26.6.1 Fixation
It is therefore nonnally necessary to use some kind of fixation for retaining the
antigens. Fixation also improves the structural preservation.
Most antigen molecules in tissue are proteins or peptides consisting of long
chains of amino acid residues. Antibodies generaIly ''recognize'' 3 to 6 of these
residues (an antigenic "site"). Dehydrating fixatives (methanol, ethanol, and ace-
tone) and the precipitating fixatives (picric acid, mercuric chloride, potassium
dichromate, and acetic acid) are preferable from a theoretical standpoint because
they denature the proteins by coagulating the secondary and tertiary structure with-
out interfering with the primary structure. The overall structural preservation of the
tissue achieved with these fixatives is, however, often unsatisfactory.
In contrast, structural preservation is excellent using the cross-linking fixatives
(e.g. parafonnaldehyde, acrolein, and glutaraldehyde). The structural alteration im-
plicit in the action of these fixatives (Sect.12.2.1) suggests that they may interfere
with the antigenicity of proteins. This problem can be partially circumvented by
strict control of the fixation time (e.g. 4% parafonnaldehyde in 0.1 mol/l phosphate
buffer for 1 to 12 hours).
If glutaraldehyde is used for fixation, the concentration should be kept as low
as possible (0.1 %) although binding of antibodies raised against dopamine, no-
radrenaline, and serotonin has been demonstrated even after fixing with 5% glu-
taraldehyde.
One way of circumventing the decrease in antigenicity caused by fonnaldehyde
fixation iso to pretreat the section with proteolytic enzymes before immunostaining.
This enzyme pretreatment probably breaks the cross-links between the fixative and
the reactive groups of the antigenic sites. The three most commonly used enzymes
are trypsin, pepsin, and a non-specific protease named "Pronase". Whereas most
antigens are "demasked" by treatment with trypsin and "Pronase", some antigens
require treatment with pepsin in order to obtain a satisfactory demasking.
In contrast, some antigens are destroyed by enzyme pretreatment. It is therefore
important to detennine the extent to which enzyme pretreatment enhances staining
before it is used routinely. A further disadvantage of the enzyme pretreatment is
that sections easily detach from the slides. This problem can be solved by coating
the glass slides with gelatin or poly-L-Iysine.
It should be stressed that protease pretreatment should be applied only when tis-
sues have been fixed in fonnalin (parafonnaldehyde). Enzyme treatment of sections
fixed in coagulative fixatives usually causes total "digestion" of the section.
Finally, it is important to relate the time of enzyme treatment to the fixation
time. Too long an enzyme treatment on sections fixed for a short period of time
may give weaker staining reactions, due to digestion of the tissue material. Too
short an enzyme treatment of tissues fixed for longer periods of time will give
weak reaction due to insufficient demasking of antigenic detenninants.
386 P.P. Clausen, M. Mf.\ller, B. van Deurs, O.W. Petersen
26.6.2 Embedding
26.7 Quantitation
More detailed expositions on the issues raised below may be found in Weakley
(1981) and Glauert (1977).
27.1.1 Embedding
IL Lyon (Ed.)
Theory and Suategy in HistochemiSU)'
@ Springer Verlag 1991
388 B. van Deurs. M. Moller, O.W. Petersen
27.1.3 Fixation
Additional details and reaction types are given in Weakley (1981) and Glauert
(1977).
27.2.1 Carbohydrates
Carbohydrates on the cell surface with terminal sialic acid (sialoglycoproteins, etc.)
can be demonstrated with a number of cationic "dyes" such as Ruthenium Red,
Alcian Blue, and colloidal iron. Excellent ultrastructural localization of anionic
sites on the cell surface can be obtained with cationized ferritin, a polycation of
around 500 kDa, which due to its iron core, can be located direct1y by EM.
Various lectins (Sect.22.4) conjugated with gold or horseradish peroxidase are
useful for demonstrating carbohydrate components on the cell surface. For instance,
concanavalin Abinds to mannose, and Ricinus communis lectin, a toxic 60 kDa
protein from castor beans, binds to terminal galactose. The specificity of binding
can be checked by preincubating with unlabelled lectin or the sugar residue to
which it binds (e.g. 0.1 m01l1lactose in the case of Ricinus communis lectin).
The PASM method (Sect.9.2.2) for detecting 1,2-glycol groups by light mi-
croscopy can also be modified for EM work. Following oxidation with periodic
acid the free aldehyde groups reduce silver methenamine to metallic silver. This
produces a deposit of electron dense silver grains at the reaction sites and can there-
fore be used to detect various carbohydrate-containing molecules such as intracel-
lular glycoproteins in the Golgi complex, in secretory vesic1es, and in lysosomes
(lysosomal hydrolases are glycoproteins). The reaction is performed direct1y on
grid-supported ultrathin sections and usually produces a great deal of non-specific
staining. This can be reduced by treating the sections with chromic acid between the
periodic acid and the silver methenamine steps. Chromic acid acts by further oxi-
dizing some of the aldehyde groups generated by the periodic acid step to carboxyl
groups.
390 B. van Deurs. M. M~ller, O.W. Petersen
27.2.2 Receptors
The principles for these reactions have been discussed in detail in Chap.24 (Hydro-
lases) and Chap.25 (Oxidoreductases). For further references see Glauert (1977).
In studies of endocytosis and intracellular transport exogenous peroxidases are
frequently used as markers or tracers. The principle is that the cells are exposed
to peroxidase molecules for varlous periods of time and, following fixation (for
which glutaraldehyde can be used), the specimens are incubated in media containing
H202 and 3,3'-diaminobenzidine (DAB) and finally postfixed in OS04. The electron
dense reaction product is readily visible in the EM, and it is therefore possible to
obtain very detailed infonnation about the pathways and compartments involved in
intracellular transport of an endocytosed protein. Exogenous peroxidases are also
used as tracers in studies of capillary (or epithelial) penneability. For this purpose
aseries of peroxidases with increasing molecular weights are available:
27.3 Immunocytochemistry
ment. In order to attain this idealized goal, many factors have to be considered and
technical problems solved.
While the detection of antigens on the free surface of sampies such as cells in
monolayer cultures or single cells in suspension is fairly straightforward and can
even be performed on living cells (at 4°C), detection of antigens in tissue sampies
such as biopsies, partieularly intracellular antigens, is technically demanding.
Techniques for the ultrastructural detection of intracellular antigens can be
subdivided into preembedding and postembedding techniques. In the first case,
immunocytochemistry is performed on pieces of tissue, or on cells in culture,
be/ore the specimens are embedded and further processed for electron microscopy.
In the second case, tissue sampies or cells (cell pellets) are first embedded, ultrathin
sections cut, and the immunocytochemieal reactions are then done directly on the
thin sections.
The "label" used in ultrastructural immunocytochemistry must be electron dense.
Two types of label are of interest, enzymes and metal formulations. The enzyme
peroxidase (horseradish peroxidase, HRP) is used extensively (Kuhlmann, 1977).
When DAß is used as substrate, the polymerized oxidized diaminobenzidine, en-
hanced by treatment with OS04, provides an electron dense reaction product which
is readily observed by EM, provided the amount of label (HRP) exceeds a critical
level. The "metal" labels include ferritin and colloidal gold, the latter being most
popular at present. Gold partieies are normally used in sizes from around 3 um up
to 12-15 nm (larger sizes should not be used). In general, the larger the size, the
more homogeneous the preparation with respect to size, the easier it is to see at low
magnifications by EM. Either a (secondary) antibody directed towards the primary
antibody or, most often, protein A (a glycoprotein extracted from the capsule of
the bacterium Staphylococcus aureus) are adsorbed to the gold particles. In the
case of protein A-gold, the probe is referred to as PAG. Colloidal gold partieies of
different sizes can be used for double labelling experiments (see below).
Excellent results can be obtained with this technique. The main advantage is
that, once embedded in Epon, the block can be used again and again for making
numerous sections, even to the extent that serial sections can be processed for
three-dimensional reconstructions. The disadvantages of the technique are:
1. A gradient of immunolabelling is obtained from the exposed surface of the
specimen (from the cell surface for cuItured cells) to the centre due to the
decreasing degree of penetration of antibodies
2. Various intracellular compartrnents may be permeabilized and rendered accessi-
ble to antibodies to differing degrees. These factors make it critical to consider
for each section its original location within the tissue block, particularly when
any form of comparative study is proposed
3. Since HRP is the most commonly used label in preembedding techniques, quan-
titation cannot be performed It should be stressed that the amount or density of
the HRP reaction product cannot in any way be used to interpret the observa-
tions in quantitative terms. Even when using a particulate marker (e.g., ferritin)
quantitation is very difficuIt due to the penetration problems mentioned above
(points 1 and 2)
Before exposure to immunological reagents, the cells or tissue are embedded in ei-
ther EponR, LowierylR, LR WhiteR, or other hydrophilic resins, or iee (by freezing)
for sectioning. Because the sections are very thin, most of the intracellular com-
partments are "opened" as far as exposure of antigens to antibodies is concerned
(cf. permeabilization in the preembedding proeedures). Typieally the sections are
collected directly on grids for electron microscopy and then treated with antibodies,
etc.
In general, immunocytochemistry on sections of Epon-embedded specimens,
performed after etching the section surface in order to remove the hydrophobie
plastic around the antigens, is not considered a satisfactory technique although
satisfying and convincing resuIts have sometimes been obtained especially when
large amounts of antigen are present in the sections.
Embedding of specimens in hydrophilie resins such as Lowieryl (or LR White)
is often used for postembedding immunocytochemistry and, in general, good or
even excellent resuIts can be obtained, depending on the amount of antigen and
the (sub-)cellular structure to be studied. These hydrophilic plastics destroy less
antigens during the embedding procedure and probably allow a certain degree of
penetration of the antibodies. However, the sensitivity of the immunoreaction is
decreased as compared to the ultracryosections described below.
In principle, the best approach involving the least perturbation is to freeze the
cells or tissue, to cut Jrozen ultrathin seetions and use the thawed cryosections for
immunocytochemistry while they are hydrated. Abrief description of the principles
of applying immunocytochemieal techniques to ultracryosections in ultrastructural
work is given below under the following headings:
27 Ultrastructural Cytochemistry and Immunocytochemistry 393
Fixation and Cryoprotection. Small sampies of tissue, or cell pellets from cell
cultures are used. The fixative can be formaldehyde (1-8%) or formaldehyde-
glutaraldehyde mixtures (with 0.1-2.5% of the latter) in a "standard" buffer, de-
pending on the antigen to be localized In some cases the use of more specialized
fixatives may be warranted. The effect of various fixatives (as weIl as fixation
time and temperature) on antigenicity should be examined by ligbt microscopical
immunocytochemistry (immunoperoxidase or immunoftuorescence). Following fix-
ation and washing, the specimens are cryoprotected by incubation in 1.8-2.3 mol/l
sucrose for at least 15 min depending on the size of the specimens. Several changes
of sucrose are recommended.
Freezing. The cryoprotected specimens are mounted on special metal stubs (silver
or copper), the size of which depend on the cryoultramicrotome to be used, and
frozen in liquid nitrogen. When frozen on the stubs specimens can be stored in-
definitely in liquid nitrogen-containers, and be used for sectioning as required. In
this way, in principle, a "library" of specimen sampies can be obtained. This can
be used for controls, comparisons, etc., just as is the case with Epon-embedded
material for example.
are removed from the cryochamber of the microtome, the sections thaw and must
therefore be kept hydrated for the rest of the procedure
H. Lyon (Ed.)
Theory and Strategy in Histochemistry
© Springer Verlag 1991
398 P.E. HfiSyer, L. Kayser, M.R. Barer, H. Lyon
28.1.1 Prerequisites
instance by proteins), and the compound should not be lost from the sections
or cells (for instance by diffusion)
28.1.2 Errors
Table 28.2. The effect of cutting speed on section dimensions. Liver tissue cut with the micrometer
screw adjusted to 121!m (from Altman, 1980; reproduced by courtesy of the Royal Microscopical
Society,Oxford).
This variation also depends on the type of tissue. For quantitative work mi-
crotomes should be set to give sections that fall within the range yielding a linear
relationship between absorbance and section thickness (Fig. 28.1).
Absorption photometry can be used for the quantitation of all histochemical re-
action products that absorb light (ultraviolet 10 infrared) inc1uding native reaction
products such as NAD+. The basis of absorption cytophotometry is measurement
of light intensity in incident and transmitted light at a particular wavelength using
a beam of light directed at portions of the sampie containing the reaction product.
The usefulness of absorption photometry depends on whether the chromophore
400 P.E. Hfilyer, L. Kayser, M.R. Barer, H. Lyon
E
1.0
E = € X M / Ad x d = € X M/A or
(28.2)
M = EA/€
As the area over which measurements are made can be defined and € is known for
several final reaction products in tissue, it is possible to express absorbance values
in units of mass.
charged metal plate inside the glass sphere. This photoelectric effect results in a
current that is directly proportional to the intensity of the incident light.
As the precipitated chromophore is not uniformly distributed over the area, A
(equation 28.2), it is necessary to divide the area into a large number of small
regions of area, a. Otherwise, a very serious error may occur, distributional error
(Sect.28.2.2). In some commercial instruments the smallest area that can be mea-
sured equals the resolution of the light microscope (0.1-0.2 JLm), so ensuring that
the field measured is always optically homogeneous.
Most instruments make individual measurements on a very small area of the
section (e.g. a circular area with a diameter of 0.2-0.5 JLm; see also Sect.28.2.3).
Measurements over larger areas are achieved by taking multiple readings, either
by moving the section in small jerks between measurements (scanning stage), by
moving the illuminated area in the microscope in a similar pattern (ßying spot), or
by image analysis (Sect.28.2.3).
The central measurement provided by these instruments is the mean integrated
extinction (MIE) or as it is also called mean integrated absorbance (MIA). This is
given as relative values and is effectively the sum of all the individual absorbance
values obtained (corrected for any overlap between adjacent measurements) divided
by the number of small regions measured:
E = MIE = el + e2 ... e n
n
where el, e2 . .. are the individual absorbances measured. Substitution in equation
28.2 gives:
If, however, the same amount of reaction product is only distributed over half
of the microscope field then the absorbance in this part of the field would be
0.301 x 2 = 0.602. The corresponding transmission (1) may be calculated from
equation 28.1:
100
0.601 = log T or I = 25%
The amount of light transmitted through the entire microscope field would therefore
be:
100+ 25 . 100
2 = 62.5% corresponding to E = log 62.5 = 0.204
instead of 0.301.
In Table 28.3 some further examples are given of the distributional errors aris-
ing when absorbance of a heterogeneously distributed final reaction product is
determined by simple photometry without scanning.
Prom Table 28.3 it appears that the more unevenly the product is distributed
and the higher the true absorbance of the microscope field is, the more serious will
be the distributional error using simple photometry.
As shown in Table 28.3 the resultant error (distributional error) can be very sub-
stantial. This problem is approached by using a measuring spot that is sufficiently
small to make the area illuminated for one measurement effectively homogeneous
28 Quantitation in Histochemistry 403
Absorbance
Amount of determined
True absorbance transmitted light by simple Distributional
of microscope field (1%) photometry error in %
Glare and Diffraction Errors. Glare error (stray light) is caused by optical im-
perfections (dirt, etc.), particularly in the substage optics. It can be minimized by
ensuring that all optical parts are scrupulously clean, by eliminating surfaces (oil
instead of a dry objective), and by closing the field diaphragm appropriately.
Diffraction error on the other hand is not due to the optics of the system but
is due to the wave characteristics of light (Duijndam et al., 1980b). This error is
dependent on the shape and absorbance of the object and on the step size to object
size ratio and can be quite substantial for small and intensely stained objects.
Duijndam et al. (1980b) mention three ways of minimizing the effects of resid-
ual distributional error, glare error, and diffraction error in their conclusion. All
these errors tend to give lower mean integrated apparent absorbance values as
compared to the true absorbance values. The remedies suggested are:
1. To incorporate corrections for the errors in a computer scanning program
404 P.E. H~yer, L. Kayser, M.R. Barer, H. Lyon
2. To use a confocal scanning microscope with a laser beam as the light source.
Due to its coincident illuminating and measuring diaphragms, this instrument
is essentially free from glare and diffraction errors
3. To obtain corrected results directly by using the two-wavelength scanning pro-
cedure program (Bicoscan) developed by van der Ploeg et al. (1979) employing
the principles of two-wavelength scanning developed by Patau (1952) and Orn-
stein (1952)
Chromatic Error. Chromatic error may occur if a wide spectral band is used for
the measurements. This error will be most pronounced when the chromophore has
a narrow absorption band. Chromatic error may consequently be reduced by using
either an interference filter, a filter monochromator, a prism monochromator, or a
grating monochromator with narrow spectral bandwidths.
It is far beyond the scope of this chapter to describe these instruments in any detail.
They can in principle be classified into three groups:
1. The jlying spot principle. A stationary object is scanned by a moving light
beam either in the image plane (Barr and Stroud GN5: smallest measuring spot
0.25 p.m) or in the object plane (Vickers M85/85N86 series: smallest measuring
spot 0.2 p.m)
2. The scanning stage principle. A moving object scans across a stationary measur-
ing beam (Zeiss SMP 01; SMP 05; MPM series; Leitz MPV2: range of smallest
measuring aperture in the specimen plane is 0.2-0.5 p.m with a minimum step
width of 0.25-0.5 p.m)
3. The image analysis principle. The object is by means of a video camera con-
verted into a television image where each point is measured by densitometry.
(The minimum effective diameter of each image point is 0.1 p.m). A video
camera can act as a two-dimensional photometer. Two different principles of
video cameras are used. These are:
406 P.E. H\!Iyer, L. Kayser, M.R. Barer, H. Lyon
a. Image tube based or solid-state video cameras, e.g. charge coupled devices
(CCD-cameras)
b. slow scan solid-state cameras
Video cameras of the first kind produce images with a speed of 25 or 30
frames/second and can be USed with conventional image analysis systems. The
limitations of these video cameras are a limited range of linear relations hip between
input and output signals, geometric distortions, and a medium resolution (512 X 512
pixels). The advantage is low cost compared to slow scan cameras. Slow scan
cameras do not use television standards but produce images with a speed of 4-5
frames/second. The advantages of slow scan cameras are a high resolution (greater
than 1000 x 1300 pixels), a large range oflinearity, and lack of geometric distorsion.
The disadvantages are high cost and the need for special computer and image
analysis systems. It is, however, to be anticipated that as resolution for the slow
scan cameras further increases and their cost decreases, their use will become
more and more widespread. This is all the more probable when the high speed of
analysis together with the possibility for combining the analysis with morphometric
measurements are taken into consideration.
The use of video cameras in microscopy is described in detail by Inoue (1986)
and the use of solid state imagers by Aikens et al. (1989). Because of ongoing
changes in the supply of commercial instrument systems, no specific instruments
will be mentioned.
Two further quantitative techniques are relevant at this juncture, staining in artificial
gel systems and quantitation of staining by elution combined with conventional
absorption photometry. Both techniques are valuable for investigating quantitative
aspects of staining processes.
Stain-Elution. This was in fact one of the first methods applied to quantitation of
histochemical reactions. The preparation is stained in the usual manner and may
be inspected to determine the distribution of stain on a qualitative basis. Elution
of the dye or coloured end product is then achieved using a known volume of a
suitable solvent and absorbance determined in the extract. The amount of primary
reactant present in the section as a whole can then be inferred taking into account
28 Quantitation in Histochemistry 407
many of the considerations discussed in 28.1. Clearly, great care must be taken to
ensure the comparability of sections subjected to this procedure.
Recently the application of the stain-elution technique to cells cultured in vitro
in microtiter plates has raised some interesting possibilities (Barer et al., 1986a;
1986b). Firstly the use of clonally derived cells and distribution of identical aliquots
into wells of the microtiter plates provides high replicate numbers of an essentially
identical biological substrate for staining (up to 5000 identical monolayers can
easily be prepared at one sitting). Fixation, staining, and elution are all performed
in the microtiter plate and procedures are therefore limited to reagents compat-
ible with the plate plastic. Secondly the microtiter plate (an 8 by 12 array of
approximately 300J-l1 ftat-bottomed wells) provides a very convenient format for
studying the quantitative effects of varying most aspects of a staining procedure
(fixation, staining time, reagent concentration, etc.) on the formatioll of cell asso-
ciated end product. This aspect is greatly facilitated by the existence of dedicated
multi-channel microtiter pipetting devices.
Finally, relatively inexpensive equipment is available to obtain absorbance read-
ings of eluted stain in situ (ELISA or microplate readers); a standard plate reader
would read 96 wells in under 60 seconds and a more sophisticated instrument
could provide values at up to 5 different wavelengths simultaneously and transfer
the readings to a computer for further analysis.
The microtiter plate system is a somewhat heretical form of cytochemistry in
that it provides no indication of the subcellular location of staining (some infor-
mation can be obtained by direct inspection using an inverted microscope). Its use
is not confined to investigating quantitative aspects of staining processes, in fact
it was developed in order to study the effect of biologically active substances on
cells cultured in vitro. These effects are characterized in terms of changes in the
integrated staining properties of cell monolayers (Barer et al 1986a; Barer, 1987).
28.3 Fluorimetry
When molecules emit light solely due to high temperature, they are said to emit
incandescence. All other forms of light emission are called luminescence. The
emission of light as a result of absorbed radiation is called photoluminescence.
Fluorescence occurs when light is emitted following absorption of light from an
outside source of energy. Fluorescence is a type of luminescence in which light
is emitted from molecules 10- 8 seconds or less after they have absorbed light. If
the delay is about 10-6 seconds, it is termed delayed ftuorescence, while a delay
of greater than 10-6 seconds results in phosphorescence. In ftuorescence the re-
emitted light is of a longer wavelength than that absorbed. This shift in wavelength
is called Stoke's shift (Fig. 28.2).
The ftuorescence microscope differs from the light microscope in heing equip-
ped with two filter systems. There are also differences as regards the light source,
408 P.E. H!1lyer, L. Kayser, M.R. Barer, H. Lyon
Stokes shift
nm
Fig. 28.2. Excitation (Exc) and emission (Ern) spectra for a fluorochrorne. I = relative intensity;
abscissa = wavelength (nm)
illumination, and condenser systems as weIl as objective and ocular lenses. For
details, see for instance Taylor and Salmon (1989).
The filter systems are excitation (or primary) and barrier (or secondary) filters.
As excitation ideally should occur at a specific wavelength an interference filter is
used as the excitation filter and is inserted between light source and specimen. To
protect this filter, and also the specimen, it is necessary to position a heat protection
filter between the light source and subsequent functional filters. Moreover, neutral
density filters are often used for reduction of excitation light energy to avoid fad-
ing of fluorochromes. The barrier filter system is placed between specimen and
detection system and should ideally remove all light with shorter wavelength than
the light emitted from the specimen. Several different kinds of illumination may
be used (e.g. full aperture and dark field). The most commonly used are, however,
dia-illumination and epi-illumination.
The intensity of the emitted light depends on the numerical aperture of the
objective and on the magnification as follows: Idia ~ N A 2 M- 2 and Iepi ~ N A 4
where I = intensity of emitted light in dia-illumination or epi-illumination; NA =
numerical aperture of the objective; M =magnification.
There are a number of advantages of incident or epi-illumination. These include:
a far greater light intensity than trans-illumination at objective magnifications of
more than about IOx (cf. above), much less absorption of the emitted fluorescence
(Ploem et al., 1974), the possibility of using an inverted microscope, and finally the
possibilities for combination with conventional trans-illumination methods such as
phase contrast, polarization, and differential interference contrast.
In epi-illumination the objective functions as both condenser and objective lens.
Light from the light source passes into the microscope at a right angle to the axis
through objective, specimen, and detection system and having passed the excitation
28 Quantitation in Histochemistry 409
filter meets the chromatic beam splitter (dichroic mirror) placed at an angle of 45°
to the light beam. This mirror has an interference coating that has a high reflectance
at 45° for the wavelength transmitted by the excitation filter, but not for the longer
wavelength passed by the barrier filter. Conversely, the dichroic mirror transmits
the wavelength passed by the emission filter to the detection system, but not the
wavelength passed by the excitation filter.
A
A F
N
Fig. 28.3. Absorbance ( - - ) and fiuorescence ( ......) as functions of number of molecules
(N). A = absorbance, F = fiuorescence.
Reflectance is defined as the ratio between the intensities of reflected and incident
light. It may be calculated by using the fonnula:
xR = (n - N? + k2
(n - N)2 + k2
where R = specular reflectance at nonnal incident light, n = refractive index of
reflecting material, k = absorption coefficient of reflecting material, N = refractive
index of the medium into which the light is reflected. The prescript x of R indicates
this medium, such as immersion oll, water, etc. (e.g. oilR) (Piller, 1977, p.135).
The refractive index (n) depends, among other factors, on the wavelength. It is
referred to as normal dispersion when n increases with the frequency of light and
anomalous dispersion when n decreases.
Ploem (1975) proposed two contrast-enhancing devices which fundamentally
improved this approach for biological work:
1. An aperture central field-stop inserted in front of the collecting lens of the lamp
housing in order to limit unwanted reflection of the central beam of light
2. An objective with high numerical aperture and a rotatable quarter wavelength
plate mounted on the front lens. Such objectives were then produced by Leitz
(Patzelt, 1977) and Zeiss (Pentz and Schulle, 1980) and the method was spec-
ified as reflection contrast microscopy
Apart from adding these two devices to a fluorescence microscope with epi-
illumination it is necessarily to replace one of the fluorescence blocks containing a
dicbroic mirror and the fluorescence filters with a polarization block containing a
50% dividing mirror and two fixed and crossed polarizers (Cornelese-ten Velde et
al., 1989). To avoid reflection at glass-air interfaces, immersion oll with a refractive
index equal to glass is used between the objective and the specimen, that is, no
coverslip is used (Cornelese-ten Velde et al., 1988).
In reflection contrast microscopy bright objects are seen against a relatively dark
background, and in this respect the image may superficially resemble that seen with
fluorescence microscopy. Contrast with the latter procedure is due to luminescence
that occurs when a molecule, after absorbing radiation, emits a quantum of lower
energy (van der Ploeg and van Duijn, 1979). The emission is almost always of
Ion ger wavelength than the exciting radiation.
In reflection contrast microscopy of stained biological material, the image is,
however, either due to spectrally selective reflection or interference after reflection.
Spectrally selective reflection may occur in many types of stained material, for in-
stance in Feulgen stained cell nuc1ei or Giemsa stained chromosomes, and is due
to anomalous dispersion (van der Ploeg and van Duijn, 1979). Selective reflection
of light takes place at wavelengths near the absorbance peak of the chromophore
bound to macromolecules embedded in a not completely matching medium. Inter-
ference after reflection is, however, the main reason for the whitish yellow appear-
ance of the polymerized oxidation product of diaminobenzidine (DAß) in reflec-
412 P.E. H~yer, L. Kayser, M.R. Barer, H. Lyon
tion contrast microscopy and is due to the high refractive index of this compound
(Cornelese-ten Velde et al., 1988).
Reftection of light by bound chromophores may be used for the visualization
of microscopic material that may be difficult to distinguish using other methods.
Reftection contrast microscopy is a sensitive method for detecting small amounts
of the insoluble polymerized DAß oxidation product formed by the peroxidase re-
action with DAß and hydrogen peroxide as cosubstrates, allowing the detection of
details in the image having reftectance values of no more than 0.01 % (Comelese-
ten Velde et al., 1988). As reftectance of the polymerized DAß oxidation product
is mainly due to interference of reftections, the stain must be present in a very thin
layer, as in ultrathin sections, in the upper layer of semithin sections of LOwicrylR
embedded material, or on preparations of chromosomes (Cornelese-ten Velde et al.,
1988; 1989). Cornelese-ten Velde and Prins (1990) have, using post-embedding,
stained different types of antigens in LOwicrylR embedded material with immuno-
gold for electron microscopy and with subsequent silver enhancement for reftection
contrast microscopy. In principle, this makes it possible to use both reftection con-
trast microscopy and electron microscopy for visualizing the colloidal gold or silver
enhanced gold particles on the same ultrathin section.
Accordingly, reftection contrast microscopy is a valuable tool in immunohis-
tochemistry and in situ hybridization histochemistry when very small amounts of
oxidized DAß are present, for example detection of single copy DNA sequences
in chromosomes and quantitation of cell membrane receptors or other constituents
in different compartments of the cell (Landegent et al., 1985; Cornelese-ten Velde
et al., 1988; 1989).
Cornelese-ten Velde et al. (1989) compared the absorbance of oxidized DAß
with its reftectance. Slides coated with peroxidase-conjugated immunoglobulins
served as a model system. In addition, optimum conditions for the detection of sin-
gle copy DNA sequences in .metaphase chromosome preparations were established
by using in situ hybridization and reftection contrast photometry. The intensities
of reftected and incident light were measured using a 100% reftecting mirror in
the object plane for calibration. The authors found a linear relationship between
the amount of peroxidase and the reftectance determined for oxidized DAß. 1t was
concluded that quantitative immunoperoxidase studies with the reftection contrast
microscope are feasible.
28.5 Interferometry
ter depending predominantly on the local protein concentration in the section. This
approach can therefore be used to measure local protein concentrations and section
thickness, though it has largely been superceded by more specific techniques for
the former application.
The spectra of dyes and tissue bound chromophores can be used in several different
ways. These include simple comparisons, colorimetry, spectra1 subtraction, and
component spectral analysis.
The simplest approach to spectral analysis is to record transmittance, absor-
bance, or reflectance spectra of the chromophores in the tissue or cells under study.
The general shape of the spectrum is observed and maxima, minima, and shoulders
are determined. This makes possible the identification of a certain dye and can be
used for determining the purity of dyes (cf. Sect.3.3.6). The absorption spectra
of various blood cells stained with Azure B and Eosin were analyzed by Zipfel
et al. (1984) who found that three absorption bands could be distinguished in
the nuclei corresponding to DNA-bound monomers and dimers of Azure B and
to the Romanowsky-Giemsa band corresponding to a complex of DNA, higher
polymers of Azure B, and Eosin. In the spectrum of cytoplasm, absorption bands
corresponding to monomers, dimers, and polymers of Azure B bound to RNA were
identified (cf. Sects.6.1.1 and 6.3.2).
Further references to work of this kind of Romanowsky stained blood films
are Marshali et al. (1981), Marshall and Galbraith (1984a), and Flanders et al.
(1984). Galbraith et al. (1979) have analyzed the spectra of nuclei and cytoplasm
of different types of cervical cells stained by the Papanicolaou technique. It is
concluded that for automated image processing of this kind of material, the most
suitable wavelength is 531 nm to maximize the contrast of cell against background
and a wavelength in the range 560-605 nm to maximize the contrast of the nucleus
against the cytoplasm. A further reference to work on the Papanicolaou stain is
Galbraith and Marshall (1984).
and y values for a given specimen into this diagram. It should, however, be noted
that equal distances on this diagram do not represent equal perceived differences
in colour appearance. Several attempts have been made to transform the CIE 1931
Chromaticity Diagram by calculation to chromaticity diagrams with uniform chro-
maticity spacing. Standardized procedures have, however, so far not been agreed
on. For instance, Marshall and Galbraith (1984b) employed four different sets of
formulae to perform colour difference calculations for histological objects stained
according to two different Romanowsky methods.
CIE coordinates contain less information than the spectrum from which they
are derived. It is thus possible for two different spectra to give the same colour and
coordinates, a phenomenon termed metamerism. More detailed spectral informa-
tion may, however, be obtamed by spectral subtraction or by component spectral
analysis.
molecule. The latter has already been treated in Chaps.4--9. In Sect.12.3.11 factors
governing the diffusion rate of fixatives are discussed The arguments advanced
here also hold true for the diffusion of dyes.
Bentley et al. (1979) have performed a microspectrophotometric study of Azure
B Eosin staining of blood cells. The influence of stain formulation and staining
technique was studied by varying stain concentrations, staining time, pH, metal
salt contamination, dye contamination, buffer concentration, and fixation time.
Goldstein (1980) determined half-staining times for cytoplasmic RNA and nu-
clei with Azure A at pH 4 and 25°C using microdensitometry. Wmzek etal. (1987)
described the combination of a microspectrophotometer with a perfusion cuvette.
The cuvette made it possible to ensure laminar ftow of the staining solution in a
completely closed system containing cytological or histological material and thus
enabled continuous observation of a cell during the complete staining process.
A detailed review of factors pertaining to the kinetics of staining is given by
Horobin (1982a), pp.56-110.
Very little work has been done on quantitation of standard techniques applied to his-
tological sections. An example is, however, the work by Wohers et al. (1983) who
examined the effect of gradual, tolbutamine induced, degranulation and subsequent
regranulation on the histochemical1y detectable zinc (dithizone, cf. Sect.17.7.5) and
calcium (glyoxal-bis-(2-hydroxyanil) (GBHA), cf. Sect.17.7.1) content of B cells in
rat pancreatic islets. The staining intensities were determined cytophotometrical1y
and compared to the insulin content determined by staining with Aldehyde Fuchsin
and by radioimmunoassay of insulin extracted from pancreatic tissue.
Although tissue-bound metal salts have received little attention, over the last
ten years a number of ftuorescent probes have been developed for the specific
demonstration of individual diffusible ions in living cells. These techniques, which
include methods for monitoring Ca2+, H+, K+, Na+, Mg2+, and Cl- in real time,
are discussed in section 28.8.10.
28 Quantitation in Histochemistry 417
28.8.2 Pigments
In principle, several methods could be made quantitative but this has not been
attempted to any substantial degree. The ferric ferricyanide reaction (Sect.8.3.1)
has been used for the quantitative assessment of glutathione (cf. Sect28.8.1O).
It is, however, not specific for this compound as it also demonstrates ascorbic
acid, cysteine residues, lipofuscin, melanin, and biogenic amines. This raises the
possibility of quantitating some of the latter non-extractable substances in paraffin
sections.
The biogenie amines can be demonstrated and quantitated by formaldehyde
induced ftuorescence (Sects.18.4.1O, 30.2.1). Arecent example has been given by
Hougaard and Larsson (1990) who found a linear relationship between cellular
polyamine concentration and formaldehyde-ftuorescamine ftuorescence yield.
28.8.3 Lipids
Staining with lysochromes can be combined with a subsequent extraction and chro-
matographie quantitation.
The acid Haematein method of Baker (cf. Sect.19.6.7) was applied to cryostat
sections of human rheumatoid synovial membranes by Henderson et al. (1978a).
Prior to the acid Haematein procedure sections were extracted by treatment with
1:3 (v/v) methanol:chloroform (cf. Sect.19.3). Measurements were performed at
the absorption maximum of 575 nm and the non-specific staining measured over
collagenous stroma was subtracted.
Quantitation of cellular DNA has been used for studies on cell growth, tumour biol-
ogy, and phylogenetic relationships. The DNA content of individual chromosomes
can also be measured.
Measurements of DNA have been made with the Feulgen reaction, ftuorescing
Schiff reagents, autoradiography, with Cr-gallocyanin, with various cationic dyes
such as Azure B and Methyl Green-Pyronin, and by ultraviolet spectrophotometry.
"older" DNA (cf. Sect.31.8.3). Note, however, that differences in hydrolysis times
of these "different" types of DNA only are apparent when using cold hydrolysis
(5 mol/l HCI at 20°C; cf. Sect.9.9).
Errors have been considered in detail. Theoretical methods were developed for
correcting absorption values determined with a scanning stage microspectropho-
tometer and these were subsequently tested in a practical examination of con-
densed, normal, and dispersed Feulgen stained chicken erythrocyte nuclei (Duijn-
dam et al., 1980a;b). They found a considerable degree of concordance between
the predicted and experimentally determined magnitudes of systematic errors due
to inhomogeneous distribution, diffraction, and glare; these sources of error are
often pronounced when small, intensely stained objects are measured.
Standards that have been used include Feulgen stained nucleated chicken ery-
throcytes, neutrophil granulocytes, and lymphocytes (2n DNA) oNpermatozoa (1n
DNA).
In quantitative studies, the Feulgen reaction has been combined with several
other methods for the simultaneous demonstration of dry mass, RNA, his tones,
non-histone proteins, enzyme activity, and immunoreactive substances (Welch et
al., 1979; Herzog, 1982; Olenev, 1983; Fukuda, 1983; Schiek et al., 1987).
Fluorescent reagents. Two approaches have been used for the quantitation of DNA
using fluorescent reagents. The first involves the use of fluorescent Schiff reagents
prepared from dyes such as Acriflavine and Auramine O. These are applied after
acid hydrolysis and are subject to similar considerations to those discussed under the
Feulgen reaction. The second technique is the application of fluorescent compounds
that intercalate between adjacent base pairs (e.g. ethidium bromide and propidium
iodide). These reagents have primarily been used in flow cytometry but they can
also be used for fluorimetric determinations on tissue sections and cytological
preparations.
Few comparisons have been made between the results obtained using Feulgen
techniques and DNA intercalating fluorochromes. Scherini et al. (1988) found that
DNA content determined by fluorimetry of cells stained with Hoechst 33342 gave
lower values than the Feulgen technique in cerebellar cells from rats treated with
bleomycin. The authors put forth the hypothesis that the observed differences in part
are due to differences in dye-DNA interaction, and that the reduction in Hoechst
staining at least is partly due to a decrease in the content or accessibility of adenine-
thymidine base pairs of DNA to the fluorochrome in cerebellar cells in animals
treated with bleomycin.
Cationic Dyes. Methyl Green-Pyronin (Sect.6.1.5) has been used in several stud-
ies. The problems include, the purity of the dyes used, cross-over in absorbance
spectrum between the two dyes, and uncertainties concerning the mechanism and
28 Quantitation in Histochemistry 419
28.8.5 Proteins
Anionic Dyes. These have been used extensively. For some of these reagents,
problems due to relatively low dye-substrate affinities have to be circumvented
(Klunk et al. 1989a). One approach has been to omit the washing step after staining
and to separate the staining solution from the substrate (e.g. cells) by filtration. Dye
binding is then determined by measuring the decrease in dye concentration in the
filtrate (Klunk et al., 1989a). The problem can be avoided altogether when spectral
changes are induced in the dye when it binds to certain proteins (e.g. Congo Red
and amyloid-like proteins, Klunk et al., 1989a; b).
420 P.E. Hf2Iyer, L. Kayser, M.R. Barer, H. Lyon
At pH 2.8 (aqueous 1% acetic acid), Naphthol Yellow S, Light Green SF, and
Orange 11 can all be used for quantitative staining of total protein (see Sect.21.4.2;
Deitch, 1955; Oud et al., 1984; Oud, 1986).
Nucleolar proteins have been quantitated using Phloxine B at pH 6.8. This
procedure has been developed for automatic assessment of nucleolar size by image
analysis (Boon et al., 1988).
Amido Black lOB has been widely used for staining protein in chromatog-
raphy (Lillie, 1977, p.142). Histologica1ly the dye has been used as astain for
haemoglobin and as a highly selective stain for basement membranes, reticulum,
and collagen where it is applied in Picric Acid mixtures of the van Gieson type
(Sect.21.6; Lillie, 1977, p.142). Schauenstein et ale (1980) determined the molar
extinction coefficient of complexes of Amido Black lOB and bovine serum albu-
min and went on to describe how a quantitative microspectrophotometric protein
determination could be done on single cells. A modification of the method for
quantitative microspectrophotometric determination of proteins in tissue sections
was presented by Nöhammer (1984).
28.8.6 Carbohydrates
PAS Reaction. The applicability of the PAS (periodic acid Schiff) reaction to
quantitative studies has been discussed extensively. Clearly the widespread accep-
tance of the Feulgen technique for quantitative work must have some bearing on
these discussions. Moreover, both Hoogwinkel and Smits (1957) and Olsson and
Dahlqvist (1967) have shown that glycogen solutions react stoichiometrically and
that the reaction product conforms to the Beer-Lambert law.
Several studies demonstrate the importance of the relative availability of gly-
cosyl units for oxidation. Halkjler-Kristensen and Ingemann-Hansen (1979) found
that only 15% of the glycosyl groups were demonstrated in tissue sections, while
Ovadia and Stoward (1971) showed that only 15-23% of the total glycogen content
in liver sections as determined chemically could be oxidized by periodic acid. (For
further discussion of this paper, see Sect.9.2.1). In molecular model experiments,
Puchder et al. (1974) showed that Schiff's reagent cannot diffuse freely through
polysaccharides and that only every second glycosyl unit on the surface of glycogen
granules can combine with Schiff's reagent for steric reasons.
Gahrton and Yataganas (1976) have tried to circumvent these problems by de-
veloping a system in which the total glycogen content of cells can be determined
by reference to a microdroplet standard. Development of the microdroplet model
allowed these workers to study the relationship between dry mass, as determined
by X-ray absorption, and PAS staining using both standard and ftuorescent Schiff
reagents (Sect.28.8.4). Results using the two types of Schiff reagent gave a highly
correlated linear relationship and the authors concluded that the microftuorimetry
method could be used for glycogen determination in single cells providing certain
procedural constraints are closely adhered to. These workers then went on to de-
28 Quantitation in Histochemistry 421
Cationic Dyes. Quantitative work with cationic reagents that demonstrate sulphate
and carboxyl groups in acid glycosaminoglycans has been sparse. An analysis of
Alcian Blue and combined Alcian BIue-Safranin 0 staining procooures has been
made by Tas (1977) using polyacrylamide films containing different glycosamino-
glycans. When used alone, Alcian Blue staining followed the Beer-Lambert law
making quantitation possible, while the combined Alcian Blue-Safranin 0 proce-
dure could not be used as results were highly dependent on the amount of Alcian
Blue bound to the glycosaminoglycan during the first step of the staining pro-
cedure. Safranin 0 can, however, be used alone for the microspectrophotometric
quantitation of glycosaminoglycans in cartilage (Kiviranta et al., 1985a).
Quantitative studies in this field include, Zoller et al. (1981) who used Alcian
Blue pH 2.5, Berrisford et al. (1985) who used both PAS and Alcian BIue pH 2.5,
and Dunham (1979) who used the Alcian Blue critical electrolyte concentration
method (cf. Sect. 6.1.2). Finally, Harkema et al. (1987) have used automated image
analysis of high iron diamine (Sect.22.2.2) stained material to quantitate sulphated
mucins; however, it is uncertain whether this method fulfils all the requirements
stated in Sect.28.1.1.
28.8.7 Enzymes
2. Reproducibility
a. The mean values of measurable parameters do not vary significantly in
repeated experiments
b. The individual measurements of a given parameter within apreparation
form a statistically-defined unimodal population
A method for testing these parameters is given by Stoward (1980).
3. Validity
a. No enzyme is lost during incubation or if some is lost, the loss is small,
constant, and known
b. The specific FRP has its expected chemical composition
c. There is a stoichiometric relationship between the amount of specific FRP
in the specimens and the amount of primary reaction product formed by the
enzyme
d. The mean absorbance or ftuorescence emission of the specific FRP is pro-
portional to its mean concentration in the specimen
e. Apart from ftuorimetry (Sect28.3), the mean absorbance of specific FRP
is proportional to the thickness of the section up to a certain critical level
when using zero order kinetics
f. The reaction rate in a particular site is direct1y proportional to the specific
activity of the enzyme at this site
g. Mean absorbance or ftuorescence emission of the specific FRP increases
uniformly with incubation time (preferably linearly)
h. When performing regression analysis of the absorbance- or ftuorescence-
incubation plots, the line should preferably extrapolate through the origin
i. When the substrate concentration is above a certain level, the reaction rate
should reach a constant maximum rate (zero order kinetics)
j. The Michaelis constants should preferably be comparable to those deter-
mined biochemically. It should, however, be noted that the KM values de-
termined using PVA media often are an order of magnitude higher
k. Enzyme modifiers (e.g. inhibitors) should have the same effect on rate of
final FRP formation as known from biochemistry, and this should preferably
be of the same order of magnitude
1. Control experiments should indicate that the amount of non-specific final
reaction product is small and preferably constant (not more than 5-10% of
the specific FRP
4. Specificity. A technique is judged to be specific if:
a. No specific FRP is formed in controls
b. The reaction conditions that give rise to maximum formation of specific
FRP in situ are the same or very similar to those used biochemically
c. Modifiers (inhibitors) of the enzyme act in a similar manner as observed
biochemically
d. Potentially interfering enzyme systems have either been suppressed, elimi-
nated or shown to be absent, or can be distinguished from the enzyme under
study and measured separately
28 Quantitation in Histochemistry 423
At present several techniques have been tested against these criteria. Examples
are acid phosphatase (Stoward, 1980), the mixed-aggregation immunocytochemi-
cal technique (MAGIC) described by Wachsmuth (1980), and glucose-6-phosphate
dehydrogenase (van Noorden, 1984).
These criteria have inspired much of the work on quantitative enzyme cyto-
chemistry perfonned during the last ten years as exemplified by the fact that most
of the papers referred to in Tables 28.4-28.6 to varying degrees have used these
criteria.
At this point the reader is reminded that the sensitivity of the technique consti-
tutes an additional criterion (Sect.28.1.1).
At the light microscopicallevel enzyme activity may be quantitated by absorp-
tion cyrophotometry (Sect.28.2.1) or by microftuorimetry (Sect.28.3).
At present, even though microftuorimetry is a very sensitive technique, no
ftuorometric enzyme histochemical method adequately fulfils the above criteria.
In particular, precise localization presents a problem as the available ftuorogenic
reagents lack sufficient substantivity and are not sufficiently insoluble in water
(Raap, 1986).
Kinetic Measurements. There are two ways to make kinetic measurements; these
are end-point measurements and continuous monitoring.
In end-point measurements incubation is perfonned for an appropriate certain
time after which the reaction is stopped and the section mounted.
The advantages of this approach are that kinetic values can be obtained from
as many areas as found necessary, and the approach is not limited to one-step
procedures (i.e. post-coupling, metal salt, as well as plateau absorbance methods
can all be used). If, however, the final reaction product diffuses into the solution
used for stopping the reaction, continuous monitoring is the only possibility.
Table 28.4. Examples of quantitative cytochemical demonstration of hydrolases.
ß-D-galactoside/ferricyanide in
N-acetyl-ß-glucosaminidase 3.2.1.30 post-coupling/naphthol AS-BI N-acetyl-ß-glucosaminide/ Robertson, 1980
Fast Gamet GBC 1
ß-glucuronidase 3.2.1.31 simultaneous coupling/naphthol AS-BI glucuronide/ Henderson, 1984; Schofield f
hexazotized Pararosanilin et al., 1983 r
Aminopeptidase A 3.4.11.7 simultaneous coupling/ex·L-glutamic acid-4-methoxy- Lojda and Gossrau, 1980;
2-naphthylamide/Fast Blue B Kugler, 1982a, b, c ~
Dipeptidylpeptidase II 3.4.14.2 simultaneous coupling/Lys-Ala-4-methoxy-2-naphthyl- Gossrau and Lojda, 1980 ~
amine/hexazotized Pararosanilin s:;:
Dipeptidylpeptidase IV 3.4.14.5 simultaneous coupling/Gly- Pro-4-methoxy-2-naphthyl- Ruhnke and Gossrau, 1989 ;:0
amine/Fast Blue B I:tl
Na + /K + -transporting ATPase 3.6.1.37 metal salt/adenosine-5' -tri phosphate/lead ammonium Chayen et al., 1981
citrate-acetate complex/H 2 S ~
metal salt/p-nitrophenyl phosphate/SrCI 2 /CoCI 2 / Firth, 1987 ;:t:
(NH 4 hS
Cal+ -transporting ATPase 3.6.1.38 metal salt/adenosine-5' -tri phosphate/lead citrate/ EI-Sherif et al., 1990 ~
ammonium sulphide 3 15
Primary i
Trivial name E.c. No. acceptor Function/location References e.
S
Lactate dehydrogenase 1.1.1.27 NAD+ glycolysis/cytosol Evans et al., 1980; Stoward and g'
Nakae, 1988; van Noorden 5r
and Vogels, 1989b g;
3-hydroxyacyl-CoA 1.1.1.35 NAD+ final step in ß-oxidation of fatty acids to acetyl Chambers et al., 1982
dehydrogenase CoA/mi tochondria
6-phosphogluconate 1.1.1.44 NADP+ pentose shunt/cytosol Altman, 1972; langes and van is.
dehydrogenase Noorden, 1989
Glucose-6-phosphate 1.1.1.49 NADP+ pentose shunt/cytosol van Noorden, 1984; langes and 3
dehydrogenase van Noorden, 1989
20Cl-hydroxysteroid 1.1.1.62 NADP+ progesterone is converted to a less active meta- Robertson et al., 1982b
dehydrogenase (formerly (NAD+) bolite/not established
1.1.1.149)
3ß-hydroxy-t1 5 -steroid 1.1.1.145 NAD+ synthesis of steroid hormones/smooth endoplas- Robertson, 1979
dehydrogenase (NADP+) matic reticulum, mitochondria (in some cells)
Glycerol-3-phosphate 1.1.99.5 FAD transfers reducing equivalents into mitochondria/ Lomax and Robertson, 1990
deh ydrogenase outer surface of the inner mitochondrial mem-
(mitochondrial) brane
Aldehyde dehydrogenase 1.2.1.3 NAD+ oxidation of aldehydes, e.g. metabolites of biogenie Chieco et al., 1986
1.2.1.4 NADP+ amines or corticosteroids/mitochondria and cy-
1.2.1.5 NAD(Pj+ tosol
Glyceraldehyde-3-phos- 1.2.1.12 NAD+ glycolysis/cytosol Henderson, 1976
phate dehydrogenase
Succinate dehydrogenase 1.3.5.1 or FAD citric acid cycle/mitochondria Evans et al., 1980; van Noorden
1.3.99.1 and Vogels, 1989b; Baker and
Santer, 1990
Glutamate dehydrogenase 1.4.1.2 NAD+ catalyzes the interconversion of L-glutamate and 2- Kugler, 1990b
1.4.1.3 1 NAD(P)+ oxoglutarate/mitochondrial matrix
1.4.1.4 NADP+
NADH dehydrogenase 1.6.5.3 and FMN oxidation of NADH/mitochondria, (smooth endo- Altman, 1972; Robertson, 1979
1.6.99.3 plasmatic reticulum)*
NADPH dehydrogenase 1.6.99.1 FMN oxidation of NADPH/smooth endoplasmatic re- Butcher et al., 1979; Robertson
ticulum, (mitochondria)* et al., 1982b
Cytochrome-c oxidase 1.9.3.1 O2 catalyzes the final step in the mitochondrial respir- Old and Johnson, 1989
atory chain/mitochondrial inner membrane
Iodide peroxidase 1.11.1.8 H 20 2 oxidation of iodide and formation of iodotyrosyl Ealey et al., 1984
(thyroglobulin) and iodothyronine/rough endo-
plasmatic reticulum, Golgi complex, peroxi-
somes, luminal side of plasma membrane (thy-
roid)
~
fr
is::
~
tD
~
;x:
~
g
Table 28.6. Examples of quantitative cytochemical demonstration of enzymes other than hydrolases and oxidoreductases. ~
~
428 P.E. H~yer, L. Kayser, M.R. Barer, H. Lyon
With this latter approach an appropriate area in the section is selected. At zero
time the incubation medium is either poured into a PerspexR ring surrounding the
section (cf. Fig. 25.3) or applied between slide and coverslip, preferably separated
from each other by thin spacers. When using a PerspexR ring it is necessary to use
a water immersion objective with a long working distance (van Noorden, 1988b).
Usually, the first measurement is made after 15 seconds and subsequent measure-
ments may for example be made every 15 seconds. The "blank" value obtained
after the first 15 seconds is subtracted from each subsequent measurement (van
Noorden, 1988b).
This approach is limited to one-step reactions. If the microspectrophotometer is
equipped with a computer-controlled fast scanning stage, it is possible to perform
sequential monitoring of multiple fields in each section (Lomax et al., 1989).
Incubation on the microscope stage of a cytophotometer can be performed at
room temperature, in a hot room, or using a heated stage (Altman, 1978).
Measurements may be performed using an image analyzer or a microspec-
trophotometer. The latter may use either the flying spot principle or a fast scanning
stage (cf. Sect.28.2.3). Usually, the measurement is made at the absorption max-
imum of the final re action product. In case of high mean integrated absorbance
values (MlE), it may, however, be necessary to make the measurements off-peak
in order to avoid out of range error of the instrument (cf. Sect.28.2.2).
In the case of ditetrazolium salts measurements must be made at the wavelength
(the isobestic point) at which the corresponding formazans have the same molar
extinction coefficient (cf. Sect. 25.1.4).
To illustrate how absolute MIE (MIA) values can be converted to absolute
units that may finally be compared to corresponding biochemical results, a simple
example is calculated below:
Suppose that a specimen of diameter 10 p,m has an absolute MIE of 0.15 at
the isobestic point for NBT after 5 min incubation, then
and with € at 585 nm = 16000 litre mol- 1 cm- 1 (Butcher, 1978) substitution in
equation 28.3 gives:
most suitable approach is 10 relate the quantity measured to total protein content,
or perhaps even better total DNA content, per unit volume or per cel1.
Computer programs have been developed for cytophotometric measurements
of double stained preparations allowing the simultaneous determination of protein
content and enzyme activity. Proteins are stained with Naphthol Yellow Safter
the sections have been reacted for the enzyme activity. Thus a dual wavelength
scanning cytophotometry program (Bicoscan) developed by van der Ploeg et al.
(1979) may be employed for this purpose. This program will compute the corrected
local absorbance values at specified wavelengths for each chromophore at each
measuring spot and integrate these values over the total object to give separate 10tals
for each chromophore. Van der Ploeg et al. (1979) originally used this program for
simultaneous determination of DNA (Feulgen) and protein (Naphthol Yellow S).
Biochemistry normally correlates to total volume though from the cytochemical
viewpoint correlation to total volume of cells would seem more reasonable when
enzyme activities are concerned.
It would appear that image analysis is the easiest way of determining total
volume of cells.
The importance of measuring the specific reaction rate cannot be sufficiently
stressed. The specific initial reaction rate is the test reaction (complete incubation
medium) minus the control reaction (medium without substrate). In most cases
test reaction minus control reaction is a sufficient correction. Van Noorden and
Vogels (1989b) found, however, that the test minus control reactions for lactate
dehydrogenase were distinctly non-linear for an tissues tested. This appeared 10
be due to product inhibition by pyruvate generated during the reaction, and it
was concluded that the appropriate control reaction in this case was the reaction
obtained by adding sodium pyruvate to a final concentration of 20 jlmol/l in the
complete incubation medium.
Finally, it should be stressed that if an overall understanding of the key metabolic
pathways is to be achieved the activities of flux-generating (rate-limiting) enzymes
should be determined (Newsholme and Start, 1976; Newsholme et al.,1980). Un-
fortunately only a few quantitative methods exist for these enzymes at present.
Van Noorden and Butcher (1991) provide arecent general review of quantitative
enzyme cytochemistry.
28.8.8 Immunohistochemistry
1. Microftuorimetry
2. Counting of gold or ferritin particles using the colloidal gold or immunoferritin
methods
3. Scanning and integrating microspectrophotometry
This section is, however, only concerned with the microspectrophotometric
aspect Qnly relatively few quantitative immunohistochemical studies of this kind
have been published. Some examples are given in Table 28.7.
When performing immunocytochemistry it is imperative to have as much infor-
mation as possible conceming the antigens and antibodies used in the study. While
this has been universally accepted the establishment of relevant criteria has. met
with considerable difficulty. The Immunocytochemistry Editorial Sub-Board of the
Histochemical Journal agreed by majority in 1985 (News and Views, 1985) that
immunocytochemical manuscripts should fulfil the following criteria as a minimum:
1. The antibody used must be characterized as to the specificity of its affinity for
the antigen in the cell or tissue under study. In addition, molecular character-
istics (e.g. molecular weight) should be provided.
2. Any quantitative or semiquantitative statement must be justified by appropriate
measurements.
In the same publication, (News and Views, 1985), Sternberger pointed out that
the above criteria only consider serum-derived antibodies and ignore the existence
of monoclonal antibodies where the establishment of specificity is redundant. Stern-
berger further points out that most specificity tests in fact are less sensitive than
immunocytochemistry .
In response to the above, Montero, (News and Views, 1986), recommends
that an additional criterion should be inclusion of the parameters of any fixation
procedure used.
At this point it is worth remembering that, whether or not fixation is an integral
part of the immunohistochemical method adopted, there are some further problems
in the quantitation of antigens in tissue sections. These, according to Larsson (1988),
pp.193-194, include:
1. Epitope preservation by fixation and post-treatment
2. Penetration of antibodies and detection reagents
3. Steric hindrance (including prozone phenomena)
4. Efficiency of the detection procedure
Moreover, it is necessary to optimize every step in the ~ubsequent immuno-
cytochemical procedure. In addition, it is extremely important to include relevant
controls (cf. Sect.26.5). A very thorough review of control procedures at different
steps of immunohistochemical methods is given by Larsson (1988), pp.19-36.
When enzyme labels are used in microspectrophotometric quantitation, it is
aprerequisite that there is a stoichiometric relationship between the amounts of
antigen and final reaction product.
Sternberger and Sternberger (1986) compared the peroxidase-antiperoxidase
(PAP) technique (Sect.26.3.3) with the avidin-biotin complex (ABC) technique
(Sect26.3.4) and concluded that in principle the PAP method, unlike the ABC
method, is suitable for making quantitative estimates of the concentration of an
&
IV
Class II HLA-DR human tonsil mouse monoclonal MTT microspectrophoto- direct glucose oxi- Poulter et al., 1987
metry dase
Oestrogen receptor human breast rat monoclonal DAB microspectrophoto- PAP Petersen et al., 1987
metry
Immunoglobulins human synovial rabbit polyclonal AEC microspectrophoto- PAP Fritz et al., 1988
membrane metry
NADPH-ferrihaemo- rat liver sheep polyclonal DAB microspectrophoto- a) PAP Smith et al., 1983 ."
protein reductase fluorescence metry; micro- b) indirect FITC tn
(EC 1.6.2.4) fluorimetry
Neurofilaments rat brain mouse DAB image analysis a) ABC-peroxidase Sternberger and
$
.~
monoclonal DAB b) PAP Sternberger, 1986
Metallothionein-l rat placenta rabbit polyclonal DAB microspectrophoto direct peroxidase Roelfzema et al., r
metry 1989
Mye1in basic protein rat brain rabbit polyclonal DAB image analysis PAP Sternberger et al., ~
1978 ~
Tyrosine 3-mono- rat brain rabbit polyclonal DAB image analysis PAP Benno et al., 1982a, b s::
oxygenase ~
(EC 1.14.16.2) 1;1:1
~
;:r:
~
§
28 Quantitation in Histochemistty 433
antigen. The authors showed that when staining intensity was plotted against anti-
body dilution a Bat curve was obtained with the ABC technique. In contrast staining
intensity with the PAP technique showed an initial increase with progressive di-
lution of the antibody. Staining intensity then progressively decreased following
an S-shaped curve. The initial increase in staining intensity has been explained by
Bigbee et al. (1977) who noted that with the PAP technique decreased staining
intensity or even false negative results might arise when a high concentration of
antigen was combined with a high concentration of the primary antibody probably
giving rise to steric hindrance. They believe that in this situation the secondary an-
tibody may form cross-bridges to two antibody moleeules from the first antiserum
and therefore lose its ability to bind the PAP complex.
When measuring the activity of an enzyme it is normally aprerequisite that ini-
tial maximum rate kinetics are used A plot of the time-course of product formation
for an HRP catalyzed re action is shown in Fig. 28.4.
EXTINCTION
0.4
0.3
0.2
o. -
~o 20 30 40 50 60
MIN
Fig. 28.4. Produet formation (extinetion) as a funetion of time for a horse radish peroxidase (HRP)
eatalyzed reaction. Pancreatie islets, rat.
Primary antibody: Specifie islet eell eytoplasmie antibody (lCA)
Visualization: HRP eonjugated protein A.
Cosubstrates: DAß Img/ml; H202 0.02%
(Marshali and Hjilyer, unpublished results, 1989)
In this example initial rate conditions only last a very short time (15-20 seconds)
after which levelling off occurs. Several possible causes for this departure from
linearity which are equally well-recognized in biochemistry, are listed below:
1. The reaction may be running out of substrate
2. The reaction may be approaching equilibrium
3. One of the products of the reaction may inhibit the enzyme
434 P.E. H~yer, L. Kayser, M.R. Barer, H. Lyon
4. One of the components in the system is unstable under the assay conditions
and is steadily being decomposed
5. An inhibitor of the enzyme may be present in the incubation medium
6. Loss of linear response by the microspectrophotometer (out-of-range error, cf.
Sect.28.2.2)
7. The pH or ionic strength may change during the assay
These causes are elaborated on below:
1. In enzyme cytochemistry it is unlikely that the reaction will run out of substrate
as the volume of incubation medium usually far exceeds that of the tissue.
Theoretically, it could become a problem during continuous monitoring when
the layer of incubation medium applied between slide and cover slip is thin
2. Equilibrium conditions do not occur with irreversible reactions, while for re-
versible reactions the probability of equilibrium is normally small as the reaction
products diffuse from the enzyme site into the incubation medium. Equilibrium
may, however, OCCur if one of the reaction products is unable to diffuse away,
or does so very slowly, as may be the case if aPVA technique is used (Gordon
and Robertson, 1986)
3. Inhibition due to one of the reaction products is a serious problem in the
demonstration of 3ß-hydroxy-~S-steroid dehydrogenase, even when the use of
a PVA technique is avoided, and may also cause problems in biochemistry
(Caffrey et al., 1979)
4. Instability of one of the components is well exemplified by heat inactivation of
alkaline phosphatase (van Duijn and van Noorden, 1989)
5. The presence of an inhibitor in the incubation medium may explain levelling
off in the case of HRP as one of the substrates (H202) has an inhibitory effect
on the enzyme
6. Out-of-range error has been dealt with in Sect.28.2.2
7. Changes in pH taking place during the assay are easily checked. Possible
changes in both pH and ionic strength can be calculated using the computer
program suggested by Clancy (1987)
Apart from these seven factors, the possibility of the enzyme becoming "clog-
ged" has been suggested as a reason for departure from linearity. Clogging of
the enzyme site might be due to precipitation of the final reaction product here
or to the cell and tissue matrix becoming obstructed. As noted by van Duijn and
van Noorden (1989), no direct evidence for this phenomenon has, however, been
presented.
On the basis of both theoretical calculations and practical experiments van Duijn
and van Noorden (1989) point out that if plateau formation is due to gradual enzyme
inactivation, it is possible to use plateau absorbance measurements as a parameter
for enzyme activity, provided that the following requirements are fulfilled:
1. Enzyme inactivation should follow first order kinetics (Sect.23.1.3)
2. Other possible reasons for levelling off can be excluded
3. There exists a stoichiometric relationship between plateau absorbance values
and amount of enzyme activity present
It has been demonstrated that these requirements are complied with in the case
28 Quantitation in Histochemistry 435
28.8.9 Glutathione
Smith et al. (1979) used the ferric ferricyanide method (Sect.8.3.1) to quantitate
the distribution of reduced glutathione in cryostat sections of rat liver. Tbe results
obtained by microdensitometry were compensated for light scatter by subtracting
the absorbance at the minimum absorption of Prussian Blue (475 nm) from the
absorbance at the maximum absorption (675 nm). Although the Schmorl reaction is
not specific for glutathione, the authors conclude that most of the reaction product
fonned is due to glutathione as there is a 70-fold greater concentration of this
substance in the liver than of ascorbic acid.
Infonnation about the local concentrations of diffusible ions can be obtained using
fluorescent probes or X-ray microanalysis. Tbe fluorescent probes constitute an
exciting new development that has received little application in conventional histo-
chemistry. This is partly due to their dependence on intact, living cells as staining
substrates.
By loading cells with the appropriate fluorescent probe in vitro, it is possible
to make dynamic quantitative estimates of diffusible ions in living cells or tissues
with a high degree of time resolution. Tbe probes, which may also be tenned
indicators or dyes, are either introduced into cells by microinjection or by incu-
bation with their respective membrane-penneable acetoxymethyl tetraesters. This
latter approach effectively traps the probe within cells since cytoplasmic esterases
produce tetra-anionic fonns that are no longer membrane-penneable (Tsien, 1981;
Grynkiewicz et al., 1985).
Tbe actual measurement of ion concentrations is based on either one of two
phenomena arising from the binding of the probe to its specific ion; these are:
436 P.E. H~yer, L. Kayser, M.R. Barer, H. Lyon
(1) a change in intensity of the emitted light and (2) a shift in the emission or
excitation wavelength. By measuring light emission instead of absorbance, high
sensitivity is achieved and ions can be measured in singlecells. The equipment
needed to measure the dynamic changes in ion concentrations at the single cell
level comprises an epi-fluorescence microscope system connected to either a pho-
tometer or a video system. An inverted microscope is preferable or alternatively
a water immersion objective can be used (Tanasugarn et al., 1984). A high level
of environmental control (temperature and C02) greatly facilitates the interpreta-
tion of results. Several investigations have been done at room temperature using
HEPES buffered media without bicarbonate. These unphysiological conditions are
to be avoided particularly when measuring intracellular pH (Thomas, 1989).
The photometer-based system is interfaced with a comput.er and this makes
it possible to collect data at regular intervals and to calculate the intracellular
concentrations based on the measured light intensity. While this system is cheaper
than the video system and offers good time resolution, it only gives information
about a single location (e.g. a cell or an organelle). The photometer approach and
apparatus are described in detail by Tsien et al. (1985) and Cobbold and Rink
(1987).
The video system is based on a very sensitive video camera (silicon intensified
target (SIT) video camera) connected to an image analysis system. This system has
a lower time resolution than the photometer system, but gives a visualization and
a localization of the changes between cells or between different points in the cello
Video set-ups are described by Arndt-Jovin et al. (1985) and Poenie et al. (1986).
Measurements are made at either single or dual wavelengths depending on the
probe used. The probes are classified as shown in Table 28.8.
Group I. Probes Changing Intensity of Light Emission on Binding to Ions. The
concentration of the unbound ion is proportional to the intensity of emitted light
from the probe after excitation of the probe at another wavelength. The intensity
is measured at a single wavelength. Examples are quin-2 and fluo-3 (Fig. 28.5) for
Ca2+, and SPQ and SPA for CI-.
Group 11. Probes Changing Excitation Wavelength. These probes change their
excitation wavelength on binding to a specific ion. The emission wavelength re-
mains the same but with an increasing concentration of the ion, the light emitted
from the unbound probe will decrease at one excitation wavelength with a con-
comitant increase in light emitted from the bound probe at the other excitation
wavelength. In principle, the concentration of the ion can be measured by register-
ing the intensity of emitted light upon excitation of the bound probe. However, the
change in excitation wavelength offers an opportunity of making a ratio between
the intensity of emitted light during alternating excitation of the cells at the two
excitation wavelengths, thus making the measurements insensitive to changes in the
concentration of the probe, e.g. by leakage or photobleaching. For measurements
with these probes more advanced equipment is necessary. The microscope must be
equipped with a computerized filterchanger or two monochromators at the illumi-
nation site with equipment synchronizing the photometer or video equipment with
28 Quantitation in Histochemistry 437
Table 28.8. Excitation and emission wavelengths of probes for the quantitative assess-
ment of diffusible ions.
Excitation Emission
wavelength(s) wavelength(s)
Probe Group nm nm References
Calcium:
BAPTA I 254/274 363 Tsien, 1980
Quin-2 I 339 492 Tsien, 1980
Fura-2 11 360/340 505-12 Tsien et al., 1985
Indo-l III 331 - 49 485/410 Grynkiewicz et al., 1985
Fluo-3 I 500-505 525 Tsien, 1988
Hydrogen:
BCECF 11 508/? 530 Rink et al., 1982
SNARF-l IV 579/518 640/587 Haugland, 1989
SNAFL-l IV 537/479 623/543 Haugland, 1989
Potassium:
PBFI 11 346/334 500 Haugland, 1989
Sodium:
SBFI 11 346/334 500 Haugland, 1989
Magnesium:
Furaptra 11 370/335 510 Raju et al., 1989
Chloride:
SPQI 320--50 445 Illsley et al., 1987
SPA 2 400 490 Krapf et al., 1988
EX= 490 NM
61.6~M
0.88
0.57
I 0.38
0.25
0.16
0.095
~..-:'---..,O .042
EM=510NM
43.5 f.lM
0.756
0.441
0.284
0.189
0.126 TTfT7-"+___......\ll
I 0 .08 1 -rrtt-r-r..""---,,,
0.047
0.021--tf/f;/i-f--;/-o./
Fig. 28.6. Excitation spectra of fura-2 at various concentrations of calcium from 0 - 43.5 11M
(11M = j1ffiol/l of Ca2+) obtained while monitoring the emission at 510 nm (EM = 510 NM). The
excitation maximum shifts from 362 nm to 340 nm on binding of Ca2+. I = excitation intensity
(arbitrary units).
(Reproduced by courtesy of Molecular Probes Inc., Eugene, Oregon).
Group IV. Probes Changing both Excitation and Emission Wavelength. The
probes in this group have the properties of both group n and In. The probes can
be used in any of the described techniques (single or dual wavelengths). Examples
of these probes are SNARF-l and SNAFL-l for determination of H+.
28 Quantitation in Histochemistry 439
Specific ßuorescent probes for ions. The fluorescent probes have been extensively
reviewed by Tsien, 1989. Commonly used probes are described below, whereas new
and not so well documented probes are indicated in Table 28.3.
Calcium. So far, for the detennination of diffusible ions, calcium probes have
probably been the most frequently used. For additional infonnation, see Cobbold
and Rink, 1987.
BAPTA. This probe, BAPTA (group 1), was the first rationally designed fluorescent
probe showing bigh selectivity for calcium (Tsien, 1980). The intensity of light
emission at 363 nm changes on binding to Ca2+. The structure of BAPTA was based
on the Ca2+-chelator EGTA wbich in turn is a derivative of EDTA (cf. Sect.15.4).
The stoicbiometry of calcium binding is 1:1 with full saturation at 1 mmol/l Ca2+
and an effective dissociation constant KcJ (cf. Sect.12.3.11) of 107 nmol/l. The
wavelengths of excitation and emission are, however, too low to give BAPTA any
practical value compared to the newer indicators, but BAPTA has served as a model
for the other Ca2+ probes.
Quin-2. The probe Quin-2 (group I) is a derivative of BAPTA (Tsien, 1980) which
emits at 492 nm on excitation at 339 nm (Tsien, 1982) and has a saturation level of
100 jlmol/l. Quin-2 is pH insensitive at pH levels above 7.05. Leakage is less than
5%/h at room temperature. Quin-2 has unfortunately two major disadvantages:
1. KcJ is low (115 nM) making Quin-2 unsuitable for intracellular Ca2+ -measure-
ment of levels above 1 jlmol/l as found in excited cells
2. Since it is the absolute intensity that is measured and not a ratio, the probe is
sensitive to leakage, photobleaching, and to differences in cellular uptake
Fura-2. This probe (Fig 28.6) has several advantages compared to quin-2 (Gryn-
kiewicz et al., 1985; Tsien et al., 1985):
1. It belongs to probe group 11, changing excitation wavelength (from 362 to
340 nm) upon binding to Ca2+ with emission at 505-512 nm. To avoid loss
of intensity in microscopes without quartz optics, measurements can be made
using excitation wavelengths at 355 nm and 380 nm (Tsien et al., 1985)
2. The intensity of emitted light is about thirty times stronger than that obtained
with quin-2
3. The discrimination between Ca2+ and other divalent ions is better
4. The KcJ is bigher (224 nmol/l) than for quin-2
Indo-l. This probe has nearly all the advantages of fura-2 with a comparable
increase in intensity (Grynkiewicz et al., 1985). The main differences between the
two probes are:
1. Indo-l belongs to probe group m changing emission wavelength on Ca2+
binding (from 485 to 410 nm) on excitation at 331-349 nm
2. Photobleaching is much faster with this probe than with fura-2 (Tsien, 1988).
KcJ is 250 nmol/l
440 P.E. HjiSyer, L. Kayser, M.R. Barer, H. Lyon
Fluo-3. This probe (Fig. 28.5) belongs to probe group 1 like quin-2, with the
same disadvantages of not using the ratio technique. Fluo-3, however, offers other
advantages such as a high K.i (400 nmol/l) and a wavelength of excitation in the
visible spectrum (500-505 nm) making it suitable for confocal scanning microscopy
(Tsien, 1988).
pH. Probes for B+ make the determination of intracellular pH possible.
BCECF. The probe BCECF, belonging to group n, is an analogue of the pB sen-
sitive 6-carboxyfluorescein (Rink et al., 1982). The latter compound is not suitable
for measurement of internal pB due to pronounced leakage. BCECF does not leak
and has a pKA (6.97) ideal for measurement of intracellular B+ concentration.
BCECF has an emission wavelength of 531 nm, and exhibits a shift in excitation
wavelength upon binding to B+ ions. Ratio measurements are possible when using
excitation wavelengths of 440 and 500 nm.
2. The assays only measure the response in the target cel1s so that it is not confused
by including non-responsive cel1s in the measurements
3. Cytochemical bioassays are extremely sensitive and are, for example, about
one thousand times more sensitive than the equivalent radioimmunoassays for
various polypeptide hormones (see Table 28.9)
4. Cytochemical bioassays measure the functional activity of molecules in contrast
to radioimmunoassays. Cytochemical bioassays do thus not determine "big-
ACTH", "big-gastrin", prohormones, or hormones that are otherwise changed
metabolical1y or configurationally
On the other hand, cytochemical bioassays are not suitable for large scale
screening assays as they are more time-consuming than radioimmunoassays. More-
over, these assays are very demanding on the skills of the investigator.
The advantages have nevertheless led to the World Health Organization recom-
mending that these techniques be the international microassays of choice (World
Health Organization, 1975).
Extensive reviews of the cytochemical approach to hormone assays are given
by Chayen (1978; 1980).
Other examples of segment microassays include: The effect of pentagastrin
and plasma gastrin on carbonic dehydratase in parietal cells of the guinea-pig
fundus (Loveridge et al., 1974); the influence of physiological or subphysiolog-
ical concentrations of corticotropin on 3ß-hydroxy-.6.5-steroid dehydrogenase in
cells of the fascicular and reticular zones of the adrenal cortex (Loveridge and
Robertson, 1978); determination of the concentration of a sodium transport in-
hibitor in human plasma and urine by measuring the stimulation of renal glucose-
6-phosphate dehydrogenase or the inhibition of renal Na+, K+ -transporting ATPase
(Fenton et al., 1982); the acute effect of physiological concentrations of thyroid
stimulating hormone on ß-galactosidase, N-acetyl-ß-glucosaminidase, and leucyl-
ß-naphthylamidase activities in thyroid follicular cel1s (perrild et al., 1989).
The concentration of certain immunoglobulins, the so-called thyroid growth
stimulating immunoglobulins that occur in Grave's disease have also been investi-
gated using the cytochemical bioassay technique (Bitensky et al., 1974; McKenzie
and Zakarija, 1977; Drexhage et al., 1989).
In principle, not only segment assays and section assays, but also cultures of
single cells, may potentially form the basis for assays for investigating the effect
of any stimulating or inhibiting compound, such as hormones, drugs, toxic agents,
cytokines, growth factors, or antibodies.
In conclusion, we agree with Bitensky and Chayen (1978), that quantitative
cytochemistry is now a highly precise and remarkably versatile way of examining
changes in cellular biochemistry induced by biologically active substances. It is
an extension of conventional biochemistry to the cellular level which avoids many
of the statistical problems inherent in conventional biochemistry; moreover, the
technique is non-disruptive.
Table 28.9. Examples of microbioassays.
~
Peptide hormone Biological function tested
assayed Target cells by quantitative cytochemistry Sensitivity' References
Corticotropin zona reticularis cells of adre- depletion of ascorbate determined by 5 fgjmI Chayen et al., 1972; Loveridge
nal cortex Prussian blue or silver nitrate reac- et al., 1975
tion
Luteinizing hormone corpus luteum cells of ovary depletion of ascorbate determined by 10- 4 mUjmI Rees et al., 1973; Buckingham
Prussian blue reaction et al., 1979
Thyroid stimulating hor- follicle cells of thyroid lability of lysosomal membrane de- 4 x 10 - 5 JlU jmI Chayen et al., 1980
mone termined by testing the permeabil-
ity for the substrate used for
measuring leucyl-ß-naphthylami-
dase activity
Gastrin parietal cells of stornach production of H + indirectly meas- 5 fgjml Loveridge et al., 1978; Lover-
ured by carbonate dehydratase idge et al., 1980
Parathyroid hormone cells of distal convoluted tu- transport of Ca2+ indirectly deter- 5 fgjmI Chambers et al., 1978; Golt-
bules of kidney mined by glucose-6-phosphate de- zman et al., 1980
hydrogenase activity
"0
Corticotropin releasing cells of the anterior pituitary corticotropin liberated Buckingham and Hodges, 1977
hormone
in
Thyrotropin releasing cells of the anterior pituitary thyroid stimulating hormone Iiber- 5 fgjmI Gilbert et al., 1977
hormone ated fr
, According to Bitensky and Chayen (1978) and Chayen 1980.
~
~
s::
~
t:Il
~
;:r:
~
g
29 Autoradiography 443
29 Autoradiography
29.2 Application
29.3 Isotopes
Most elements (E) occur as several isotopes each of which possess the same chem-
ical properties due to their identical complement of electrons but different physical
properties, due to the differences in their nuclei.
An element is defined by its atomic number (Z) which corresponds to the num-
ber of positively charged protons in the nucleus. The element is also characterized
by a mass number (A), which comprises the sum of the protons and the uncharged
neutrons (N) within the nucleus. Neutrons have the same mass as protons.
Atoms with the same atomic number but different mass numbers are described
as isotopes of the element concerned. If the number of neutrons is different from
the number of protons, the isotope may be unstable and radioac-tive. Sooner or later
such unstable isotopes transform into more stable isotopes by emission of ionizing
radiation (a-, ß-, or ')'-radiation). The stable and most common isotope of carbon
possesses a mass number of 12, formally written ~2C (~E). The atomic number
is often omitted because C alone specifies carbon. When denoting radioactive iso-
topes, however, the mass-number (A) is always included (e.g. 14C or sometimes
C-14).
The type and the energy of the radiation is different for each isotope. Only some
forms are suitable for autoradiography.
a-Radiation. This comprises helium atomic nuclei ~He2+. All a-particles from
a particular isotope are emitted with the same energy (monoenergetic radiation).
Traces made by a-particles in a photographic emulsion appear as lines of identical
length. This form of radiation is rarely used for autoradiography because it is
only emitted by some heavy elements that are of litde or no interest in biological
systems.
29.3.2 Half-Life
Sooner or later unstable atoms transform into stable atoms by the emission of ioniz-
ing radiation. Although, apriori, it is not possible to calculate when emissions will
occur, the time taken for half the atoms to decay into stable elements can be deter-
mined. This time is called the half-life (Tl/2) of the element and is characteristic
for each isotope. The half-life is of 14C is very long (Tl/2 = 5,730 years) while that
of 3H is shorter (T1 / 2 = 12.3 years). Suitable isotopes for autoradiography do not
have short half-lives. Table 29.1 reviews the characteristics of some biologically
important isotopes.
Consideration must be given to both the physical and histological aspects of the
technique:
1. The isotope must be retained and remain at the same tissue location during the
histological procedure (fixation, dehydration, embedding, etc.). Several small
hydrophilic molecules are extracted during these processes and their detection
requires special procedures
2. It is essential to determine whether the radioactive isotope remains attached
to the molecule being studied or whether it can be incorporated into another
metabolite. Isotopes may also detach from the target molecule and bind to other
molecules or lie free in the section. Controls can be performed to identify these
pitfalls but their exact nature varies depending on the labelled molecule of
interest
3. Sections should not contain substances that either sensitize the photographic
emulsion directly (positive chemography) or that interfere with the sensitized
bromide crystals in the emulsion so that they cannot be developed (negative
chemography). (cf. Sect.29.7.2)
After tissue sections containing the radioactive isotope have been cut and placed
on gelatinized glass slides, they must be de-waxed and placed in contact with an
emulsion. The application of the photographic emulsion can be performed in two
ways:
Emulsions can also be made as gels. The gel is melted by heating to around 40°C
and the emulsion diluted with an equal amount of distilled water. The glass slides
29 Autoradiography 447
bearing the labelled sections are dipped into the melted emulsion and placed on a
horizontal table to dry. The thickness of the emulsion is dependent on the degree
of dilution and the temperature of the emulsion. After drying the emulsions are
exposed as in the stripping-film technique.
29.5.3 Emulsions
During development the silver bromide crystals in the emulsion are reduced to
metallic silver. As shown in Fig. 29.1 it is important to control the temperature
and the development time in order to obtain the optimum balance between the
development of grains initiated by specific radiation and the background.
29.5.5 Staining
A number of histological stains can be used after the autoradiographs have been de-
veloped. Stains have to pass through the gelatin layer of the emulsion and problems
448 M. M~ller, I.M. Krogh
s
I
I
+ I
I "
I
I
I
I
~//
... - .. - ------------'
o 2 3 4 5 min
Fig. 29.1. The kinetics of development of specific grains compared with the background. Number
of silver grains developed (S), development time (min).
(- - -) number of specific grains
( - - ) number of background grains
(U) optimum time for development
may arise with acidic reagents which may remove silver grains and with staining
of the gelatin itself.
It is also possible to perform enzyme reactions through the gelatin layer but
incubation times have to be prolonged. Eosin can be recommended for routine
counterstaining because careful washing in tap water removes the stain from the
gelatin before it is washed out of the section.
~I--- Emulsion
Grid
000
00 +r---Grid
-+---Emulsion
or by dipping the sections in the emulsion. In the latter method the grids are
placed on a collodion coated glass slide and the slide and attached grids are dipped
into the emulsion (Fig. 29.3).
The autoradiographs are developed in light-tight boxes at 4°C containing a
desiccant such as silica gel. The exposure times are much longer than for light
microscopic work. This is due to the extreme thinness of the sections (80-100
nm) which contain very small amounts of isotope. The autoradiographs can be
developed for 5 min in a D-19 developer at 20°C and fixed with thiosulphate as
in the light microscopical procedure. After fixation sections can be contrasted with
lead and uranyl acetate.
29.6 Resolution
The emissions from a radioactive isotope are random with respect to time and
direction. Depending on the exposure time, each source produces a number of
grains at different locations in the emulsion with the isotope at the centre. For a
ß-emitting source the distribution of grains would appear as shown in Fig. 29.4.
50%
Fig. 29.4. Grain density (D) distribution as a function of distance (abscissa) from a radioactive
source (x). Resolving power is the distance (a) from source to 50% peak grain density.
The curve shows a nonnal distribution within which the resolution of the system
is defined as the distance (a) from the source at which the grain density falls to
half of that detected over the source itself.
Because the resolution depends on several factors (isotope, grain size of the
emulsion, thickness of the emulsion, developer) a related value based not on grain
density but on total number of grains called the half radius (HR) is used instead
(Williams, 1977, part n, pp.89-94). In a specific autoradiograph 50% of the grains
generated by a point source should be located within a distance indicated by HR.
The HR for 3H detected with an llford LA emulsion and developed in Microdol X
is 280 nm. For 14C the HR is 390 nm.
29 Autoradiography 451
29.7 Artifacts
In every autoradio graph some of the grains are not caused by the presence of the
isotope in the section. The reverse problem may also occur (Le. sensitized silver
bromide crystals disappearing or failing to give rise to silver grains by the end
of development). Some of these artifacts will be discussed further below. They
are, however, often unpredictable and careful autoradiographic controls are alwilYS
required.
29.7.2 Chemography
Positive Chemography. This is caused by components of the section other than the
specific isotope. These are often reactive molecules with reducing groups which can
direcdy sensitize the crystals in the emulsion. The process is temperature dependent
and can be decreased by exposing the emulsion at low temperatures. It is also
possible to decrease the chemography by placing a layer of evaporated carbon
between the sections and the emulsion.
Nuc1ear emulsions are sensitive to pressure and scratches and even fingerprints
can produce grains. Lateral stress due to shrinkage during drying can also produce
non-specific grains.
29.8 Quantitation
All countries have strict regulations with regard to storage and handling of radionu-
clides. These regulations generally follow the those laid down by the International
Atomic Agency in Vienna (IAEA) in their booklet "Safe Handling of Radionu-
clides".
Radionuclides can be classified into four groups of radiotoxicity:
a. Very high
b. High
c. Moderate
d. Low
Isotopes used for autoradiography mosdy belong to group d. Tritium and 14C
present absolutely no radiation hazard, however, 32p and 125 1 are in group c. Because
125 1 can be concentrated in the thyroid gland this isotope should be handled in a
fume hood. For further reading the booklet from IAEA is recommended.
30 Fillorescence Microscopic Methods in
Histochemistry
M. Mt/Jller, H. Lyon
30.1 Autoßuorescence
The majority of tissue components in unstained untreated sections show some de-
gree of ftuorescence. This is called primary ftuorescence or autoftuorescence. Such
autoftuorescence is particularly pronounced in plant tissues, while in animal tissue,
collagen, elastin, and lipofuscin are noted for this property. Collagen and elastin
(Sect21.5) show a blue-green ftuorescence, while lipofuscin (Sect18.2.2) gives an
orange ftuorescence. The red autoftuorescence of porphyrins (Sect18.2.1) may be
useful in diagnostic work (Sect31.6.2). A number of drugs also ftuoresce and it
is therefore possible to follow their fate in tissues using ftuorescence microscopy.
Examples include tetracyclines (Sect.15.9) and Acridines.
Background autoftuorescence can be a nuisance in catecholamine work (Sect.
30.2) and immunoftuoresence (Sect.26.2.1). In these circumstances Cowen et al.
(1985) suggest the use of Benzo Sky Blue (pontamine Sky Blue 5BX, C.!. 24400)
as a counterstain. This reduces autoftuorescence.
Biogenic amines can be divided into catecholamines and indolamines. The most
important representatives of the catecholamines are adrenaline and noradrenaline
which are localized to the nervous system and the adrenal medulla. Dopamine,
which is found in the nervous system, may also be included in this group. The
H. Lyon (Ed.)
Theory and Sttalegy in Histochemistty
@ Springer Verlag 1991
454 M. M01ler, H. Lyon
most important indolamine is serotonin which is found in the nervous system and
in enterochromaffin cells (SectI8.1.4). The formaldehyde induced ftuorescence
method (FIF) was first described by Eränkö (1955; 1967) and further developed
by Falck and coworkers (1961; 1962). The method is very sensitive and highly
specific. Due to the very high solubility of biogenic amines, it is usually necessary to
perform the method on freeze-dried tissue (Sect.l1.3.2). It is possible, particularly
wfth serotonin, to obtain good results with standard paraffin-embedded material
(Lyon et al., 1982) or even better, material embedded in hydroxyethylmethacrylate
(Lyon et al., 1981) (Sect.14.4.2).
a b
HOJQQR'
R'
HO~ HoJQQ
HO
J0 ~H2 o o HO
NH
HO
bN
R" Rn
R"CHO
The condensation is usually performed with formaldehyde (R" = H), but can
also be performed with vapours of acetaldehyde, glutaraldehyde, formic acid, and
acetic acid all with excellent results.
With regard to adrenaline the dehydrogenation (autooxidation) (a-+b) requires
far more energy (Jonsson, 1967) than dopamine or noradrenaline.
For serotonin the ftuorophore formation occurs in two steps which are similar
to the steps for noradrenaline.
30 Fluorescence Microscopic Methods in Histochemistry 455
HO~H2
l8J--NH;J ~
~HCHO
HO~H
l8J--NH~~
~ -2H
HO~
l8J--NH~
Excitation Emission
Biogenie amine (nm) (nm)
Mechanism. Glyoxylic acid reacts with biogenie amines as seen in the following
scheme with noradrenaline as an example.
30 Fluorescence Microseopie Methods in Histochemistry 457
OH
HO~
HO~ ~H2
HOOCCHO ~
OH
HO~H
HO~~
COOH
HOOCCHO ~
HO~'-CH'COO
HO~~
1~ COOH
OH
O~_CH'COOH
HO~~
COOH
Sensitivity. The glyoxylic acid method is even more sensitive than the FlF-method,
noradrenaline can be detected down to a concentration of 10-7 pmol/l.
The o-phthalaldehyde method (OPT) was first described by Häkanson et al. (1970).
Deparaffinized seetions are treated with the vapour of o-phthalaldehyde.
458 M. Mlilller, H. Lyon
N::;:::'\
~NH
CH z rQYCHO
I
CHzNH z ~CHO
histamine o-phthalaldehyde
Mechanism.
a b OH
:~H'
HO:©n
HO 0 NH z
Hom
~ ~
,
H0lQ()
HO 0 NH o
HO
NH
Hom HOOO
~ ~
"
HO©:)
HO 0 .&N
o
HO ~N
-HP
~O
0 ·...-;::N
0&
1~ 1~
0:CO
HO:::--""
'"
:::--.... NH HO:::--"" :::--.... NH
/H
R-N -CH 2CH 2CI
"'--CH 2 CH 2CI
2CI-
Dopamine (a) and noradrenaline (b) fonn the isoquinolines I and IV on conden-
sation with HCHO. Dehydrogenation leads to the fluorescent dihydro-derivatives
in their quinone fonn III and VI with excitation maxima at 410 nm. Short treatment
with HCI-vapour leads to fonnation of the non-quinone fonns 11 @d V with exci-
tation maxima at 370 nm. Further treatment with HCI leads to the transfonnation
of V to VII with an absorption maximum of 320 nm; 11 cannot react in this way,
and it thus possible to differentiate ßetween dopamine and noradrenaline.
This means staining with fluorescent dyes. Dyes may be subdivided into diachromes
and jluorochromes (Sect.26.2.1) depending on whether they can seen by ordinary
light or fluorescence microscopy. Many diachromes are also fluorochromes, e.g.
Congo Red, Eosin, and Pararosanilin.
Fluorochromes are extremely sensitive stains. Compared to non-fluorescent
dyes, very small amounts of dye can be demonstrated by fluorescence microscopy.
Fluorescent dyes are very useful for screening procedures since the fluorescence is
read against a dark background. Specific staining can often be rapidly detected at
low magnifications even by inexperienced workers (Wachsmuth, 1988).
Staining can be quantitated using microfluorimetry (Chap.28).
Some examples of direct fluorochromy are given below.
The antimalarial drug Atabrine (Quinacrine, Mepacrine) and the cIosely related mi-
tosis inhibiting substance Atabrine Mustard (Quinacrine Mustard) are fluorochromes
which stain chromosomes in mitotic cells and give rise to highly characteristic
bands which make it easier to identify the individual chromosomes.
Staining is usually perfonned on material which has been pretreated with
trypsin. Q-bands and QM-bands are referred to depending on which fluorochrome
has been used. Similar results can be obtained with the Giemsa-method, G-bands.
460 M. Mpller, H. Lyon
It is possible to "reverse" the bands, so-called R-bands (dark bands become light
and vice versa) by denaturation (e.g. with a strong salt solution or heat prior to
staining).
30.3.2 Amyloid
Supravital staining must be performed with very low dye concentrations to avoid
toxic effects. There are therefore considerable advantages in using fluorochromes
such as Neutral Red, Acridine Orange, Congo Red, and Evans Blue for this purpose.
Enzyme activity in cells can be detected when the enzyme causes a change in fluo-
rescence by its action on a substrate or a coenzyme. The redox kinetics of enzyme
systems in which pyridine nucleotides participate can be followed using the fluo-
rescence of NADH. In the case of hydrolases it is possible to use the fluorescence
of the enzyme product (naphthol derivative) instead of coupling reactions. This
makes increased sensitivity and optical specificity possible and facilitates microflu-
orimetric quantitation (Chap.28).
Part 7
An Introduction to Applied Histochemistry
31 Applied Histochemistry - An Overview
The morphological evaluation of cells and tissues requires that key structures are
demonstrated by staining. The chosen methods should give reproducible results
irrespective of the source of material, manner of fixation, or thickness of the prepa-
ration.
The most frequently used general oversight stains are "Haematoxylin"-Eosin,
Romanowsky-Giemsa, and Papanicolaou. For oversight staining of plastic embed-
ded material Toluidine Blue 0 may be used.
lL Lyon (Ed.)
Theory and Strategy in Hi8lOchemistry
@ Springer Verlag 1991
466 H. Lyon, E. Schulte, J. Visfeldt, E. Hasselager
31.2.1 ''Haematoxylin''-Eosin
The basic principles for the demonstration of ionized or ionizable groups are dis-
cussed in Chap.6 where also the definitions for acidophilia and basophilia are given.
31.3.1 Basophilia
Basophilic Granules. These are found in mast cells and in the basophils in blood.
They chiefly contain a proteoglycan in which heparin forms apart (Sect.2.1.5). This
is of importance in identification as the granules are stained metachromatically even
at low pH (particularly if preceded by deamination). Mast cells and basophils are
encountered in increased amounts in inflammatory processes, particularly acute,
allergic reactions. It should be noted that heparin is soluble in alcoholic fixatives.
Therefore, in cytological specimens (blood and bone marrow smears) dye-fixation
as in Giemsa stains is preferable to fixation in pure methanol (cf. Appendix A).
Virus Particles. These may be found as basophilic inclusion bodies in for instance
warts.
31.3.2 AcidophiIia
Induction. The cytoplasm of liver cells can show increased acidophilia on induc-
tion. This results from exposure to various metabolites and exogenous compounds
and involves an increase in the amount of smooth endoplasmatic reticulum bearing
the enzymes which decompose and/or detoxify these molecules.
Paneth Cells. These cells in the intestinal mucosa contain coarse acidophilic gran-
ules. These consist of the strongly basic, arginine rich enzyme lysozyme (mu-
raminidase) which is able to degrade the material of the cell wall surrounding
Gram positive bacteria (Sect.2.3.1).
Zymogen Granules. These acidophilic structures are seen in areas such as the
apical part of the exocrine gland cells of the pancreas.
Acidophilic (Eosinophilic) Bodies. These are seen in acute viral hepatitis and
probably represent dead cells.
Oedema Fluid. With a sufficiently high content of protein, oedema fluid can be
seen as an extracellular, light, acidophilic material.
Fibrinoid. This stains as fibrin and is seen in different immune reactions such as
in rheumatoid nodules and in the vessel wall in polyarteritis nodosa.
Viral Inclusion Bodies. These can be acidophilic, for instance molluscum bodies
in molluscum contagiosum or Negri bodies in the cytoplasm in rabies.
31.4.1 Bacteria
The structure and chemical composition of bacteria is dealt with in Sect.2.3. The
Gram and Ziehl-Neelsen methods are considered in Sect.6.1.6.
A number of other methods for the demonstration of mycobacteria have been
described. Fluorescence microscopy of mycobacteria is possible using Auramine 0
and Rhodamine B (red-golden) or Thioflavine TG (blue-green fluorescence) in the
acid fast staining method. The advantage of these methods compared with the
classical Ziehl-Neelsen method is that the acid fast organisms are seen strongly
fluorescent on a dark background and are therefore easily demonstrated using
dry objectives (25x or 40x) with resultant larger visual field and time saving.
470 H. Lyon, E. Schulte, J. Visfeldt, E. Hasselager
Bartholomew (1981) insists that the resu1t should be confirmed with a standard
Ziehl-Neelsen staining.
31.4.2 Fungi
The pathologist may wish to confirm the presence of calcium in lesions suggestive
of dystrophic or metastatic calcification (Sect.17.7.1). With the exception of iron
(cf. Sect.31.6.3), demonstrations of other forms of metallic deposit such as copper
are rarely performed Details are given in Chap.17.
31.6.1 Haemoproteins
31.6.2 Porphyrins
31.6.6 Lipofuscin
Lipofuscin (Sect.18.1.2) may be seen in the Leydig cells of the testis and in tumours
derived from these cells. Nerve cells contain increasing amounts of lipofuscin with
age as do several cell types undergoing atrophy (e.g. heart musc1e cells).
Lipofuscin can be demonstrated by its orange autofluorescence (Sect.30.2) and
its acid fast basophilia (Sect.l8.4.1). In case of a doubtful reaction the Sudanophilia
of lipofuscin may be of value (Sect.l9.6.1), as this is preserved even in paraffin
embedded material. Both Sudanophilia and PAS positivity decrease with the age
of the lipofuscin.
31.6.7 Melanin
Melanin (Sect.18.1.3) is a reducing agent and very small amounts can be demon-
strated using the ferric ferricyanide method (Sect.8.3.1). With slightly larger
amounts, the ferrous ion uptake method of Lillie is a valuable, highly selective
method (Sect.18.4.9). The ability to produce melanin (e.g. in malignant melanomas)
can be demonstrated with the DOPA reaction (Sect.l8.4.13) (cryostat sections of
fresh or formaldehyde fixed material are necessary). Immunohistochemical tech-
niques can also be useful in the diagnosis of malignant melanomas.
31.6.8 Serotonin
31.6.9 Catecholamines
31.6.10 Asbestos
31.6.11 Silica
31.6.12 Urate
Crystals of monoso,dium urate occur in deposits, in and around joints and tendons
and several other sites, in patients with gout. Although monosodium urate is only
slightly soluble in water, losses during tissue processing involving aqueous solu-
tions may be extensive unless the urate is bound to protein. Identification can be
achieved by demonstration of birefringence and by the argentaffin reaction with
Gomori's silver methenamine method (Sect.8.3.2).
The fonnation of fat droplets in cells (fatty degeneration) is usually due to the ac-
cumulation of triglycerides. The droplets are usually seen as empty vacuoles in the
cells in paraffin sections. Frozen sections are required for a positive demonstration
using lysochromes (Sect.19.6.1).
The degeneration of myelin can be investigated using the Marchi methods
(Sect.19.6.5). The use of OS04 with another oxidant enables differentiation between
hydrophobie (black-stained) and hydrophilic lipids. In degenerative conditions (de-
myelination) the lipid content is more hydrophobie.
In general, methods for the demonstration of lipids are only occasionally useful
in the diagnosis of lipid containing tumours. This is due to the fact that many
malignant soft tissue tumours give a positive reaction, while some liposarcomas
give a negative reaction.
31.7.3 Lipidoses
Lipid storage diseases or lipidoses are familial disorders of lipid metabolism char-
acterized by enzyme defects (Sect.31.11.2). They are classified according to the
nature of the accumulated lipids.
For example the diagnosis of type B Niemann-Pick disease can be made on bone
marrow aspirates or liver biopsies where the defect in sphingomyelinase activity
leads to abnonnal deposition of sphingomyelin and cholesterol. With this and other
lipidoses detennining the exact distribution and nature of the lipid deposits enables
an accurate diagnosis.
Sphingomyelin is stained weakly with Sudan Black B (Sect.19.6.1) and ap-
pears red in polarized light. In the central nervous system the sphingomyelin
stains blue with the chromation-acid Haematein reaction or NaOH-acid Haematein
(Sect.19.6.7) reaction. Cholesterol is demonstrated with the PAN method (Sect.
19.6.11) and gangliosides can be demonstrated with the modified PAS reaction, but
the borohydride-periodate-Schiff method (Sect.19.6.9) of Buk and Bayliss High
(1986) is probably better.
The amount of DNA present in the nucleus changes with the cell cycle. In mam-
mals, for rapidly dividing cells, the cycle lasts about 24 hours; the GI-phase (G =
gap) takes about 10 hours (lowest DNA content), in the S-phase (S = synthesis),
which lasts about 8 hours, there is a gradual increase in DNA to twice the GI
level and this level then remains constant through G2 (4 hours) until-the end of the
M-phase (M = mitosis) which takes about 2 hours. In more slowly dividing cells
the GI-phase is prolonged.
Pathological variations include malignant ceIls where abnonnal numbers of
chromosomes may occur (aneuploidy) and where the amount of DNA per chromo-
some is often elevated.
The staining properties of DNA vary according to its physicochemical state. For
example DNA in condensed chromatin (non-S-phase) is hydrolyzed more slowly
by Feulgen hydrolysis (Sect.9.9) than newly synthesized DNA (S-phase, partially
476 H. Lyon, E. Schulte, J. Visfeldt, E. Hasselager
condensed) so that in cell smears the optimal hydrolysis times are 60 and 20
min respectively. This sort of differential characteristic can be used to assess the
content of different forms of DNA and may be useful in detecting neoplastic and
preneoplastic cells where, in addition to the normal forms of DNA, small amounts
of a more labile form with an optimum hydrolysis time of 3-5 min are found
(Husain, 1983).
The amount of ribosomal RNA in a cell usually parallels protein synthesis (Sect
2.2.2). Thus, large amounts of RNA are found in the cytoplasm of cells producing
proteins for export, such as liver cells (albumin), glandular cells (enzymes and
mucus), and plasma cells (antibodies).
In haemolytic anaemias, where the need for new red blood cells is high, an
increased number of young erythrocytes (polychrome erythrocytes or reticulocytes)
is found in peripheral blood. These cells have, in contrast to mature erythrocytes,
rather large amounts of RNA in their cytoplasm as haemoglobin synthesis is still
taking place.
Polychrome erythrocytes give a faint bluish colour with the Romanowsky-
Giemsa stain. Supravital staining with the oxazin dye Brilliant Cresyl Blue, yields
blue granules and threads in the cytoplasm.
Table 31.1. Methods for the demonstration of specific amino acid residues.
The mucous secretory cells of the nonnal gastric epithelium almost exclusively
contain neutral mucins, while traces of sialomucins and perhaps sulphomucins can
be found in the mucous neck cells of the fundus. Sialomucins and sulphomucins are
sometimes also found in the antrum and in the glands of the oesophageal-gastric
junction.
Table 31.3. The results of some carbohydrate reactions on the mucosa of normal stornach and
stornach with intestinal metaplasia.
intervening
goblet cells epithelial cells
Mucous type type
membrane
Method normal 11 III 11 III
PAS + + + + + + +
Alcian Blue pH 3 0 + + + + +/0 +
Alcian Blue pH 5.7 +0,2 mol/I MgCI 2 0 0 0 + 0 0 +
Alcian Blue pH 5.7 + 0.7 mol/I MgCI 2 0 0 0 +/0 0 0 +/0
Periodic acid /borohydride / potassium 0 + 0 0 0 0 0
hydroxide/PAS
+ = positive reaction.
+ /0 = positive or negative reaction.
o = negative reaction.
Goblet cells in the lower two-thirds of the colonic crypts predominantly secrete
sulphomucins, while the goblet cells in the upper one-third and surface epithelium
secrete N- and especially O-acylated sialomucins (Filipe, 1983).
Acid mucins are rarely present in prostate gland without neoplastic changes. Con-
versely, these mucins are found in two thirds of prostatic carcinomas. Tbey are
usually not present in the most poorly differentiated tumours (Jöbsis, 1983).
In the breast both neutral mucins and sialomucins are found in the epithelial cells
in normal tissue, in fibrocystic disease, and in carcinoma (Ormerod and Sloane,
1983). In carcinoma of the breast a combined Alcian Blue pB 3-PAS reaction will
frequently demonstrate large globules with a blue rim and a magenta spot in the
480 H. Lyon, E. Schulte, J. VIsfeldt, E. Hasselager
middle, so-called bull's eye appearance. These vacuoles are particularly associated
with in situ and infiltrating lobular carcinoma.
Table 31.5. The histochemical characteristics for myxoid tumours and chondrosarcomas.
31.10.8 Glycogenoses
Cell organelles may be selectively demonstrated at both the light and the EM level
by the presence of certain "marker" enzymes (Novikoff, 1976). Typical examples
are given in Table 31.6.
482 H. Lyon, E. Schulte, J. Visfeldt, E. Hasselager
propria and muscularis mucosae are more numerous and often also increased in
size (Lake, 1983).
Table 31.7. Methods for histochemical assessment of musc1e biopsies (according to Johnson, 1983).
Method Purpose
Striated muscle comprises approximately equal numbers of type 1, type 2A, and
type 2B fibres. Normally, there should not be more than 3% type 2C in adults. The
random distribution of fibre types in adults corresponds to the random distribution
of the muscle fibres from different motor units. The fibres from one motor unit
all belong to the same type, but as the motor unit territories overlap each other, a
mosaic is formed.
Neurogenie Muscle Diseases. In these conditions, where the most prominent patho-
logical change is denervation, there are many common characteristics regardless
of where in the nerve the lesion is localized (central or peripheral). If only a few
motor units are denervated the distribution of the atrophied muscle fibres is scat-
tered. The atrophied muscle fibres include both type 1 and type 2 fibres as there is
generally no selective effect of the denervation process on either type 1 or type 2
motor units. If, however, the denervated fibres are reinnervated by branches from
one single axon the result will be a group of fibres of the same histochemical type.
This phenomenon with grouping of fibres of the same type is a clear indication of
reinnervation in neurogenic muscle diseases.
Quite frequently biopsies from well-compensated (i.e. reinnervated) disorders
do not show any signs of fibre atrophy, and the diagnosis can only be made by
the histochemical demonstration of fibre type grouping. Durlng the process of rein-
nervation fibres change from one type to another if the reinnervating axons are
of a different motor type from the original axons. Fibres, which are undergoing
change from type I to type 2 or vice versa, pass through a transition stage. This
31 Applied Histochemistry - An Overview 485
Investigations of bone marrow and peripheral blood smears are based on staining
with the Romanowsky-Giemsa method (Sect.6.3.2). Supplementary histochemical
staining methods can facilitate differentiation of the haemopoietic cells and are
of particular value in the subgrouping of leukaemias. These methods inelude Su-
dan BIack B, Neutral Red, PAS, Toluidine BIue metachromasia, and reactions
for myeloperoxidase, a-naphthylacetate esterase, chloroacetate esterase, alkaline
phosphatase, and acid phosphatase (Hayhoe and Quaglino, 1980). It is important
to perform the staining reactions as soon as the specimens have been prepared.
Table 31.9 shows the results expected for the above methods applied to normal
haemopoietic cells (after Smith, 1983).
The histochemical classification of leukaemias is beyond the scope of this chap-
ter. While the techniques comprising Table 31.9 are all useful, immunohistochem-
ical elassification schemes are now pre-eminent in this field (cf. Sect. 32.5.5).
~
;.
~
~
f
f
31 Applied Histochemistry - An Overview 487
32.1 Introduction
The basic prerequisites for obtaining reliable staining results with immunohisto-
chemical staining methods are the use of fixation and processing methods of the
specimen compatible with optimal preservation of the antigen under investigation
(Sect26.6).
As fixation in routine diagnostic pathology is usually carried out with neutral
buffered fonnaldehyde solution, it is extremely important to know whether the
antigen under investigation can withstand this process. Although many antigens
have been shown to tolerate routine fixation and processing, some do not (e.g.
H. Lyon (Ed.)
TheoIy and Sttategy in Histochemistry
@ Springer ~lag 1991
490 M. Vyberg, P.P. Clausen
Electron microscopy
The choice of primary antibodies is another important factor. In recent years huge
numbers of commercial antibodies have been released on the market, and it is often
very difficult to select from different antibodies to the same antigen.
As there is no general agreement on quality specification, individuallaboratories
have to make their own pilot studies in order to select the most suitable antibodies
with respect to economy, specificity, and lack of non-specific staining and cross-
reactivity.
Qnly very few antigens can be considered as cell or tissue specific. The diagnostic
conclusion drawn from an immunohistochemical staining should therefore not be
based on only one staining reaction but on the results obtained using a panel of
positive and negative reactions which will increase the diagnostic specificity of the
staining results. In this context it is of great importance to correlate the staining
results with the histomorphological picture, and with results obtained using other
histochemical staining methods and/or electron microscopy.
In the following sections of this chapter the most common areas, in which
immunohistochemical staining methods are used, will be summarized. Following
a short description of the pathological problems, the typical staining profiles will
be presented in tables.
vase ulitis is an essential element. A survey is given in Table 32.2 of the most im-
portant reaction patterns in skin disorders in which immunohistochemical demon-
stration of immunoglobulin and complement is especially important Deposits may
be encountered in conditions other than those shown in the table, thus deposits of
C3 in the basement membrane in urticaria, IgM intercellularly in the epidermis in
pemphigoid, lupus erythematosus, and lichen planus as weIl as in elderly people
without skin disease. In connection with diseases accompanied by vase ulitis de-
posits of IgG, IgM, and complement are seen. Granular deposits of IgA in vessel
walls is a characteristic reaction in allergie purpura (Henoch-Schönlein).
Table 32.2. Typical reaction patterns in different skin disorders with deposits in the epidermis,
basement membrane, and/or dermal papillae. One or more reactions may be aQsent. In the basement
membrane granular deposits are seen in lupus erythematosus and dermatitis heq,etiformis, while the
deposits are linear in the other disorders.
1 Pemphigus erythematosus.
2 Reaction mayaiso be seen in skin outside the lesion.
3 Usually massive deposits in affected skin in discoid lupus erythematosus; in systemic lupus
erythematosus deposits mayaiso be seen in unaffected skin exposed to light in two thirds of the
cases.
Disorder Antibodies to
[ Thyrotrophin-receptors.
Disorder Antibodies to
Deposits of ~
~
~
Diagnosis IgA IgG IgM C3 Fibrin
Grading of reactions: « +» occasional weak; ( + ) occasional moderately positive; + moderately positive; + + strongly positive;
MM: mesangial matrix; SL: sclerotic lesion; BM: basement membrane; GR: granular; L: linear; SE: subepithelial; SEn: sub-
endothelial; CR: crescentic.
~
496 M. Vyberg, P.P. Clausen
Keratin Leukocyte
low/high common
molecular weight Vimentin antibody S·lOO
Thyroglobulin. Carcinomas derived from follicle cells of the thyroid can be specif-
ically characterized by the presence of thyroglobulin, a protein with a mol.wt. of
670 kDa. In addition to the more highly differentiated papillary and follicular
thyroid carcinomas thyroglobulin can be demonstrated in a number of anaplastic
thyroid carcinomas.
Thmours Derived from Endocrine Glands. These can be subdivided on the basis
of their production of hormones. Epithelial cells belonging to the diffuse neuroen-
docrine system (DNES) can be characterized, partly by staining with general neu-
roendocrine markers such as neuron specific enolase (NSE), chromogranin (CG-A),
and synaptophysin (SY), and partly by demonstration of their polypeptide hormone
production.
Germ Cell Thmours. These comprise seminoma (in the ovary dysgerminoma)
and non-seminomatous tumours. Seminoma/dysgenninoma are characterized by a
positive reaction for placental alkaline phosphatase.
Thmours Derived from Gonadal Stromal Cells. In the ovary and testis these
tumours can be immunohistochemically c1assified by their hormone production. In
the ovary oestradiol and progesterone can be demonstrated in granulosa and theca
lutein cells. In the testis oestradiol and testosterone can be demonstrated in Sertoli
cells and oestradiol (progesterone) and testosterone in Leydig cells.
Thmours Derived from the Surface Coelom Epithelium of the Ovary. These are
characterized by positive reactions for keratin (Sect.32.5.1) and EMA (Sect.32.5.2),
but only seldom by reactions for PLAP, HCG, and AFP (Table 32.7).
Seminoma/d ysgerminoma + 01 0 0 0
Embryonie carcinoma +/0 01 + + 0
Volk sac tumour +/0 01 + 0/+ 0
Choriocarcinoma + + 0 + 0
Non-germ cell carcinoma 0/+ 0/+ 0/+ + +/0
1 Positive in syncytiotrophoblast like cells.
+ /0: usually positive; 0/ + : usually negative.
Antibodies to:
PLAP: placental alkaline phosphatase; HCG: human chorionic gonadotrophin; AFP: r:J.-
foetoprotein; KER: keratin; EMA: epithelial membrane antigen.
500 M. Vyberg, P.P. Clausen
Mesenchymal tumours comprise tumours derived from connective tissue cells, en-
dothelial cells, muscle cells, cartilage and bone cells, as well as cells in the lym-
phatic and haemopoietic tissue. The majority of mesenchymal tumours are charac-
terized by giving a positive immunohistochemical reaction with antibodies directed
against vimentin (Sect.32.5.1). A number oftumours derived from muscle cells form
an exception. Some mesenchymal tumours such as synovial sarcoma, epithelioid
sarcoma, and chondroma express keratin filaments (Sect.32.5.1) in addition to vi-
mentin. The preponderant part of tumours derived from muscle cells give a positive
reaction for desmin.
Desmin. This is a polypeptide (mol. wt 50-55 kDa) found in nearly all normal
smooth muscle cells, striated muscle cells and in a few other kinds of cells. Some
few smooth muscle cells, especially those in relationship to vessels, do not contain
desmin, but only vimentin. Desmin can thus be demonstrated in benign and malig-
nant tumours derived from smooth muscle (leiomyomas, leiomyosarcomas) as well
as in tumours derived from striated musc1e (rhabdomyoma and rhabdomyosarcoma)
and also in tumours where a myogenic differentiation takes place, for example in
mesodermal mixed tumour of the ovary or uterus. Desmin is the best general myo-
genic marker. More well differentiated myogenic tumours express actin in addition
to desmin.
This positive reaction is, however, only of limited value in the differential
diagnosis, as positive reaction also may be seen in many other cell types.
Lymphoid and myeloid neoplasia include neoplastic conditions derived from lym-
phoid cells (malignant lymphoma and leukaemias) and from myeloid cells (leukae-
mias and, some localized tumours derived from early stages in the development of
granular leukocytes, monocytes, red blood cells, and platelets).
Malignant lymphomas can be morphologically classified into two main groups:
Hodgkin's lymphoma and non-Hodgkin lymphoma. Non-Hodgkin lymphomas can
further be differentiated, depending on whether they are derived from T-cells or B-
cells, into T- and B-celllymphomas. The earliest B- and T-cells are characterized
by a positive reaction for the enzyme terminal deoxynucleotidyl transferase (TOT)
which may be used as a marker for lymphoblastic lymphomas and leukaemias (see
Table 32.8).
IgM, {j in IgD, and € in IgE. There are two types of light chains - K, and A. An
Ig-producing cell only produces one kind of light chain. The ratio between K,-
and A-chains in the organism is about 3:2. Immunoglobulins are produced by B-
lymphocytes. The earliest chains that can be detected are JL-chains in the cytoplasm.
Later immature and mature B-cell surface chains, that may be detected, are JL-, {j-,
and L-chains. In the further development into plasma cells intracytoplasmatic Ig
can be demonstrated.
Joining Chains. The joining chains (J) which link the subunits of IgM and IgA
are produced in all B lymphocytes and plasma cells but are only secreted bound
to IgA or IgM. The demonstration of J-chains can be used in the identification of
B cells and in the c1assification of lymphomas. The phenotypic expression of T
and B ceIl neoplasias with the reaction pattern for the most important CD-antigens,
immunoglobulins, and TDT are given in Table 32.8.
CD 1-8. These are chiefly present in T lymphocytes, while CD 10, 19, and 22 are
chiefly connected with B lymphocytes. While these antigens can only be satisfac-
torily demonstrated in cryostat sections the foIlowing antigens (CD 15, CD 30, CD
45, and L 26) can all be demonstrated in paraffin sections of formaldehyde fixed
material.
CD 30. CD 30 (105 kDa) is also called X-I antigen. It is found on a small, not
more precisely defined, group of large lymphocytes. In malignant lymphomas CD
30 is found on Hodgkin and Reed-Sternberg cells and on most cells in pleomorphic
T cell lymphoma as well as large cell anaplastic lymphomas.
CD 45. This antigen (200 kDa), also called leukocyte common antigen (LCA)
(Sect32.5.1), is - as previously described - an important marker in the differenti-
ation of malignant lymphomas and other tumours. CD 45 R is adesignation for a
number of CD 45 proteins.
Antibodies can demonstrate either CD 45 R-B (205 and 220 kDa) which is
found on most B lymphocytes and a smaller proportion of T lymphocytes or CD
45 R-T (185 kDa) found on most T lymphocytes. Both CD 45 R-B and CD 45
R-T react, as opposed to most other CD antibodies, on paraffin sections.
L 26. This is a monoclonal antibody which reacts with a, as yet not characterized,
antigen found in the cytoplasm of most B lymphocytes. With L 26, 90-95% of all
B lymphomas can be detected.
Table 32.9 shows the immunohistochemical c1assification of malignant lym-
phomas using antibodies which work on paraffin sections.
32 Applied IIrunWlOhistochemistry 503
Table 32.8. Antigen patterns characterizing cell differentiation (and corresponding neoplasms) of
the lymphoid system.
T Iymphocyte identification
Cells Neoplasms CD: 1 2 3 4 5 6/7 8 TdT
Early T-Iymphoblastic + + + +
thymocyte lymphoma
Cortical + + + + + + + +
thymocyte Acute T-Iymphoblastic
Medullary leukaemia + + +1 + + +2 +
thymocyte
Peripheral Peripheral + +1 + + +2
T cell T-Iymphoma; T-CLL 3
B Iymphocyte identification
Cells . Neoplasms CD: 10 19 22 ~ S-Ig ' C_Ig 2 TdT
?
Erythroblasts
CD13 CD15
CD41 CD42
Megakaryoblats I------t~
Table 32.10. Antigen patterns characterizing cell differentiation of the myeloid system.
The neoplasias of the nervous system consist of tumours derived from the cen-
tral and peripheral nervous system. Tumours of the central nervous system com-
prise in part tumours derived from neurons and in part tumours derived from glial
cells and cells in the ependyma and meninges. The neurone derived tumours in-
clude, according to increasing maturity: neuroblastomas, ganglioneuroblastomas,
gangIioneuromas, gangIiogliomas, and gangliocytomas. These tumours, to increas-
ing degrees, show positive reactions for the neuronal markers: neuron specific
enolase (Sect.32.5.2), synaptophysin (Sect.32.5.2), and neurofilament.
Neurofilament. This comprises three polypeptide types (68, 150, and 200 kDa)
which can be demonstrated in neurons and neuronal processes in the central ner-
vous system and in peripheral nerves as well as ganglion cells and the suprarenal
medulla. In addition to the neoplasias mentioned, positive reactions are also found
in paragangIiomas and phaeochromocytomas.
32 Applied Immunohistochemistry 505
GliafibriIlary Acidic Protein. Astrocytomas are the largest group of tumours de-
rived from the neuroglia. The most valuable marker for both normal and neoplastic
astrocytes is gliafibrillary acidic protein (GFAP).
GFAP is a polypeptide (51 kDa) which can be demonstrated in astrocytes
and a few ependymal cells. A positive reaction is also found in Schwann cells,
folliculostellate cells in the pituitary gland, and satellite cells in ganglia. Posi-
tive reactions can be demonstrated in astrocytomas and tumours with an astrocyte
component (glioblastoma, gliosarcoma) and in a few cases of ependymomas and
oligodendrogliomas. Furthermore, in the central nervous system in a few non-glial
tumours: medulloblastoma, neuroblastoma, retinoblastoma, and pinealoma as well
as choroid plexus papilloma. Outside the central nervous system positive reaction
has been demonstrated in schwannomas and pleomorphic adenomas.
S-lOO Protein. S-100 protein (Sect.32.5.1) is in the nervous system mainly found
in the same cells that express GFAP with the difference that the positive reaction
in addition to being found in the cytoplasm is also often observed in the nuelei.
Tumours derived from oligodendroglia frequently show a positive reaction on
staining with the antibody anti-LEU-7 that reacts with a, as yet not characterized,
membranous antigen found in NK cells (natural killer lymphocytes). A positive
reaction can be shown in elose to 90% of oligodendrogliomas but also in some
astrocytomas, schwannomas, neurofibromas, a few malignant melanomas, and neu-
roendocrine tumours.
Meningeomas show a positive reaction for epithelial membranous antigen and
vimentin (Sect.32.5.1).
One of the most important characters of malignant tumours is their capacity for in-
vasive growth. In tissues delimited by a basement membrane (epithelia) (Sect.2.4.3)
infiltration of this boundary can be considered an early marker of invasive growth.
The basement membrane consists of the basal lamina and the reticular lamina. The
structure and antigenic composition of basement membranes of different tissues
have been determined in considerable detail (Sect.2.4.3). Immunohistochemical in-
vestigations of tumour invasion depend on detecting the disruption of this ordered
structure. The changes detected can be divided into those affecting the intrinsic and
the extrinsic components, respectively.
506 M. Vyberg, P.P. Clausen
The diagnosis of many infectious diseases depends on serologie tests (Le. demon-
stration of serum antibodies against microorganisms) and eulture of microorganisms
from tissue or body fluids. While these diagnostie techniques are likely to remain
imponant in the future, the possibility for identifying viruses and other parasites in
situ combined with the inereased frequeney of opportunistie infections in immuno-
eompromised patients has increased the need for reagents and methods for the in
situ identifieation and localization of these microorganisms.
32 Applied Immunohistochemistry 507
In recent years a number of antisera have been developed which make the
immunohistochemical identification of microorganisms possible. These techniques
have to be viewed in parallel with the potential of in situ hybridization particularly
for viruses (Sect.20.6). Table 32.11 gives a survey of important microorganisms
that can be demonstrated in routine treated tissue.
Campylobacter Leishmania
Candida albicans Morbilli virus
Chlamydia Mycobacteria
Cytomegalovirus Mycoplasma
Hepatitis virus B Polio virus
(Surface and core antigens) Rotavirus
Herpes simplex land II Rubella virus
Human immunodeficiency virus Toxoplasma
Influenza virus Trichophyton
Klebsiella Varicella-zoster virus
Legionella
Appendix A: Standardization of Staining Methods
No detailed descriptions of staining methods are provided in this book. The reader
is referred to one or more of the excellent texts covering this field (cf. Preface).
Nevertheless, we have feIt it appropriate to inc1ude this appendix which gives
some of our views on the technical aspects of procedures which should be given
particular emphasis. It is our opinion that the need for standardization and quan-
titative methods in daily work is pressing. This appendix sets out the appropriate
general considerations followed by a few selected methods in order to cover this
area.
According to Boon and Wittekind (1986) the principle aim of standardizing staining
methods is to render their application reproducible and therefore reliable. This is
of the utmost importance when dyes and stains are used for automated cell pattern
recognition (Wittekind, 1985; Wittekind and Schulte, 1987).
The theoretical background for standardization of cell and tissue preparation
sterns from the fact that any preparatory step - from cell sampling to mounting
of the stained slide - will somehow affect the structure of the cell and ultimately
lead to the production of a particular staining pattern which, in strict tenns, is an
artifact. What we eventually observe by microscopy is, from the perspective of a
cell, the product of a rather violent procedure: In cytological preparations the cells
have been isolated from their tissue, spread out on the surface of a glass slide and
immersed in a liquid poison which abruptly arrests and - sensu stricto - "fixes"
the cell in the very last moment of its life. During fixation some components of the
cell might be removed by the fixative (alcohols remove lipids), and proteins may
be precipitated or cross-linked. In histology the tissue is embedded in molten wax
or in polymerizing plastic; the cell is cut into thin slices on a microtome. Finally -
and this is true in both cytology and histology - dyes dissolved in aqueous or
alcoholic solutions are bound by sometimes unknown mechanisms to one or the
other substrate in the cell thus yielding "contrast" between stained and unstained -
or less stained - components of the cello
H. Lyon (Ed.)
Theory and Sttategy in Histochemistry
© Springer Verlag 1991
510 H. Lyon, D. Wittekind, E. Schulte
What we finally look at has not much to do with the "true" structure of the
living cell; this can be easily confirmed by the comparison of a living cell in phase
contrast microscopy with the same cell after a conventional fixation and staining
procedure in usual light microscopy.
Thus the aim of standardization in cell and tissue preparation is to make the
staining pattern, Le. the artifact, reproducible; in other words standardized prepara-
tory techniques should guarantee standard artifacts.
Finally, we have to answer the question: "Which staining pattern should we
choose as the standard". Variation of the preparatory technique will result in varia-
tion of the staining pattern: if for instance the pH of the staining solution is changed,
hue and/or intensity of staining can change, and a whole palette of staining patterns
will be found when several preparatory steps are changed simultaneously.
It seems logical both from a practical and a theoretical point of view to choose
as standard the staining pattern which best fulfils the requirements of the observer,
be it human or a computer; correspondingly, the preparatory technique yielding
this staining pattern is defined as the standard teChnique. This is explained in detail
below.
Standardization of a method for staining cytological or histological material
requires consideration of all steps in the procedure.
Cytological material may primarily be air-dried and then fixed or fixation may
be carried out on the wet material. The results achieved with a staining method
may be quite different according to which of these procedures has been chosen.
For instance with the Romanowsky-Giemsa method air drying is preferred, while
for the Papanicolaou method air drying has adetrimental effect on the staining
result. The choice of fixative also has a pronounced effect on the staining result;
with the Papanicolaou method Boon and Drijver (1986) recommend the use of
ethanol in concentrations between 50% and 70% and with added polyethylene
glycol (300-1000), whereas this fixative cannot be recommended for use with the
Romanowsky-Giemsa method.
Histological material may also primarily be treated in two different ways. These
are either the preparation of cryostat sections (Sects.11.2.2 and 11.3.1) or the use of
a chemical fixative (Chap.12) which is usually followed by dehydration (Sect.14.2),
clearing (Sect.14.3), and embedding (Sect.14.4) for light microscopy. All of these
procedures have a profound effect on the material before the final sections are cut.
Rigorous standardization is therefore of paramount importance to achieve repro-
ducible results irrespective of the subsequent staining procedure.
Appendix A: Standardization of Staining Methods 511
Concerning the staining methods themselves, assuming that the preparatory tech-
niques have been standardized, the most important reagents are dyes (Sect.3.3).
Dyes. In general, dyes are coloured organic molecules with large systems of delo-
calized electrons (conjugated 7r-electronic systems). Dyes are available as crystals
or as powders which on solution in a suitable solvent may bind by physico-chemical
attractions to a substrate and impart colour to the latter. A stain is a solution of dye
in a suitable solvent. Stains may be subdivided into stock solutions and working
solutions, where a stock solution is a stable solution of one or more dyes at a con-
centration, which is higher than that usually employed for staining, while a working
solution is a solution of one or more dyes in a suitable solvent at concentrations
adapted to staining purposes. Finally, it should be mentioned that a chromogenic
reagent is a colourless reagent which can react with suitable groups present or in-
duced in the biological substrate with the formation of a dye in situ. Standardized
dyes, stains, and chromogenic reagents are of course essential to a standardized
method.
Staining Time. The importance of this factor varies a great deal depending upon the
complexity of the staining process. However, it should be ensured that the staining
times are not too short. It should be appreciated that a staining equilibrium is very
rarely achieved with most staining methods. Extended staining times nearly always
result, therefore, in staining patterns which deviate considerably from the results
obtained with "normal" staining times. Standardization of staining time should not,
however, normally present any problem.
Staining Temperature. The temperature of the staining bath is very important and
will of course effect the staining time. At elevated temperatures the transport of
molecules and ions and reactions between these are accelerated. This means that
increased staining temperature leads to shorter staining time. However, the increase
in temperature may have deleterious effect on the staining result as the proteins of
the tissue become denatured. It is usual to stain at room temperature which probably
can vary between 18°C and 28°C. At the higher temperature chemical processes
take place at double the speed of those at the lower. Control of staining temperature
can therefore be quite important. The temperature of interest is that of the staining
solution and not so much that of the surroundings. It should be remembered that
buffers or other reagents taken direct1y from the refrigerator take a considerable
amount of time to come into temperature equilibrium with the surroundings.
Contact of Stain with the Section or the Cells. This is a difficult factor to control
as it is influenced by movement of the slides in the staining bath and also by
the space between them in the staining rack. Standardization of this factor can,
according to Zimmermann (1983), only be achieved when staining machines are
employed
case when polychrome staining procedures are used; the different dyes may be
differentially affected by the rinsing process.
Reagents.
Ethyl Green, C.!. 42590, or Methyl Green, C.!. 425851)
Pyronin Y, C.I. 450052)
Polyamide, e.g. Polyamide 006, Machery-Nagel, cat. no. 81562
I-propanol
Formic acid
NaBF4 -solution
0.1 mol/l potassium hydrogen phthalate
0.1 mol/l HCI
DNase I
RNase A
99% vIv ethanol
95% vIv ethanol
70% vIv ethanol
Distilled water
I-butanol
Xylene
Hydrophobie synthetic resin
1) Ethyl Green, C.!. 42590, or Methyl Green, C.I. 42585, in pure form. If this is
not available, commercial dyes may be purified by dissolving a quantity equivalent
to 0.15 g pure Ethyl Green or Methyl Green in 50 m1 warm distilled water. 5
g polyamide (Polyamide CC6, Macherey-Nagel, cat. no. 81562) is added. After
stirring for 30 min with a magnetic stirrer, suction filtration is carried out through
a glass filter (diameter 7 cm, pore size 3). The bottom of the glass filter is covered
with a layer of polyamide (approximately 3 g) which has been ftushed with water
and dried by suction. This layer is covered by a piece of filter paper to prevent
the polyamide from being suspended during filtration. The dye solution is poured
over the filter without suction. When the funnel is filled up, suction is applied
514 H. Lyon, D. Wittekind, E. Schulte
gently. The residue in the flask is washed with a little water, and the filter cake is
rinsed with water to obtain a total volume of 90 ml. The degree of purification is
tested by performing thin layer chromatography with I-propanol:formic acid, 4:1,
as the eluting solvent. Crystal Violet is easily detected as it has a higher Rf value
than Methyl Green and Ethyl Green as weIl as a characteristic colour. If proper
purification has not been obtained, the solution is filtered once more through a layer
of fresh polyamide on the filter. The solution is subsequently precipitated with a
NaBF4-solution and recrystallized from ethanol giving analytically pure Methyl or
Ethyl Green tetrafluoroborate. (Andersen et al., 1986).
2) Pyronin Y, C.IA5OO5. Certified sampies containing at least 45%, and preferably
between 80% and 100%, of the pure dye should be used.
Staining Procedure.
1. The sections are hydrated
2. Stain for 5 min at room temperature (22°C) in the working solution
3. Rinse in two changes of distilled water, 2-3 seconds in each
4. Shake off surplus water
5. Agitate in three changes of 1-butanol until clear
6. Mount direct1y from 1-butanol or via two changes of xylene in a hydrophobie
synthetic resin
Note that if glycol methacrylate sections are used instead of paraffin sections
dehydration cannot be performed with I-butanol. In these circumstances the sec-
tions should be dried over silica gel in athermostat oven at 37°C.
Staining Results.
Camoy Formaldehyde
chromatin green blue
nucleoli red (lllac)red
cytoplasmic RNA red (lllac)red
cartilage matrix orange orange
mast cells orange unstained
other structures unstained unstained
Appendix A: Standardization of Staining Methods 515
Reagents.
Azure B-SCN, C.I. 52010, DIN 5898111, in pure form l )
Eosinic acid, C.I. 45380, DIN 58981/2, in pure form l )
Dimethylsulphoxide (DMSO)
0.03 mol/l N-(2-hydroxyethyl) piperazine-N-2-ethanesulphoxide acid (HEPES)
buffer, pH 6.5
Para-toluenesulphonie acid 0.1 %
NaHC030. 1%
DistilIed water
2-propanol
Toluene
Hydrophobie synthetic resin
1) Both dyes are easily soluble in dimethylsulphoxide (DMSO) at the appropriate
concentration.
700 ml distilled water is added to 300 ml of this buffer stock solution to obtain
the buffer working solution of 0.03 mol/l pH 6.5 (which is simply referred to as
"buffer" in the following).
Material. Air dried cytological material (smears or imprints) may be fixed rou-
tinely in methanol p.a. 10 min. Stock solution diluted by methanol is strongly
recommended as stainlfixative solution. The stainlfixative solution is prepared by
dilution of 1 vol stock solution with 15 vol methanol analytical grade. Slides are
immersed for three min in stainlfixative solution. Use of the stainlfixative solution
avoids the loss of basophil granules which unavoidably occurs in methanol. His-
tological material can be fixed in 4% formaldehyde containing 1 g ZnS04 in 100
ml formaldehyde. Both paraffin and glycolmethacrylate (GMA) embedded material
can be used.
Staining Results.
Cytology: Nuclear chromatin: purple; nucleoli: light blue; basophilic cytoplasm:
blue; basophilic granules: purple-black; eosinophilic granules: red-orange; neu-
trophilic granules: purple; erythrocytes: pink-orange. Note: Staining has not been
completed unless cell nuclei are purple.
Histology: Nuclear chromatin: purple; RNA-rich cytoplasm: blue; cytoplasm with-
Appendix A: Standardization of Staining Methods 517
out or with low content of DNA: greenish-blue; intestinal goblet cells: purple/blue;
eosinophil granules: orange; basophiVmast cell granules: red purple; cartilage: red
purple; bone: violet; basement membranes: purple; elastic fibres: green; axons:
purple.
Appendix B: Quantitative Morphological Methods in
Microscopy
E. Hasselager
B.l Definitions
B.2 Observations
IL Lyon (Ed.)
Theory and Sttalegy in Histochemislty
© Springer Verlag 1991
520 E. Hasselager
B.3 Stereology
In the context of stereology, unbiased values are "without systematic deviation from
the true value". To ensure this observations must be done on isotropic, uniform,
random sections.
Depending on the objectives of the study, tissue blocks must be selected either
at random or systematically so that they are demonstrably representative of the
structure or element under investigation. Fields for observation on individual sec-
tions must also be sampled systematically, independent of content and observer,
and inside a well-defined region.
To determined the number of profiles (i.e. structures of any kind) per area, unbiased
counting frarnes are used as discussed by Gundersen et al. (1988a).
Appendix B: Quantitative Morphological Methods in Microscopy 521
The area of profiles per section is easily obtained by throwing a lattiee of test
points over the section and simple eounting of points over profiles over total points
on a seetion giving the area fraction of profiles.
The boundary of profiles per area ean be measured by eounting intersections of
test lines and profile boundary. This is then related to area measurement as above.
B.3.4 Anisotropy
If anything in the tissue or eells has a preferred orientation, for example musele
fibres, eapillaries, or mitochondria, the tissue is not isotropie. The eontribution of
these eonstituents to the seetion depends on the orientation of the section plane.
Isotropie seetions ean be prepared by using the Uorientator" where two eonseeutive
seetions at right angles to each other are made under random guidanee (Gundersen
et al., 1988a).
In many biologieal tissues a partieular orientation of seetions is neeessary to
obtain information absent in other sections. Examples are skin eovered with ep-
ithelium, intestinal erypts, eontractile fibers in musele eells, or the long axis of
microvilli to identify their eellular origin.
A vertical seetion is a section plane perpendicular to aplane of reference
for instanee a basal lamina or the surfaee of an organ. Struetures with preferred
orientation ean always be seen in one vertieal seetion properly rotated. The use of
test line systems for surfaee parameters must be weighted by the sine of the angle
between test line and vertieal plane axis, or the use of a special eyeloid test line
system (Baddeley et al., 1986).
A set of new stereologieal methods for eorrect sampling and sizing of cells or
partieles has been reviewed by Gundersen et al. (1988b).
522 E. Hasselager
This is a pair of parallel section planes a known distance apart. A partic1e is sampled
if it hits one of the planes (reference plane) and not the other (look-up plane). By
counting the number of partic1es only seen on reference planes the exact number
of partic1es in the physical volume between the two planes is obtained.
The optical disector uses thin optical seetion planes produced inside one thick
section (5-100 pm) of a block. The optical sections are produced by moving the
focus plane down through the object. Objectives with high numerical aperture and
small focal depth give well-defined thin focal planes. The confocal microscope
gives excellent optical sectioning of thick blocks, but only in reflectance or epiflu-
orescent microscopy (cf. Sects.28.4 and 28.3).
When measuring the distance I from any point to the boundary of a profile in any
isotropic direction, and averaging the squared distances, then 11'12 is an unbiased
estimate of the area. The area of a circle is a special case. This is the basis for
the nucleator where measurements are carried out isotropically with respect to a
fixed point, for example a nuc1eus in mononuc1eated cells, or more efficiently a
nuc1eolus if only one is present in every cello The nucleus in an oocyte can be used
for follicular parameters.
With the fractionator it is possible to sampie uniformly at random with a prede-
termined probability. Tissue or organs are cut into slabs, some of these are sampled
uniformly, cut into bars, sampled again and cut into blocks for final sampling. The
final blocks are embedded together as a conglomeration. Sections of the conglomer-
ation will be effective unbiased sampies of the original material with "concentrated"
information.
After proper sampling and sectioning something has to be measured on the sections.
The simplest way of doing this is to have the section projected onto a test lattice
with points, lines, etc., or to have the test system superimposed in the microscope.
Areas and boundary lengths are estimated as number of points and intersections
which in turn are transformed to volume and surface area respectively.
Appendix B: Quantitative Morphological Methods in Microscopy 523
A digital image is a body ofbinary data from picture elements (pixels) each contain-
ing information of various local parameters describing intensity of light, colours,
invisible parts of the spectrum, extinction, secondary electrons, x-rays or element
analysis. The pixels are arranged in a 2-dimensionaI square, trlangular, or hexag-
onal matrix consisting of for instance 512x512 elements. With grey levels from
o to 255 (8 bits) in each element, one digital image needs approximately 0.25
Megabytes of memory. A colour image based on three colours in each pixel needs
0.75 Megabytes storage.
To handle single stationary images a limited amount of conventional computer
power and memory is needed. If, however, the images arrive at video rate, 50 Hz,
about 15 Megabytes information per second has to be put into a digital frame
store, interfaced to and processed by a high performance computer, and, finally,
the results have to be returned to the frame store for inspection on a monitor at
video rate.
Conversion of continuous (analogue) images into a digital form acceptable to
computers involves two important processes: sampling and quantitation, jointly
referred to as "digitization". Sampling represents the image by measurements at
spaced intervals traversing the image in a specific pattern. The distance between
sampie points determines spatial resolution. Quantitation transforms brightness val-
ues into a range of integer numbers. The number of bits used governs greyscale res-
olution. A diagram for obtaining and showing digital images is shown in Fig. B.l.
VIDEO video
.• .·• I video
~-----.
.... output
VIDEO
r--t-.~
IN input ••• OUT
,--u_n_i_t---ll : • I unit
Imemory I •
••••• ,.(frame ••••••
image/
signal
I I·
scanner.. • .
I
store) I I •
. • ... printer
I
image
I I I I
analogue
•••• : digital.
Fig. AppB.l. Flow chart and handling of different signal types in elcttonic image processing.
When stored in digital form the image can be manipulated in many ways. For
an introduction to digital image processing, see Niblack (1985).
524 E. Hasselager
,-----------, 20 20 20 20 20
I I 20
I
160 160 160 160 20 20
80 80 80 80 160 20
C 80
220
80
220
80 80 160
80 80 160
20
20
220 220 220 80 160 20
220 220 220 80 50 20
220 220 220 80 160 20
220 220 220 80 160 20
220 220 80 80 160 20
,I
, 80 80 80 80 160 20
I 80 80 80 80 160 20
I 160 160 160 160 20 20
I
--------- 20 20 20 20 20 20
Fig. AppB.2. A: Section of 3 liver cells (only partly shown). Nuclei (N), cytoplasm (C), and
background with discrete level of intensity. B: Corresponding digital array of numbers representing
grey levels from rectangle indicated in A.
number of pixels
250
200
150
100
50
o
o 50 100 150 200 250
grey level
Grey level histograms. One of the simplest yet most useful tools are grey level
histograrns. For each grey level, the number of pixels having that grey level are
shown, see Fig. B.3. The grey levels can be redistributed by non-linear greyscale
transformation. By stretching the grey levels in the populated regions of the his-
togram and compressing them in others, the contrast may improve tremendously.
Setting a threshold for grey levels is a simple case of this principle.
Point operation. In point operation each output pixel is related in a systematic way
to the corresponding input pixel. By algebraic operations images may be added,
subtracted, multiplied, etc.
Appendix B: Quantitative Morphologica1 Methods in Microscopy 525
Local or neighbourhood operation. This means, that each output pixel may de-
pend not only on the grey level of the corresponding input pixel, but also on the
grey levels of certain of the neighbouring input pixels. A mask is applied to every
single pixel in the input image, and the resulting value from the operation put into
the corresponding output pixel. The masks shown here indicate the weights and
disposition of pixels participating in the operation:
The real advantage is seen when larger masks or more elaborate filters than
mentioned above have to be applied. When a mask (m) is intended for used on an
image (i), only two Fourier Transforms (Ff) have to be calci.llated as
FT(i) x FT(m) = FT(i&m)
and the inverse FT(i&m) gives the resulting image of (i) convoluted with (m). This
approach is much faster for large masks than the use of direct spatial convolution.
More information can be found in Pratt (1978).
Erosion and Dilation. Erosion strips off layers of pixels from the object. Dilation
adds on layers of pixels. By combined sequences of erosion and dilation the con-
nectivity of elements can be investigated and noise "defects" repaired For example,
mitotic figures can be retained, whereas insignificant particles are removed.
Image Restoration. When the output digital image is to be presented the single
pixel values may be translated as they are sent to the display. This requires a
hardware look-up table (LUT) adopting the grey values to ideal settings for the
display system or transforming them to colours. False colours are useful both to
improve grey level resolution as the human eye can not distinguish more than
approximately 50 grey levels, and in order to display invisible bands or other
information.
Measurements of Height and Depth. These are made possible by viewing the
object from different viewpoints. A parallax is generated by shifting the specimen,
tilting the specimen, or by tilting the electron beam (TEM/SEM). This area is still
devel~ping.
Integrated Systems for Digital Image Analysis. The single components of a dig-
ital image processing system must be set up with protocols ensuring compatibility.
Appendix B: Quantitative Morphological Methods in Microscopy 527
With increasing demands on image resolution and the handling of images at video
rate, larger frame stores and high performance computers are needed.
Again quite a number of different systems are offered by the commercial com-
panies. Most of them are based on image registration by video cameras. All have
facilities for thresholding, filtering and contrast enhancement All are able to give
parameters for areas, lengths, numbers, grey level histograms, etc. in the two-
dimensional images. Many of them have special application programs for various
disciplines. In biology for counting and classifying blood cells, finding mitotic fig-
ures, registering chromosomes, analyzing silver grains in autoradiographs or any
kind of autometallographic technique, and contrast enhancement of radiographs. In
biochemistry for analysis of chromatograms and electrophoresis patterns.
Other types of programmes have been developed for geology, carthography,
aerial photography, and remote sensing data of any kind from satellites or space
crafts.
All these techniques give useful results for two dimensions but the whole body
of stereological methods mentioned above are required to transform the data to
meaningful parameters for three-dimensional analysis. It is absolutely essential that
the investigator should carefully consider shrinkage, anisotropy, uniform unbiased
sampling, correction procedures, etc., irrespective of the elegance and ease by which
two-dimensional results are obtained.
Abe, M., Kramer, S.P., and Seligman, A.M.: The histochemical demonstration of pancreatic-like
lipase and comparison with the distribution of esterase. J.Histochem.CytOChem. 12:364-383
(1964).
Abrahart, E.N.: Dyes and their intermediates. 2nd edit. Edward Arnold, London (1977).
Adams, C.W.M.: A p-dimethylaminobenzaldehyde nitrite method for the histochemical demon-
stration of tryptophane and related compounds. J.Clin.Pathol. 10:56-65 (1957).
Adams, C.W.M.: Histochemical mechanisms of the Marchi reaction for degenerating myelin.
J.Neurochem. 2:176-186 (1958).
Adams, C.W.M.: A histochemical method for the simultaneous demonstration of normal and
degenerating myelin. J.Path.Bact. 77:~50 (1959).
Adams, C.W.M.: Osmium tetroxide and the Marchi method: Reactions with polar and non-polar
lipids, protein and polysaccharide. J. Histochem.Cytochem. 8:262-267 (1960).
Adams, C.W.M.: A perchloric acid-naphthoquinone method for the histochemica1localisation of
cholesterol. Nature (London) 192:331-332 (1961).
Adams, C.W.M.: Neurohistochemistry. Elsevier, Amsterdam (1965).
Adams, C.W.M. and Bayliss, O.B.: The release of protein, lipid and polysaccharide components
of the arterial elastica by proteolytic enzymes and lipid solvents. J.Histochem.Cytochem.
10:222-226 (1962).
Adams, C.W.M. and Bayliss, O.B.: Histochemica1 observations on the localization and origin
of sphingomyelin, cerebroside and cholesterol in normal and arteriosclerotic human artery.
J.Path. Bact. 85:113-119 (1963).
Adams, C.W.M. and Bayliss, O.B.: Lipid histochemistry.In: D. Glick and R.M. Rosenbaum, Eds.:
Techniques of biochemical and biophysical morphology. Vol. 2. Wlley Interscience, New York
(1974).
Adams, C.W.M. and Davison, A.N.: The histochemical identification of myelin phosphoglycerides
by their ferric hydroxamates. J.Neurochem. 3:347-351 (1959).
Adams, C.W.M., Abdullah, Y.H., and Bayliss, O.B.: Osmium tetroxide as a histochemical and
histological reagent. Histochemie 9:68-77 (1967).
Adams, C.W.M., Bayliss, O.B., and Ibrahim, M.Z.M.: Modifications to histochemical methods
for phosphoglyceride and cerebroside. J. Histochem.Cytochem. 11:560-561 (1963).
Adams, C.W.M., Abdullah, Y.H., Bayliss, O.B., and Weller, R.O.: Histochemica1 detection of
triglyceride esters with specific lipases and a calcium lead sulphide technique. J.Histochem.
Cytochem. 14:385-395 (1966).
Aikens, R.S., Agard, D.A. and Sedat, J.W.: Solid state images for microscopy, pp.292-313. In:
Y.-L. Wang and Taylor, D.L., Eds.: Methods in cell biology, Vol. 29. Fluorescence microscopy
of living cells in culture. Part A. Fluorescent analogs, Iabeling cells, and basic microscopy.
Academic Press, Boston (1989).
Albedi, F.M., Cassano, A.M, Ciarelli, F., Donelli, G., Giuliani, Mingazzini, P., and Marinozzi,
v.: lnftuence of cetylpyridinium chloride on the ultrastructural appearance of sulphated gly-
cosaminoglycans in human colonic mucosa. Histochemistry 89:397-401 (1988).
Allison, R.T.: The effects of various fixatives on subsequent lectin binding to tissue sections.
J.Histochem. 19:65-74 (1987).
Altman, F.P.: Quantitative dehydrogenase histochemistry with special reference to the pentose
shunt dehydrogenases. Progr.Histochem.Cytochem. 4/4:225-273 (1972).
530 References
Baker, D.M. and Santer, R.M.; Development of a quantitative histochemical method for deter-
mination of succinate dehydrogenase activity in autonomie neurons and its application to the
study of aging in the autonomie nervous system. J.Histochem.Cytochem. 38: 525-531 (1990).
Baker, J.R. The histochemical recognition of lipine. Quart.J. Micr.Sei. 87:441-470 (1946).
Baker, J.R.: Principles of biological microtechnique. Methuen, London, (1958).
Baker, J.R.: Cytological technique. The principles underlying routine methods. 5th edit. Science
Paperbacks Methuen & Co. Ltd., London, (1966).
Bancroft, J.D. and Cook, C.: Manual of histological techniques. 2nd edit. Churchill-Livingstone,
Edinburgh, (1984).
Bancroft, J.D., and Stevens, A. Eds.: Theory and practice of histological techniques. 3rd edit.
Churchill-Livingstone, Edinburgh (1990).
Barka, T. and Anderson, P.J.: Histochemistry. Theory, practice, and bibliography. Hoeber,lIl"ew
York (1963).
Barer, M.R.: Development of a ceD culture method for the detection and quantitation of bacterial
enterotoxins. PhD Thesis. University of London (1987).
Barer, M.R., Lyon, H.O., and Drasar, B.S.: Quantitation of dye binding to ceD monolayers in a
microtiter system. HistochemJ. 18:122-128 (1986a).
Barer, M.R., Mann, G.F., and Drasar, B.S.: A semi-automated system for the assessment of
toxicity to cultured mammalian cens based on detection of changes in staining properties.
Developments in Biological Standardization 64:251-259 (1986b).
Barrett, A. and Heath, M.F.: Lysosomal enzymes, pp.I9-145. In: J.T. Dingle, Ed.: Lysosomes: a
laboratory handbook, 2nd edit., Elsevier, North Holland Biomedical Press (1977).
Barrett, AJ.: The many forms and functions of cellular proteinases. Fed.Proc. 39:9-14 (1980).
Barmett, RJ. and Seligman, A.M.: Demonstration of protein-bound sulfhydryl and disulfide
groups by two new histochemical methods. J.Natl.Cancer Inst. 13:215-216 (1952).
Bartholomew, J.W.: Stains for microorganisms in smears. In: G. Clark, Ed.: Staining procedures.
4th edit., pp.375-440. Williams & Willdns, Baltimore (1981).
Bauman, J.G.J., van der Ploeg, M. and van Duijn, P.: Fluorescent hybridocytochemical procedures:
DNAJRNA hybridization in situ, pp.41-87. In: J. Chayen and L. Bitensky, Eds.: Investigative
microtechniques in medicine and biology. Vol. l. Marcel Dekker, Inc., New York (1984).
Bayliss, O.B. and Adams, C.W.M.: Bromine Sudan Black (BSB). A general stain for tissue lipids
including free cholesterol. HistochemJ. 4:505-515 (1972).
Becher, S.: Untersuchungen über die Echtfllrbung der Zellkerne mit künstlichen Beizenfarbstoffen
und die Theorie des histologischen Farbprozesses mit gelösten Lacken. Gebr. Bornträger,
Berlin (1921).
Belt, W.D. and Hayes, E.R.: An ultraviolet - Schiff reaction for unsaturated lipids. Stain Technol.
31:117-122 (1956).
Bennion, PJ. and Horobin, R.W.: Some effects of salts on staining: Use of the Donnan equilibrium
to describe staining of tissue sections with acid and basic dyes. Histochemistry 39:71-82
(1974).
Benno, R.H., Tucker, L.W., Joh, T.H. and Reis, D.J.: Quantitative immunocytochemistry of ty-
rosine hydroxylase in rat brain. I. Development of a computer assisted method using the
peroxidase-antiperoxidase technique. Brain Res. 246:225-236 (1982a).
Benno, R.H., Tucker, L.W., Joh, T.H. and Reis, D.J.: Quantitative immunocytochemistry of ty-
rosine hydroxylase in rat brain. 11. Variations in the amount of tyrosine hydroxylase among
individual neurons of the locus coeruleus in relationship to neuronal morphology and topog-
raphy. Brain Res. 246:237-247 (1982b).
Bentley, SA, Marshali, P.N., Wade, M.J. and Galbraith, W.: Azure B-eosin staining of blood
cens. The effects of variation in stain formulation and staining technique on stain performance.
Anal.Quant.Cytol. 1:179-186 (1979).
Berrisford, R.G. Wells, M. and Dixon, M.F.: Gastric epithelial mucus-a densitometric histo-
chemical study of aspirin-induced damage in the rat. Br.J.exp.Path. 66:27-33 (1985).
Berube, G.R., Powers, M.M., Kerkay, J., and Clark, G.: The gallocyanin-chrome alum stain;
influence of methods of preparation on its activity and separation of active staining compound.
Stain Technol. 41:73--81 (1966).
532 References
Brachet, J.: La localisation des acides pentose nucleiques dans les tissus animaux et les oeufs
d'amphibiens en voil de development. Arch.Biol. 53:207-219 (1942).
Brachet, J.: The use of basic dyes and ribonuclease for the cytochernical detection of ribonucleic
acid. Quart.J.Micr.Sci. 94:1-10 (1953).
Braendgaard, H., Evans, S.M., Hovard, C.V. and Gundersen, H.J.G.: The total number ofneurons
in the human neocortex unbiasedly estimated using optical disectors. J. Microsc. 157:285-304
(1989).
Bresser, J. and Evinger-Hodges, M.J.: Comparison and optimization of in situ hybridization pro-
cedures yielding rapid, sensitive mRNA detections. Gene Anal.Techn. 4:89-104 (1987).
Brody, IA. and Engel, W.K.: Isoenzyme histochemistry: the display of selective lactate dehydro-
genase isoenzymes in sections of skeletal muscle. J.Histochem.Cytochem. 12:928-929 (1964).
Buckingham, J.C. and Hodges, J.R.: The use of corticotrophin production by adenohypophyseal
tissue in vitro for the detection and estimation of potential corticotrophin releasing factors.
J.Endocrinol. 72:187-193 (1977).
Buckingham, J.C., Chayen, J., Hodges, J.R., Robertson, W.R. and Weisz, J.: A cytochernical
section assay method for the determination of luteinizing hormone. J.Endocrinol. 81:16Op
(1979).
Buk, SJA. and Bayliss High, O.B.: Periodate oxidation of glycolipids: a borohydride-periodate-
Schiff method for ganglioside demonstration in tissue sections. Histochem.J. 18:228-232
(1986).
Bullock, G.R. and Petrusz, P., Eds.: Techniques in immunocytochemistry, voLl. Academic Press,
London (1982).
Bullock, G.R. and Petrusz, P., Eds.: Techniques in immunocytochemistry, vol.2. Academic Press,
London (1983).
Bullock, G.R. and Petrusz, P., Eds.: Techniques in immunocytochemistry, vol.3. Academic Press,
London (1985).
Burstone, M.S.: Histochemical methods for pro tein detection. In: W. Graumann and K. Neumann,
Eds.: Handbuch der Histochemie, vol. III, 2nd edit. Gustav Fischer Verlag, Stuttgart (1959).
Burtner, H.J. and Lillie, R.D.: A live-hour variant of Gomori's methenamine silver method for
argentaffine cells. Stain Technol. 24:225-227 (1949).
Butcher, R.G.: The estimation of the nucleic acids of tissue sections and its use as a unit of
comparison for quantitative histochemistry. Histochemie 13:263-275 (1968).
Butcher, R.G.: The chemical determination of section thickness. Histochemie 28:131-136 (1971).
Butcher, R.G.: The measurement in tissue sections of the two formazans derived from Nitroblue
Tetrazolium in dehydrogenase reactions. HistochemJ. 10:739-744 (1978).
Butcher, R.G.: The oxygen insensitivity phenomenon as a diagnostic aid in carcinoma of the
bronchus, pp.241-25l. In: J.R. Pattison, L. Bitensky and J. Chayen, Eds.: Quantitative cyto-
chemistry and its applications. Acadernic Press, London (1979).
Butcher, R.G.: Grades of PVA and enzyme retention in tissue sections. Histochem.J. 14:165-166
(1982).
Butcher, R.G.: Enzyme histochemistry of the myocardium. In: AJ. Drake-Holland and I.M. Noble,
Eds.: Cardiac metabolism, pp.445-469. John Wiley & Sons, New York (1983).
Butcher, R.G.: Reaction rate studies of glucose-6-phosphate dehydrogenase in rat tracheal epithe-
lium; the effect of section thickness. Histochemistry 81:567-572 (1984).
Butcher, R.G. and Altman, FP.: Studies on the reduction of tetrazolium salts. II. The measurement
of the half reduced and fully reduced formazans of neotetrazolium chloride in tissue sections.
Histochemie 37:351-363 (1973).
Butcher, R.G. and Evans, A.W.: Diffusion during dehydrogenase reactions: the effects of inter-
mediate electron acceptors. Histochem.J. 16:885-895 (1984).
Butcher, R.G. and Papadoyannis, D.E.: Histochemical measurement of myocardial phosphofruc-
tokinase activity. J.Physiol. 289:18P-19P (1979).
Butcher, R.G. and van Noorden, CJ.F.: Reaction rate studies of glucose-6-phosphate dehydroge-
nase activity in sections of rat liver using four tetrazolium salts. Histochem.J. 17:993-1008
(1985).
Butcher, R.G., Bitensky, L., Cashman, B., and Chayen, J.: Differences in the redox balance in
human rheumatoid and non-rheumatoid synoviallining cells. Beitr.Path. 148:265-274 (1973).
534 References
Cabrini, R.L, Frasch, A.C.C. and ltoiz, M.E.: Quantitative histochernistry of esterases. His-
tochem.J. 9:369-371 (1977).
Caffrey, J.L., Nett, T.M., Abel, J.H. and Niswender, G.D.: Activity of 3ß-hydroxy-A 5-steroid
dehydrogenase/A 5-A 4/isomerase in the ovine corpus luteum. Biol.Reprod. 20:279-287
(1979).
Cain, A.J.: On the significance of the plasma! reaction. Quart J.Micr.Sei. 90:75-86 (1949a).
Cain, A.J.: A critique of the plasma! reaction, with remarlcs on recendy proposed techniques.
QuartJ.Micr.Sci. 90:411-426 (1949b).
Cain, A.J.: The histochemistry of lipids in animals. Biol.Rev. (Cambridge) 25:73-112 (1950).
Cajal, S. Ramon y: Sobre un neuvo proceder de impregnaei6n de la neuroglia y sus resultados en
los centros nerviosos deI hombre y animales. Trab. Lab. Invest Biol. 11: Fase. 3, Dieiembre
(1913) Cited in: G. Clark, Ed.: Staining Procedures,4th edit., pp. 134-135. Williams & Wllkins,
Baltimore (1981).
Cajal, S. Ramon y: EI proceder deI oro-sublimado para la coloraei6n de la neuroglia. Trab. Lab.
Invest Biol. 14: Fasc. 3, (1916).
Caspersson, T.: Methods for the determination of the absorption spectra oe cell structures. J.R
Microsc.Soc. 60:8-25 (1940).
Casselman, W.G.B.: Acetylated Sudan Black B as a more specific histochemical reagent for lipids.
QuartJ.Micr.Sci. 95:321-322 (1954).
Chambers, D.1., Braimbridge, M.V., Frost, G.T.B., Nahir, A.M., and Chayen, J.: A quantitative
cytochemical method for the measurement of ß-hydroxyacyl CoA dehydrogenase activity in
rat heart muscle. Histochemistry 75:67-76 (1982).
Chambers, D.1., Dunham, J., Zanelli, J.M., Parsons, JA, Bitensky, L. and Chayen, J.: A sensitive
bioassay of parathyroid hormone in plasma. Clin.Endocrinol. 9:375-379 (1978).
Chayen, J.: The cytochemical approach to hormone assay. IntRev. Cyt. 53:333-396 (1978).
Chayen, J.: The cytochemical bioassay of polypeptide hormones. Springer-Verlag. Berlin (1980).
Chayen, J. and Denby, E.F.: Biophysical technique as applied to cell biology. Methuen, London
(1968).
Chayen, J., Altman, F.P., and Butcher, RG.: The effect of certain drugs on the production and pos-
sible utilization of reducing equivalents outside the mitochondria, pp.196-230. In: S.Dikstein,
Ed.: Fundamentals of cell pharmacology. Thomas, Springfield (1973a).
Chayen, J., Bitensky, L., and Butcher, RG.: Practical histochemistry. John Wiley & Sons, London
(1973b).
Chayen, J., Loveridge, N. and Daly, J.R.: A sensitive bioassay for adrenocorticotrophic hormone
(ACTH) in human plasma. Clin.Endocrinol. 1:219-233 (1972).
Chayen, J., Daly, J.R, Loveridge, N. and Bitensky, L.: The cytochemical bioassay of hormones.
Recent Progr.Horm.Res. 32:33-79 (1976).
Chayen, J., Bitensky, L., Johnstone, J.J., Gooding, P.E. and Slater, T.F.: The application of mi-
crospectrophotometry to the measurement of cytochrome P-450, pp. 129-137. In: J.R Pattison,
L. Bitensky and J. Chayen, Eds.: Quantitative cytochemistry and its applications. Academic
Press, London (1979).
Chayen, J., Gilbert, D.M., Robertson, W.R., Bitensky, L. and Besser, G.M.: A cytochemical
section-bioassay for thyrotropin. J.Immunoassay 1:1-13 (1980).
Chayen, J., Frost, G.T.B., Dodds, RA, Bitensky, L., Pitchfork, J., Baylis, P.H., and Barmett,
RJ.: The use of a hidden metal-capture reagent for the measurement of Na+ -K+ -ATPase
activity: A new concept in cytochemistry. Histochemistry 71:533-541 (1981).
Cheng, K.K.: Experimental studies on the mechanism of the wnal distribution of beryllium liver
necrosis. J.Path:Bact. 71:265 (1956).
Chevremont, M. and Frederic, J.: Une nouvelle methode histochimique de mise en evidence des
substances a fonction sulfhydrile. Arch.Biol. (paris) 54:589-605 (1943).
Chieco, P., Normanni, P., Moslen, M.T. and Maltoni, C.: Quantitative histochemistry of benzalde-
hyde dehydrogenase in hepatocellular carcinomas of vinyl chloride-treated rats. J.Histochem.
Cytochem. 34:151-158 (1986).
Clancy, M.J.: The formulation of buffers and media for enzyme histochemistry. Histochem.1.
19:27-34 (1987).
Clark, G., Ed.: Staining procedures. 4th edit. Williams & Wilkins, Baltimore (1981).
References 535
Cobbold, P .H. and Rink, T J.: Fluorescence and bioluminescence measurement of cytoplasmic
. free calcium. Biochem.J. 248: 313-328 (1987).
Cole, L.C. and Nettleton, G.S.: A histochemical investigation of the mechanism of Aldehyde
Fuchsin staining of pancreatic B-cell. Histochem.J. 20:635-641 (1988).
Comelese-ten Velde, I. and Prins, F.A.: New sensitive light microscopical detection of colloidal
gold on ultrathin sections by reftection conttast microscopy. Combination of refiection con-
ttast and electron microscopy in post-embedding immunogold histochemistty. Histochemistty
94:61-71 (1990).
Comelese-ten Velde, 1., Bonnet, J., Tanke, HJ. and Ploem, J.S.: Refiection conttast microscopy.
VlSualization of (peroxidase-generated) diaminobenzidine polymer products and its underlying
optical phenomena. Histochemistty 89:141-150 (1988).
Comelese-ten Velde, I., Wiegant, J., Tanke, HJ. and Ploem, J.S.: Improved detection and quantifi-
cation of the (immuno)peroxidase product using refiection conttastmicroscopy. Histochemistty
92:153-160 (1989).
Coulton, L.A., Henderson, B. and Chayen, J.: The assessment of DNA-synthetic activity. Histo-
chemistry 72:91-99 (1981). -.
Cowden, R.R. and Curtis, S.K.: Further comments on the selectivity of Mercury Orange for
protein thiols. Histochem.J. 16:1237-1240 (1984).
Cowen, T., Haven, A.J., and Bumstock, G.: Pontamine sky blue: A counterstain for back-
ground autofiuorescence in fiuorescence and immunofiuorescence histochemistry. Histochem-
istty 82:205-208 (1985).
Cremers, A.F.M., Jansen in de Wal, N., Wiegant, J., Dirks, R.W., Weisbeek, P., van der Ploeg,
M. and Landegent, J.E.: Non-radioactive in situ hybridization. A comparison of several im-
munocytochemical detection systems using refiection-conttast and electton microscopy. His-
tochemistty 86:609-615 (1987).
Croce, A.C., Bottiroli, G., Prosperi, E., Supino, R. and Stoward, P J.: Limitations of the quanti-
tative cytochemical assay of catechol oxidase in melanoma cells. Histochem.J. 20: 595-602
(1988).
Crockard, A.D.: Cytochemistty of lymphoid cells: a review of findings in the normal and leukaemic
state. HistochemJ. 16:1027-1050 (1984).
Cross, S.A.M., Ewen, S.W.B., and Rost, F.W.D.: A study of the methods available for the cyto-
chemicallocalization of histamine by fiuorescence induced with o-phthalaldehyde or acetalde-
hyde. Histochem.J. 3:471-476 (1971).
Culling, C.F.A., Reid, P.E., and Dunn, W L.: A new histochemical method for the identifica-
tion and visualization of both side chain acylated and non-acylated sialic acids. J.Histochem.
Cytochem. 24:1225-1230 (1976).
Culling, C.F.A., Reid, P.E., Clay, M.G., and Dunn, W.L.: The histochemical demonstration of 0-
acylated sialic acid in gastro-intestinal mucins. Their association with the potassium hydroxide-
periodic acid-Schiff effect. J.Histochem.Cytochem. 22: 826-831 (1974).
Danscher, G.: Autometallography. A new technique for light and electron microscopic visual-
ization of metals in biological tissues (gold, silver, metal sulphides and metal selenides).
Histochemistty 81:331-335 (1984).
Deitch, A.D.: MicrospectrophotomeU"ic study of the binding of the anionic dye Naphthol Yellow
S by tissue sections and purified proteins. Lab.Invest 4:324-351 (1955).
Deitch, A.D., Wagner, D., and Richart, R.M.: Conditions infiuencing the intensity of the Feulgen
reaction. J. Histochem. Cytochem 16:371-379 (1968).
De Jong, A.S.H., van Kessel-van Vark, M., and Raap, A.K.: Sensitivity of various visualization
methods for peroxidase and alkaline phosphatase activity in immunoenzyme histochemistty.
Histochem.J. 17: 1119-1130 (1985).
De la Torre, J.C. and Surgeon, J.W.: A methodological approach to rapid and sensitive monoamine
histofiuorescence using a modified glyoxylic acid technique: The SPG method. Histochemistty
49:81-93 (1976).
DeLellis, R.A., ed.: Diagnostic immunohistochemistty. Masson Publishing USA, New York
(1981).
DeLellis, R.A., ed.: Advances in immunohistochemistty. Masson Publishing USA, New York
(1984).
536 References
DeLellis, RA, ed.: Advances in immunohistochemistry. Raven Press, New York (1988).
De Schepper, G.G., van Noorden, CJ.F. and Koperdraad, F.: A cytochemical method for mea-
suring enzyme activity in individual preovulatory mouse oocytes. J.Reprod.Fert. 74:709-716
(1985).
Desmet, VJ., Bullens, A.M., de Groote, J., and Heirwegh, K.P.M.: A new diazoreagent for specific
staining of conjugated bilirubin in tissue sections. J.Histochem.Cytochem. 16:419-427 (1968).
Dixon, K.C.: Hydrophobie lipids in injured cells. Histochem.J. 2:151-187 (1970).
Dixon, K.C.: Solubilization and cellular fatty change. Histochem. J. 5:363-387 (1973).
Dixon, M. and Webb, E.C.: Enzymes. 3rd ediL Academic Press, San Diego, CA (1979).
Dodds, R.A., Pitsillides, A.A. and Frost, G.T.B.: A quantitative eytochemical method for ornithine
decarboxylase activity. J.Histochem.Cytochem. 38:123-127 (1990).
Doebler, J.A., Wickersham, E.W. and Anthony, A.: Quantitative cytochemical analyses of ovarian
luteal and interstitial cell RNA in relation to plasma progestins during pseudopregnancy in
the raL Cell Biochem.Funct. 1:173-178 (1983).
Drexhage, HA., Wilders, M.M., Lens, J.W. and Boyages, S.: lmmunoglobulins die das Waehstum
von Thyreozyten stimulieren. Aktuel.Endokrinol.Stoffwechsell0,supp1.2:180-184 (1989).
Dubeau, L., Chandler, L.A., Gralow, J.R., Nichols, P.W. and Jones, PA: Southern blot analysis of
DNA extracted from formalin-fixed pathology specimens. Cancer Res. 46:2964-2969 (1986).
Dubowitz, V. and Brooke, M.H.: Museie biopsy: A modern approach. W. B. Saunders, London
(1973).
Duijndam, W.A.L. and van Duijn, P.: The inftuence of chromatin compactness on the stoichiom-
etry of the Feulgen-Schiff procedure studied in model films. J.Histochem.Cytochem. 23:882-
890 (1975).
Duijndam, W.A.L., van Duijn, P. and Riddersma, S.H.: Optical errors in scanning stage absorbance
cytophotometry. 11. Application of correction factors for residual distributional error, glare and
diffraetion error in practical cytophotometry. J.Histochem.Cytochem. 28:395-400 (1980a).
Duijndam, WAL., Smeulders, A.W.M., van Duijn, P. and Verweij, A.C.: Optieal errors in
scanning stage absorbance cytophotometry. I. Procedures for eorrecting apparent integrated
absorbance values for distributional, glare, and diffraction errors. J.Histochem.Cytochem.
28:388-394 (1980b).
Dunham, J.: Metabolism of the fracture callus, pp.213-220. In: J.R. Pattison, L. Bitensky and J.
Chayen, Eds.: Quantitative cytochemistry and its application. Academie Press, London (1979).
Dunnigan, H.G.: Chromatographic separation and photometric analysis of the components of Nile
Blue Sulphate. Stain Technol. 43:242-248 (1968a).
Dunnigan, H.G.: The use of Nile Blue Sulphate in the histochemical identification of phospho-
lipids. Stain Technol. 43:249-256 (1968b).
Duve, C. de, Pressman, B.C., Gianetto, R., Wattiaux, R. and Appelmans, F.: Tissue fraetionation
studies. VI. Intracellular distribution patterns of enzymes in rat liver tissue. Biochem.J. 60:604-
617 (1955).
Ealey, PA., Henderson, B., and Loveridge, N.: A quantitative study of peroxidase activity in
unfixed tissue sections of the guinea-pig thyroid gland. Histochem.J. 16:111-122 (1984).
Ebel, J.P., Colas, J., and Muller, S.: 11. Mise au point de methodes de detection cytochimiques
specifiques des polyphosphates. Exp.Cell Res. 15:21-28 (1958).
Edgar, G.W.F. and Donker, C.H.M.: InfIuence oflipid solvents on sphingolipids (sphingomyelins,
cerebrosides, gangliosides) in tissue sections. Acta neurol.Psychiat.Belg. 5:451-461 (1957).
Ehrlich, P.: Referate aus den Verein für innere Medicin zu Berlin. Deutsche Med.Wochenschr.
8:269 (1882). Cited in: Lison p.273 (1960).
Einarson, L.: A method for progressive selective staining of Nissl and nuc1ear substance in nerve
cells. Amer.J.Pathol. 8:295-307 (1932).
Einarson, L.: On the theory of gallocyanin-chrome alum staining and its application for quanti-
tative estimation of basophilia. A selective staining of exquisite progressivity. Acta Pathol.
Microbiol.Scand. 28:82-102 (1951).
Elftman, H.: Controlled chromation. J.Histochem.Cytochem. 2:1-8 (1954).
Elleder, M. and Lojda, Z.: Studies in lipid histochemistry IX. The specifieity of the Holczinger's
reaction for fatty acids. Histochemie 32:301-305 (1972).
References 537
Ellwart, J. and Dönner, P.: Effect of 5-fluoro-2' -deoxyuridine (FdUrd) on 5-bromo-2' -deoxyuridine
(BrdUrd) incorporation into DNA measured with a monoclonal BrdUrd antibody and by the
BrdUrd/Hoechst quenching effect. Cytometry 6:513-520 (1985).
EI-Sherif, G., van Noorden, C.J.F. and Bacsy, E.: Specificity of the metal-salt method for the
localization of Ca+ -AlPase activity studied in rat hypophysis tissue in a model system of
polyacrylamide gel films. Histochem.J. 22:51-56 (1990).
Engbrek, K., Johansen, K.S., and Jensen, M.E.: A new technique for Gram staining paraffin-
embedded tissue. J.Clin.Path. 32:187-190 (1979).
Ericsson, J.L.E. and Biberfeld, P.: Studies of aldehyde fixation. Fixation rates and their relation
to fine structure and some histochemical reactions in liver. Lab.Invest. 17:281-298 (1967).
Erlich, H.A.: PCR technology. Macmillan Press, London (1989).
Erlich, HA, Gelfand, D.H. and Saiki, R.K.: Specific DNA amplification. Nature 331:461-462
(1988).
Ernst, S.A.: Transport adenosine triphosphatase cytochemistry. 1. Biochemical characterization of
a cytochemical medium for the ultrastructural localization of ouabain-sensitive, potassium-
dependent phosphatase activity in the avian salt gland. J.Histochem.Cytochem. 20: 13-22
(1972a).
Ernst, S.A.: Transport adenosine triphosphatase cytochemistry. II. Cytochemica1 localization of
ouabain-sensitive, potassium-dependent phosphatase activity in the secretory epithelium of the
avian salt gland. J.Histochem.Cytochem. 20:23-38 (1972b).
Eränkö, 0.: Distribution of fluorescing islets, adrenaline and noradrenaline in the adrenal medulla
of the hamster. Acta Endocrinol. 18:174-179 (1955).
Eränkö, 0.: The practical histochemical demonstration of catecholamines by fonnaldehyde-in-
duced fluorescence. J.Roy. Micr.Soc. 87:259-276 (1967).
Eränkö, O. and Eränkö, L.: Differentiation of dopamine from noradrenaline in tissue sections by
microspectrofluorimetry. J. Histochem.Cytochem. 19:131-132 (1971).
Esmann, V.: The glycogen content of leukocytes from diabetic and non-diabetic subjects. Scand.J.
clin.Lab.Invest. 13:134-139 (1961).
Evans, A.W., Johnson, N.W., and Butcher, R.G.: A quantitative cytochemical study of three
oxidative enzymes during experimental oral carcinogenesis in the hamster. BritJ.Oral Surg.
18:3-16 (1980).
Fahirni, H.D.: Fine-structural cytochemical localization of peroxidatic activity of catalase. Tech-
niques of biochemical and biophysical morphology 2:197-245 (1975).
Fahirni, H.D.: Qualitative cytologica1 criteria for the validation of enzyme histochemical tech-
niques, pp.33-51. In: Trends in enzyme histochemistry and cytochemistry. Ciba Foundation
Symposium 73. Excerpta Medica, Amsterdarn (1980).
Falck, B. and Torp, A.: A fluorescence method for histochemical demonstration of noradrenalin
in the adrenal medulla. Med.Exp. 5:429-432 (1961).
Falck, B., Hillarp, N.-A., Thieme, G., and Torp, A.: Fluorescence of catechol amines and related
compounds condensed with fonnaldehyde. J.Histochem.Cytochem. 10:348-354 (1962).
Faulkner, R.R. and Lillie, R.D.: A buffer modification of the Warthin-Starry silver method for
spirochaetes in single paraffin sections. Stain Technol. 20:81-82 (1945).
Feigl, F.: Spot tests in organic analysis. 5th edit. Elsevier, Amsterdarn, (1958).
Felgenhauer, K. and Glenner, G.G.: Quantitation of tissue-bound renal amino-peptidase by a
microdensitometric technique. J.Histochem.Cytochem. 14:53-63 (1966).
Fenton, S., Clarkson, E., MacGregor, G., Alaghband-Zadeh, J. and De Wardener, H.E.: An assay
of the capacity of biological fluids to stimulate renal glucose-6-phosphate dehydrogenase
activity in vitro as a marker of their ability to inhibit sodium potassium-dependent adenosine
triphosphatase activity. J.Endocrinol. 94:99-110 (1982).
Ferguson, M.M. and MacPhee, G.B.: Kinetic study of llß-hydroxysteroid dehydrogenase in rat
submandibular salivary gland. Archs. oral Biol. 20:241-245 (1975).
Feulgen, R. and Rossenbeck, H.: Mikroskopisch-chemischer Nachweis einer Nucleinsäure von
Typus der Thymonucleinsäure und die darauf be,ruhende elektive Färbung von Zellkernen in
mikroskopischen Präparaten. Z.PhysioI.Chem. 135:203-248 (1924).
Filipe, M.l.: Value of histochemical reactions for mucosubstances in the diagnosis of certain
pathological conditions in the colon and rectum. Gut 10:577-586 (1969).
538 References
Filipe, M.I.: Malignant and inflamrnatory diseases of the gastrointestinal tract, pp. 126-135. 1n: M.I.
Filipe and B.D. Lake, Eds.: Histochernistry in pathology. Churchill Livingstone, Edinburgh
(1983).
Filipe, M.I., Abbas, S., and Greaves, P.: Mucin composition and CEA content in assessing ma-
lignant potential of colonic adenomas. VIIIth European Congress of Pathology, Helsinki, p.22
(abstract) (1981).
Firth, J.A.: Quantitative assessment of Na+ ,K+ -ATPase localization by direct and indirect p-
nitrophenyl phosphatase methods. J.Histochem.Cytochem. 35:507-513 (1987).
Flanders, A., Galbraith, W. and Marshall, P.N.: Microspectrophotometric studies ofRomanowsky
stained blood cells. IV. Maturation of myeloid and erythroid celliines in bone marrow. Stain
Techno!. 59:91-103 (1984).
Flitney, F.W.: The time course of the fixation of albumin by formaldehyde, glutaraldehyde,
acrolein and other aldehydes. J. Roy.Micr.Soc. 85:353-364 (1966).
Fontana, A.: Verfahren zur intensiven und raschen Färbung des Treponema pallidum und anderer
Spirochäten. Dermatol.Wochenschr. 55:1003-1004 (1912).
Francis, R.J.: Amyloid, pp.155-175. 1n: J.D. Bancroft, and A. Stevens, EdS.: Theory and practice
of histologica1 techniques. 3rd edit. Churchill-Livingstone, Edinburgh (1990).
Franks, F.: Biological freezing and cryofixation. 1n: Low temperature biological microscopy and
microanalysis. Royal Microscopical Society, Oxford, (1978).
Frederiks, W.M.: Some aspects of the value of Sudan black B in lipid histochernistry. Histochem-
istry 54:27-37 (1977).
Frederiks, W.M. and Marx, F.: Quantitative aspects of the histochernical tetrazoliurn salt reaction
on monoamine oxidase in rat liver. Histochem.J. 17:707-715 (1985).
Frederiks, W.M. and Marx, F.: A quantitative histochernical study of 5'-nucleotidase activity in
rat liver using the lead salt method and polyvinyl alcoho!. Histochem.J. 20:207-214 (1988).
Frederiks, W.M., Marx, F., and Bosch, K.S.: Diurnal variation in glycogen phosphorylase activity
in rat liver. A quantitative histochernical study. Eur.J.Cell Bio!. 43:339-341 (1987a).
Frederiks, W.M., Marx, F., and van Noorden, C.J.F.: Quantitative histochernical assessment of the
heterogeneity of glycogen phosphorylase activity in liver parenchyma of fasted rats using the
semipermeable membrane technique and the PAS reaction. HistochemJ.: 19:150-156 (1987b).
Frederiks, W.M., Marx, F., and van Noorden, C.J.F.: Histochemical demonstration of creatine
kinase activity using polyvinyl alcohol and auxiliary enzymes. Histochem.J. 19:529-532
(1987c).
Frederiks, W.M., Marx, F., Chamuleau, RA.F.M., van Noorden, C.J.F. and James, J.: Immuno-
cytochemical determination of ploidy class-dependent bromodeoxyuridine incorporation in rat
liver parenchymal cells after partial hepatectomy. Histochernistry 93:627-630 (1990).
Freifelder, D.: Physica1 biochemistry. Applications to biochernistry and molecular biology. W.H.
Freemann, San Francisco (1976).
Fritz, P., Hönes, J., Rautenstrauch, H., Lutz, D., Tuczek, H.V., Mischlinski, A., Saal, H.J., Klein,
C. and Laschner, W.: Imrnunophotometric quantification of extravascular irnmunoglobulin
deposits in the synovia! membrane of patients with osteoarthritis and rheurnatoid arthritis.
BasApp!.Histochem. 32:455-469 (1988).
Fukuda, M.: Application of the rnicrocomputer in multiparametric fluorescence cytophotometry.
J.Histochem.Cytochem. 31:241-243 (1983).
Fukuda, M., Böhm, N. and Fujita, S.: Cytophotometry and its biological application. Prog.Histo-
chem.Cytochem. 11/1:1-119 (1978).
Gahrton, G. and Yataganas, X.: Quantitative cytochernistry of glycogen in blood cells. Methods
and clinical application. Progr.Histochem.Cytochem. 9/1:1-30 (1976).
Galbraith, W. and Marshali, P.N.: Studies on Papanicolaou staining. I1I. Quantitative investigations
of orangeophilia and cyanophilia. Stain Techno!. 59:133-142 (1984).
Galbraith, W., MarshalI, P.N. and Bacus, J.W.: Microspectro-photometric studies of Romanowsky
stained blood cells. I. Subtraction analysis of a standardized procedure. J.Microsc. 119:313-
330 (1980).
Galbraith, W., Marshali, P.N., Lee, E.S. and Bacus, J.W.: Studies on Papanicolaou stain. I. Visible-
light spectra of stained cervical cells. Ana1.QuantCyto!. 1:160-168 (1979).
References 539
Gall, J.G. and Pardue, ML.: Fonnation and detection of RNA-DNA hybrid molecules in cyto-
logical preparations. Proc.Nad.Acad.Sci. 63:378-383 (1969).
Garrett, J.R., Kidd, A., Kyriacou, K. and Smith, R.E.: Use of different derivatives of D-val-Ieu-
arg for studying kallikrein activities in cat submandibular glands and saliva. HisoochemJ.
17:805-818 (1985).
Garvey, W., Fathi, A., Bigelow, F., Carpenter, B., and Jimenez, C.: A reliable method for demon-
strating Gram-positive and Gram-negative bacteria. Stain Technol. 61:251-253 (1986).
Geerts, A. and Roels, F.: Quantitation of cata1ase activity by microspectrophotometry after di-
aminobenzidine staining. Histochemistry 71:357-367 (1981).
Gilbert, D.M, Besser, G.M. Bitensky, L and Chayen, J.: Development of a cyrochemical section-
assay for thyroid stimulators. J.EndocrinoL 75:40P (1977).
Gill, J .E. and Jotz, M.M.: Further observations on the chemistry of pararosaniline-Feulgen staining.
Histochemistry 46:147-160 (1976).
Gi11i1and, G., Perrin, S. and Bunn, H. Franklin: Quantitative amplüication of mRNA using poly-
merase chain reaction. J.Cell.Biochem. Suppl. V, 13, WH:OOl (1989)
Glauert, A.M., Ed.: Practical methods in electron microscopy, Vol. 5. North-Holland Publishing
Company, Amsterdam (1977).
Gleich, G.L.: The eosinophil: New aspects and structure and function. J.Allergy Clin.Immunol.
60:73-82 (1977).
Glenner, G.G.: The histochemical demonstration of indole derivatives by the rosindole reaction
of E. Fischer. J.Histochem. Cytochem. 5:297-304 (1957).
Glenner, G.G.: The bases of the staining of amyloid fibers: Their physico-chemical nature and
the mechanism of their dye-substrate interaction. Progr.Histochem.Cytochem. 13/3 (1981).
Glenner, G.G. and Lillie, R.D.: The histochemical demonstration of indole derivatives by the post-
coupled p-dimethylaminobenzylidene reaction. J.Histochem.Cytochem. 5:279-296 (1957).
Glenner, G.G. and Lillie, R.D.: Observations on the diazotization-coupling reaction for the histo-
chemical demonstration of tyrosine: Metal chelation and forrnazan variants. J.Histochem.Cyto-
chem. 7:416-422 (1959).
Glick, D.: Trends in quantification in histochemistry and cytochemistry. Histochem.J. 13:227-240
(1981).
Glick, D., Engstrom, A. and Malmstrom, B.G.: A critical evaluation of quantitative histo- and
cyrochemical microscopic techniques. Science 114:253-258 (1951).
Goldstein, DJ.: Ionic and non-ionic bonds in staining, with special reference to the action of
urea and sodium chloride on the staining of elastic fibres and glycogen. Quart.J.Micr.Sci.
103:477-492 (1962).
Goldstein, DJ.: A microdensitometric method for the analysis of staining kinetics. J.Microsc.
113:331-343 (1980).
Goldstein, DJ.: Errors in microdensitometry. Histochem.J. 13: 251-267 (1981).
Golgi, C.: Sulla fina struttura dei bulbi olfattorii. Riv.Sper. Freniatr.Med.Legal. 1:405-425 (1875).
Golodetz, L. and Unna P.G.: Zur Chemie der Haut, m. Das Reduktions-vermögen der histologis-
chen Elemente der Haut Monatschr.Prakt.Dermatol. 48:149-166 (1909).
Goltzman, D., Henderson, B. and Loveridge, N.: Cytochemical bioassays of parathyroid hormone.
Characteristics of the assay and analysis of circulating hormonal forms. J.Clin.Invest 65: 1309-
1317 (1980).
Gomori, G.: Microtechnical demonstration of phosphatase in tissue sections. Proc.Soc.Exp.Biol.
Med. 41:23-26 (1939).
Gomori, G.: A new histochemica1 test for glycogen and mucin. Amer.J.Clin.Pathol. 16:177-179
(1946).
Gomori, G.: Aldehyde Fuchsin: A new stain for elastic tissue. Amer.J.Clin.Pathol. 10:665-666
(1950a).
Gomori, G.: An improved technic for acid phosphatase. Stain Technol. 15:81-85 (1950b).
Gonchoroff, N.J., Greipp, P.R., Kyle, RA. and Katzmann, J.A.: A monoclonal antibody reactive
with 5-bromo-2-deoxyuridine that does not require DNA denaturation. Cytometry 6:506-512
(1985).
540 References
Gooding, P.E., Chayen, J., Sawyer, B. and Slater, T.F.: Cytochrome P-450 disttibution in rat liver
and the effect of sodium phenobarbitone administration. Chem.-BioI.Interact. 20: 299-310
(1978).
Gordon, H. and Sweets, H.H.)r.: A simple method for the silver impregnation of reticulum.
Amer.J.Pathol. 15:545-552 (1936).
Gordon, M. and Robertson, W.R.: The application of continuous monitoring microdensitometry
to an analysis of NAD+ binding and 3ß-hydroxy-D.5 -steroid dehydrogenase activity in the
regressing corpus luteum of the pro-oestrous rat ovary. Histochem.J. 18:41-44 (1986).
Gossrau, R.: Tetrazoliummethoden zum histochemischen Hydrolasennachweis. Histochemistry
58:203-218 (1978).
Gossrau, R.: Cytochemistry of membrane proteases. Histochem.J. 17:737-771 (1985).
Gossrau, R. and Lojda, Z.: Study on dipeptidylpeptidase 11 (OPP 11). Histochemistry 70:53-76
(1980).
Gossrau, R., van Noorden, C.J.F. and Frederiks, W.M.: Enhanced light microscopic visualization
of oxidase activity with the cerium capture method. Histochemistry 92:349-353 (1989).
Gossrau, R., Graf, R., Rulmke, M. and Hauski, C.: Proteases in the human full-term placenta.
Histochemistry 86:405-413 (1987).
Graf, M., Baici, A. and Strauli, P.: Histochemicallocalization of cathepsin B at the invasion front
of the rabbit V2 carcinoma. Lab.Invesl 45:587-596 (1981).
Graham, R.C. and Karnovsky, M.J.: The early stages of absorption of injected horseradish per-
oxidase in the proximal tubules of mouse kidney: Ultrastructural cytochemistry by a new
technique. J.Histochem.Cytochem. 14:291-302 (1966).
Gram, C.: Ueber die isolierte Färbung der Schizomyceten in Schnitt- und Trockenpräparaten.
Forschr.Med. 2:185 (1884).
Grange, J.M.: The mycobacteria. In: G. Wllson, A. Miles, and M.T. Parker, Eds.: Topleyand
Wilson's Principles of bacteriology, virology and immunity, vol. 11: Systematic bacteriology,
7th edit. Edward Amold Ltd., London (1983).
Gratzner, H.G.: Monoclonal antibody to 5-bromo- and 5-iododeoxyuridine: A new reagent for
detection of DNA replication. Science 218:474-475 (1982).
Gratzner; H.G., Leif, R.C., Ingram, D.J. and Castro, A.: The use of antibody specific for bromod-
eoxyuridine for the immunofluorescent determination of DNA replication in single cells and
chromosomes. Exptl.Cell Res. 95:88-94 (1975).
Gray, P., Day, M.W., Hayweiser, L.J., and Nevsimal, C.: Oxazine dyes. 2. Celestine blue B,
gallocyanin and gallamin blue with mordants other than ferric alum. Stain Technol. 32:161-
165 (1957).
Grimelius, L.: A silver nitrate stain for a2 cells in human pancreatic islets. Acta Soc.Med.Ups.
73:243-270 (1968).
Grocott, R.G.: A stain for fungi in tissue sections and smears. Amer.J.Clin.Pathol. 25:975-979
(1955).
Grynkiewicz, G., Poenie, M., and Tsien, R.Y.: A new generation of Ca2+ indicators with greatly
improved fluorescence properties. J. Biol.Chem. 260:3440-3450 (1985).
Gundersen, H.J.G., Bendtsen, T.F., Korbo, L., Marcussen, N., M~ller, A., Nielsen, K., Nyengaard,
J.R., Pakkenberg, B., S~rensen, F.B., and West, M.J.: Some new, simple and efficient stereolog-
ical methods and their use in pathological research and diagnosis. Acta Pathol.Microbiol.lm-
munol.Scand. 96:379-394 (1988a).
Gundersen, H.J.G., Bagger, P., Bendtsen, T .F., Evans, S.M., Korbo, L., Marcussen, N., M~ller, A.,
Nielsen, K., Nyengaard, J.R., Pakkenberg, B., S~rensen, F.B., Vesterby, A. and West, M.J.:
The new stereological tools: Disector, fractionator, nucleator and point sampled intercepts
and their use in pathological research and diagnosis. Acta Pathol.Microbiol. lmmunol.Scand.
96:857-S81 (1988b).
Gutschmidt, S.: "In situ" determinations of apparent KM and Vmax of brush border disacchari-
dases along the villi of normal human jejunal biopsy specimens. A quantitative histochemical
study. Histochemistry 71:451-462 (1981).
Gutschmidt, S., Emde, C.,and Riecken, E.O.: Quantification of a-glucosidases along the villus of
the small intestine in man. Introduction of a computerized histochemical method. Histochem-
istry 67:85-97 (1980a).
References 541
Gutschrnidt, S., Kaul, W., and Riecken, E.O.: A quantitative histochemical technique for the char-
acterisation of a-glucosidases in the brush-border membrane of rat jejunum. Histochemistry
63:81-101 (1979).
Gutschmidt, S., Lange, U., and Riecken, E.O.: Kinetic characterization of unspecific alkaline
phosphatases at different villus sites of rat jejunum. Histochemistry 69:189-202 (1980b).
Gutschmidt, S., Lange, U. and Riecken, E.O.: "In situ"-measurements of protein contents in the
brush border region along rat jejunal villi and their correlations with four enzyme activities.
Histochemistry 72:467-479 (1981).
Haase, A.T.: Analysis of viral infections by in situ hybridization. J.Histochem.Cytochem. 34:27-
32 (1986).
Häkanson, R., Juhlin, L., Owman, C., and Sporrong, B.: Histochemistry of histamine: microspec-
trofluorometric characterization of the fluorophores induced by o-phthaldialdehyde. J.Histo-
chem.Cytochem. 18:93-99 (1970).
Halkjrer-Kristensen, J. and Ingemann-Hansen, T.: Quantitation of the PAS reaction in skeletal
muscle of man. Histochem.J. 11:737-746 (1979).
Hamada, S.: A double labeling technique combining 3H-thymidine autoradiography with BrdU
immunocytochemistry. Acta Histochem.Cytochem. 18:267-269 (1985).
Hanker, J.S.: Oxidoreductases, pp.I-139. In: M.A. Hayat, Ed.: Electron microscopy of enzymes.
Principles and methods. Vol. 4. Van Nostrand Reinhold Company, New York (1975).
Harada, K.: The nature of mycobacterial acid-fastness. Stain Technol. 51:255-260 (1976).
Hardonk, M.J.: The use of phenazine methosulphate as an electron carrier in the histochemical
demonstration of dehydrogenases. Histochemie 4:563-568 (1965).
Harkema, J.R., Plopper, C.G., Hyde, D.M. and St. George, J.A.: Regional differences in quan-
tities of histochemically detectable mucosubstances in nasal, paranasal, and nasopharyngeal
epithelium of the bonnet monkey. J.Histochem.Cytochem. 35:279-286 (1987).
Hartmann, F. and Fleck, U.: Vergleichende chemische und histologische Analyse der Leberver-
fettung. Klin.Wochenschr. 30:652-{)54 (1952).
Haugland, R.P.: Molecular Probes. Handbook of Fluorescent probes and research chemicals.
Molecular Probes, Inc. Eugene, Oregon, USA (1989).
Hayat, M.A.: Principles and techniques of electron microscopy: Biological applications. Vol. I.
Van Nostrand-Reinhold, New York, (1970).
Hayes, E.R.: A rigorous re-definition of the plasmal reaction. Stain Technol. 24:19-23 (1949).
Hayhoe, F.G.J. and Quaglino, D.: Haematological cytochemistry. Churchill-Livingstone, Edin-
burgh (1980).
Henderson, B.: Quantitative cytochemical measurement of glyceraldehyde 3-phosphate dehydro-
genase activity. Histochemistry 48:191-204 (1976).
Henderson, B.: Quantitative dehydrogenase cytochemistry in the study of cellular metabolism
with special reference to rheumatoid arthritis, pp,83-98. In: J.R. Pattison, L. Bitensky and
J. Chayen, Eds.: Quantitative cytochemistry and its applications. Academic Press, London
(1979).
Henderson, B.: Increase in the activity of lysosomal acid hydrolases in the chondrocytes of arthritic
joints of rabbits with experimental allergic arthritis. Histochem.J. 16:287-293 (1984).
Henderson, B. and Loveridge, N.: Intermediate electron-acceptors in quantitative cytochem-
istry. Comparison of phenazine methosulphate and Meldola Blue. Histochemistry 72:617-{)23
(1981).
Henderson, B., Bitensky, L. and Chayen, J.: Altered phospholipids in human rheumatoid synovio-
cytes. Ann.Rheum.Dis. 37:24-29 (1978a).
Henderson, B., Bitensky, L. and Chayen, J.: Mitochondrial oxidative activity in human rheumatoid
synoviallining cells. Ann.Rheum.Dis. 37:548-551 (1978b).
Henderson, B., Johnstone, 1.1. and Chayen, J.: 5' Nucleotidase in the human synoviallining in
rheumatoid arthritis. Ann.Rheum.Dis. 39:248-252 (1980).
Henderson, B., Loveridge, N., and Robertson, W.R.: A quantitative study ofthe effects of different
grades of polyvinyl alcohol on the activities of certain enzymes in unfixed tissue sections.
HistochemJ. 10:453-463 (1978c).
Henderson, B., Glynn, L.E., Bitensky, L. and Chayen, J.: Evidence for cell division in acutely
inflamed rabbit joints. Ann.Rheum.Dis. 40:177-181 (1981).
542 References
Herzog, R.E.: Cytophotometric determinations of DNA, histone, arginine, lysine, and their con-
centrations in eu- and heterochrornatin of the cell nucleus of dysplasias, carcinoma in situ,
and carcinoma of the human cervix uteri. Arch.Gynecol. 231:91-98 (1982).
Hess, J., Jensen, C.V. and Diemer, N.H.: Calcium-imaging with Fura-2 in isolated cerebral mi-
crovessels. Acta Histochem. 87:107-114 (1989).
Hildebrand, R. and Schleicher, A.: Image analysis of the histochemical demonstration of glucose-
6-phosphatase activity in rat liver. Histochemistry 86:181-190 (1986).
Hillarp, N.-i\.. and Hökfelt, B.: Evidence of adrenaline and noradrenaline in separate adrenal
medullary cells. Acta Physiol. Scand. 30:5~8 (1953).
Holczinger, L.: Histochemischer Nachweis freier Fettsaüren. Acta Histochem. (Jena) 8:167-175
(1959).
Hooghwinkel, G.J.M. and Smits, G.: The specificity of the periodic acid-Schiff technique studied
by a quantitative test tuhe method. J.Histochem.Cytochem. 5:12(}-126 (1957).
Hopwood, D.: Some aspects of fixation with glutaraldehyde. A biochemical and histochemical
comparison of the effects of formaldehyde and glutaraldehyde fixation on various enzymes and
glycogen with a note on penetration of glutaraldehyde into liver. J.AnaL 101:83-92 (1967).
Hopwood, D.: Fixatives and fixation: A review. Histochem.J. 1:323-360 (1969).
Horobin, R.W.: Histochemistry. Gustav Fischer, Stuttgart and Butterworths, London (1982a).
Horobin, R.W.: Selection of optimum tetrazoliurn salts for use in histochemistry: the value of
structure-staining correlations. Histochem.J. 14:301-310 (1982b).
Horobin, R.W.: Thoughts arising from Cowden and Curtis's letter. Histochem.J. 16:124a-1241
(1984).
Horobin, R.W. and Bennion, P.J.: The interrelation of the size and substantivity of dyes; the role
of van der Waals attractions and hydrophobic bonding in biological staining. Histochemie
33:191-204 (1973).
Horobin, R.W. and Flemming, L.: Is Mercury Orange a selective stain for thiols? Histochem.J.
14: 1003-1006 (1982).
Horobin, R.W. and Flemming, L.: One-bath trichrome staining: investigation of a general mech-
anism based on a structure-staining correlation analysis. Histochem.J. 20:29-34 (1988).
Horobin, R.W. and Flemming, L.: 'Trouble-shooting' microwave accelerated procedures in his-
tology and histochemistry: understanding and dealing with artefacts, errors and hazards. His-
tochem.J. 22:371-376 (1990).
Horobin, R.W. and Goldstein, D.J.: The influence of salt on the staining of tissue sections with ba-
sic dyes: An investigation into the general applicability of the critical electrolyte concentration
theory. Histochem.J. 6:599-609 (1974).
Horobin, R.W. and Murgatroyd, L.G.: The composition and properties of gallocyanin-chrome
alum stains. HistochemJ. 1:36-54 (1968).
Hougaard, D.M. and Larsson, L.-I.: Towards microfluorometric quantitation of polyamines in
situ. Relationship hetween cellular polyamine concentration and fluorescence yield of the
formaldehyde fluorescamine method. Histochemistry 93:359-362 (1990).
Howie, A.J., Burnett, D. and Crocker, J.: The distribution of cathepsin B in human tissues. J.Path.
145:307-314 (1985).
H~yer, P.E.: Histoenzymology of the human ovary: dehydrogenases directly involved in steroido-
genesis, pp.52-67. In: P.M. Motta and E.S.E. Hafez, Eds.: Biology of the ovary. Martinus
Nijhoff, The Hague (1980).
H~yer, P.E.: A quantitative cytochemical study of estradiol 17ß-dehydrogenase in the human
term placenta. J.Histochem.Cytochem. 36:875 (1988).
H~yer, P.E. and Andersen, H.: Specificity in steroid histochemistry, with special reference to
the use of steroid solvents. Distribution of llß-hydroxysteroid dehydrogenase in kidney and
thymus from the mouse. Histochemie 24:292-306 (1970).
H~yer, P.E. and Andersen, H.: Histochemistry of 3ß-hYdroxysteroid dehydrogenase in rat ovary.
1. A methodological study. Histochemistry 51:167-193 (1977).
Hl'lyer, P.E. and Byskov, A.G.: A quantitative study of .t,5,3ß-hydroxysteroid dehydrogenase
activity in the rete system of the immature mouse ovary. In: A.G. Byskov and H. Peters, Eds.:
International Congress Series No. 559. Development and function of reproductive organs,
pp.216-224. Excerpta Medica, Amsterdam-Oxford-Princeton (1981).
References 543
H~yer, P.E, Lyon, H., Jakobsen, P., and Andersen, A.P.: Standardized Methyl Green-Pyronin Y
procedures using pure dyes. Histochem.J. 18:90-94 (1986).
H~yer, P.E., Wohlrab, F. and Stoward, P .1.: Dehydrogenases involved in sterojd and prostaglandin
metabolism. Chapter 23. In: PJ. Stoward, Ed.: Pearse, AG.E.: Histochemistry: Theoretical
and applied. 4th edit. Vol. ill. Churchill Livingstone, Edinburgh In press (1991).
Hukill, P. and Putt, F.A.: A specific stain for iron using 4,7-diphenyl-l,1O-phenanthroline (batho-
phenanthroline). J.Histochem. Cytochem. 10:490-494 (1962).
Husain, O.A.N.: Cytology: preparative techniques, pp.296-300. In: Filipe, M.I. and Lake, B.D.,
Eds.: Histochemistry in pathology. Churchill Livingstone, Edinburgh (1983).
Husain, O.A.N. and Watts, K.C.: Rapid demonstration of nucleic acids using "oxidized" gallo-
cyanin and chromic potassium sulphate: methods and applications. J.Clin.Pathol. 37:99-101
(1984).
Hussein, K.A., Milne, G., and Hopwood, D.:Glycosaminoglycans in human gallbladder basenient
membrane: nature and quantitative changes in chronic cholecystitis. Histochem.J. 20:449-454
(1988).
Ibrahim, K.S., Husain, O.A.N., Bitensky, L. and Chayen, J.: A modified tetrazolium reaction for
identifying malignant cells from gastric and colonic cancer. J.clin.Pathol. 36:133-136 (1983).
Illsley, N.P. and Verkman, AS.: Membrane chloride transport measured using a chloride-sensitive
ßuorescent probe. Biochemistry 26:1215-1219 (1987).
Imes, N.K.,Jr., Sanders, D.C., Crane, C.R., and Clark, G.: Assaying actual hematein content of
commercial hematoxylins and hemateins. Stain Technol. 44:167-172 (1969).
Impraim, C.D., Saiki, R.K., Erlich, H.A. and Teplitz, R.L.: Analysis of DNA extracted from
formalin-fixed, paraffin-embedded tissue by enzymatic amplification and hybridization with
sequence-specific oligonucleotides. Biochem.Biophys.Res.Comm. 142:710-716 (1987).
Innis, M.A., Myambo, K.B., Gelfand, D.H. and Brow, M.AD.: DNA sequencing with Thermus
aquaticus DNA polymerase and direct sequencing of polymerase chain reaction-amplified
DNA. Proc.Nat.Acad.Sci. 85:9436-9440 (1988).
Innis, M.A., Gelfand, D.H., Sninsky, J.J. and White, T.J.: PCR protocols. Academic Press, New
York (1989).
Inoue, S.: Video microscopy. Plenum Press, New York (1986).
IUB: Enzyme nomenclature. Academic Press, San Diego, California, (1984).
Jacobsen, N.O.: The histochemical localization of lactic dehydrogenase isoenzymes in the rat
nephron by means of an improved polyvinyl alcohol method. Histochemie 20:250-265 (1969).
Jawetz, E., Melnick, J.L., Adelberg, E.A. Bmoks, G.F., Butel, J.S and Ornston, L.N.: Medical
microbiology, 18th ed. Prentice-Hall International Inc., London (1989).
Jedrzejewski, K. and Kugler, P.: Peptidases in the kidney and urine of rats after castration.
Histochemistry 74:63-84 (1982).
Jeffrey, I.J.M., Mosley, S.M., Jones, C.J.P., and Stoddart, R.W.: Proteolysis and lectin histochem-
istry. Histochem.J. 19:269-275 (1987).
Jöbsis, A.C.: Prostate, pp.263-267. In: Filipe, M.1. and Lake, B.D., Eds.: Histochemistry in pathol-
ogy. Churchill Livingstone, Edinburgh (1983).
Jobst, K. and Sandritter, W.: Über den quantitativen histochemischen Nachweis von basischen
Kernproteinen mit Gallocyaninchromalaun. Histochemie 4:277-285 (1964).
Johannes, M.-L. and Klessen, C.:AlcianbluejPAS or PAS/Alcianblue? Remarks on a classical
technic used in carbohydrate histochemistry. Histochemistry 80:129-132 (1984).
John, H.A., Birnstiel, M.L. and Jones, K.W.: RNA-DNA hybrids at the cytologicallevel. Nature
223:582-587 (1969).
Johnson, M.A.: Skeletal muscle, pp.89-113. In: Filipe, M.I. and Lake, B.D., Eds.: Histochemistry
in pathology. Churchill Livingstone, Edinburgh (1983).
Johnston, K.I. and Ashford, AE.: A simultaneous-coupling azo dye method for the quantitative
assay of esterase using a-naphthyl acetate as substrate. HistochemJ. 12:221-234 (1980).
Jones, D.: The reaction of formaldehyde with unsaturated fatty acids during histological fixation.
HistochemJ. 1:459-491 (1969).
Jones, D.: Reactions of aldehydes with unsaturated fatty acids during histological fixation, pp.l-
45. In: PJ. Stoward, Ed.: Fixation in histochemistry. Chapman and Hall, London (1973).
544 References
Jonges, G.N. and van Noorden, C.J.F.: In situ kinetic parameters of glucose-6-phosphate de-
hydrogenase and phosphogluconate dehydrogenase in different areas of the rat liver acinus.
Histochem.J. 21:585-594 (1989).
Jonsson, G.: Fluorescence methods for the histochemical demonstration of monoamines. VII. Flu-
orescence studies on biogenic monoamines and related compounds condensed with fonnalde-
hyde. Histochemie 8:288-296 (1967).
Juarranz, A. Ferrer, J.M., Tato, A., Caiiete, M., and Stockert, J.C.: Metachromatic staining and
electron dense reaction of glycosaminoglycans by means of Cuprolinic Blue. Histochem.J.
19: 1-6 (1987).
Junqueira, L.C.U., Bignolas, G., and Brentani, R.R.: Picrosirius staining plus polarization mi-
croscopy, a speci1ic method for collagen detection in tissue sections. HistochemJ.ll:447-455
(1979).
Kasten, FR.: The chemistry of Schiff's reagent. Int.Rev.Cytol. 10:1-100 (1960).
Kashiwa, H.K. and Atkinson, W.B.: The applicability of a new Schiff base, glyoxal bis (2-
hydroxyanil), for the cytochemical localisation of ionic calcium. J.Histochem.Cytochem.
11:258 (1963).
Kayser, K. and Bubenzer, J.: Microwave-assisted staining procedures in routine histopathology.
HistochemJ. 22:365-370 (1990).
Kiefer, G.: Recent developments in Gallocyanine-Chrome Alum staining, pp.199-208. In: G.L.
Wied and G.F. Bahr, Eds.: Introduction to Quantitative Cytochemistry. Academic Press, New
York, London (1970).
Kiefer, R., Kiefer, G., Salm, R., Rossner, R., and Sandritter, W.: A method for the quantitative
evaluation of eu- and heterochromatin in interphase nuclei using cytophotometry and pattern
analysis. Beitr.Path. 150:163-173 (1973).
Kieman, J.A.: Histological & histochemical methods: Theory & Practice. Pergamon Press, Oxford
(1981).
Kiviranta, I., Jurvelin, J., Tammi, M., SälImällen, A.-M., and Helminen, HJ.: Microspectrophoto-
metric quantitation of glycosaminoglycans in articular cartilage sections stained with Safranin
O. Histochemistry 82:249-255 (1985a).
Kiviranta, 1., Tammi, M., Jurvelin, J., SälImällen, A.-M., and Helminen, H.J.: Demonstration of
chondroitin sulphate and glycoproteins in articular cartilage matrix using periodic acid-Schiff
(PAS) method. Histochemistry 83:303-306 (1985b).
Kjellstrand, P.T.T.: Control of extraction of deoxyribonucleic acid and apurinic acid by polyethy-
lene glycol in Feulgen hydrolysis. J.Histochem.Cytochem. 25:371-375 (1977).
Kjellstrand, P.T.T. and Lamm, CJ.: A model of the breakdown and removal of the chromatin
components during Feulgen acid hydrolysis. HistochemJ. 8:419-430 (1976).
Klunk, W.E., Pettegrew, J.W. and Abraham, D.J.: Two simple methods for quantifying low-affinity
dye-substrate binding. J.Histochem.Cytochem. 37:1293-1297 (1989a).
Klunk, W.E., Pettegrew, J.W. and Abraham, D.J.: Quantitative evaluation of Congo red bind-
ing to amyloid-like proteins with a beta-pleated sheet conformation. J.Histochem.Cytochem.
37:1273-1281 (1989b).
Koch, J.E., Kolvraa, S., Petersen, K.B., Gregersen, N. and Bolund, L.: Oligonucleotide-priming
methods for the chromosome-specific labeling of alpha satelite DNA in sim-. Chromosoma
98: 259-265 (1989).
Koch, R.: Die Ätiologie der Tuberkulose. BerLKlin.Wochenschr. 19:221 (1882).
Kok, L.P. and Boon, M.E.: Microwave in microscopy. J.Microsc. 158:291-322 (1990).
Konradi, C., Kornhuber, J., Froelich, L., Fritze, J., Heinsen, H., Beckmann, H., Schulz, E., and
Riederer, P.: Demonstration of monoamine oxidase-A and -B in the human brain stem by a
histochemical technique. Neuroscience 33:383-400 (1989).
Krapf, R., Dlsley, N.P., Tseng, H.C., and Verkman, A.S.: Structure-activity relationships of
chloride-sensitive fluorescent indicators for biological application. Anal.Biochem. 169:142-
150 (1988).
Krawetz, SA., Pon, R.T. and Dixon, G.H.: Increased efficiency of the Taq polymerase catalyzed
chain reaction. Nucl.Acid.Res. 17:819 (1989).
Krikelis, H. and Smith, AA.: Palladium toning of silver-impregnated reticular fibers. Stain Tech-
nol. 63:97-100 (1988).
References 545
Lillie, R.D. and Fullmer, H.M.: Histopathologie technie and practical histochemistry. 3rd edil
McGraw-HilI Book Co., New York, (1976).
Lillie, R.D. and Pizzolato, P.: Histochemical azo coupling reactions of the pigments of obstructive
icterus and of hematoidin. I.Histochem.Cytochem. 17:738-748 (1969).
Lillie, R.D. and Pizzolato, P.: Histochemical azocoupling reactions of the pigments of obstructive
icterus and of hematoidin. 11. Effect of blockade, bleaching and extraction procedures on the
pigments and their azo derivatives. I.Histochem.Cytochem. 18:75-79 (1970).
Lillie, R.D., pizzolato, P., Dessauer, H.C., and Donaldson, P.T.: Histochemical reactions at tissue
arginine sites with alkaline solutions of ß-naphthoquinone-4-sodium sulfonate and other 0-
quinones and oxidized o-diphenols. I.Histochem.Cytochem. 18:487-497 (1971).
Lillie, R.D., pizzolato, P., Henderson, R., and Donaldson, P.T.: The inftuence of formaldehyde
fixation time on the rate of deamination of erythrocytes and spinal cord proteins. Histochemie
24:156-166 (1970).
Lillie, R.D., Pizzolato, P., Vacca, LL., Catalano, R.A., and Donaldson, P.T.: Use of the diazo-
sulfanilic acid, pH 1 Azure A sequence on adrenal medulla and the effects of prior oxidation
and reduction on the reaction in several species. I.Histochem. Cytochem; 21:448-454 (1973).
Lison, L.: Histochemie et Cytochemie Animales. 3rd edil Gauthier Villars, Paris, (1960).
Lojda, Z.: Malabsorption. In: Filipe, M.1. and Lake, B.D., Eds.: Histochemistry in pathology,
pp.136-144. Churchill Livingstone, Edinburgh (1983).
Lojda, Z.: The importance of protease histochemistry in pathology. Histochem.I. 17:1063-1089
(1985).
Lojda, Z. and Gossrau, R.: Study on aminopeptidase A. Histochemistry 67: 267-290 (1980).
Lojda, Z., Gossrau, R., and Schiebler, T .H.: Enzyme histochemistry. A laboratory manual. Springer
Verlag, Berlin (1979).
Lojda, Z, van der Ploeg, M. and van Duijn, P.: Phosphates of the naphthol AS series in the
quantitative determination of alkaline and acid phosphatase activities "in situ" studies in
polyacrylamide membrane model systems and by cytospectro-photometry. Histochemie 11: 13-
32 (1967).
Lomax. R.B. and Robertson, W.R.: Mitochondrial a-glycerol phosphate dehydrogenase activity in
IIA fibres of the rat lateral gastrocnemius muscle; the effect of Ca2+ and ATP. HistochemJ.
22:119-124 (1990).
Lomax, R.B., Daglish, A., Taylor, R.I., Gordon, M.T. and Robertson, W.R.: Rapid continuous
monitoring of enzyme activity in tissue sections: experience with the M85A and Zeiss UMSP-
30 systems. Histochem.I.21:595-599 (1989).
Loren, 1., Björklund, A., Falck, B., and Lindvall, 0.: An improved histofluorescence procedure
for freeze-dried paraffin-embedded tissue based on combined formaldehyde-glyoxylic acid
perfusion with high magnesium content and acid pH. Histochemistry 49:177-192 (1976).
Loveridge, N.: A quantitative cytochemical method for measuring carbonie anhydrase activity.
Histochem.I. 10:361-372 (1978).
Loveridge, N. and Robertson, W.R.: Stimulation of adrenal 5-ene, 3ß-hydroxysteroid dehydroge-
nase by corticotrophin in vitra. I.Endocrinol. 78:457-458 (1978).
Loveridge, N., Alaghband-Zadeh, I., Daly, I.R. and Chayen, I.: The nature of the redox change
measured in the cytochemical bioassay of corticotrophin. I.Endocrinol. 67:28P (1975).
Loveridge, N., Bloom, S.R., Welboum, R.B. and Chayen, I.: Quantitative cytochemical estimation
of the effect of pentagastrin (0.005-5 pg/ml) and of plasma gastrin on the Guinea-pig fundus
in vitro. Clin.Endocrinol. 3:389-396 (1974).
Loveridge, N., Hoile, R., Iohnson, A.G. and Chayen, I.: The cytochemical measurement of gastrin-
like activity. IEndocrinol. 77:40P-41P (1978).
Loveridge, N., Hoile, R.W., Iohnson, A.G., Gardner,J.D. and Chayen, I.: The cytochemical
section-bioassay of gastrin-like activity. I.lmmunoassay 1:195-209 (1980).
Lowry, O.H., Rosebrough, N.I., Farr, A.L. and Randali, R.I.: Protein measurement with the Folin
phenol reagent. I.Biol.Chem. 193:265-275 (1951).
Lukaszyk, A.: The use of tris (hydroxymethyl) aminomethane as a copper complexing agent in the
histochemical method for a-glycerophosphate-ferricyanide oxidoreductase activity detection.
Histochemie 27:366-369 (1971).
References 547
Lund-Hansen, T., H~yer, P.E., and Andersen, H.: A quantitative cytochemical assay of ß-galacto-
sidase in single cultured human skin fibroblasts. Histochernistty 81:321-330 (1984).
Lyon, H. and Prent~, P.: Alcian blue-silver impregnation. A method differentiating between acid
carbohydrates, reticulin fibres and collagen fibres. Acta Path.Microbiol.Scand. SectA 81:6--8
(1973).
Lyon, H. and Prent~, P.: Aldehyde Fuchsin staining of pancreatic B cells. Reproducible high-
contrast staining of formalin-fixed and paraffin-embedded material. HistochemJ. 12:97-105
(1980).
Lyon, H., Breksted, M., and M~ller, M.: Method for direct induction of serotonin fluorescence
in a carcinoid tumor after embedding in hydroxyethyl methacrylate. Lab.Invest. 45:316-320
(1981).
Lyon, H., Breksted, M., and M~ller, M.: Induced fluorescence of serotonin in paraffin sections
from carcinoid tumors. Brain Res. Bull. 9:751-756 (1982).
Lyon, H., Schulte, E. and H~yer, P .E.: The correlation between uptake of Methyl Green and
Feulgen staining intensity of cell nuclei. An image analysis study. Histochem.J. 21:508-513
(1989).
Lyon, H., Jakobsen, P., Clausen, P .P., and Andersen, A.P.: Methyl Green-Pyronin staining with
pure Pyronin Y. Histochem.J. 15:605-606 (1983).
Mager, M., McNary, W.F.,Jr., and Lionetti, P.: The histochemical detection of zinc. J.Histochem.
Cytochem. 1:493-504 (1953).
Magun, B.E. and Kelly, J.W.: A new fluorescent method with phenanthrenequinone for the his-
tochemical demonstration of arginine residues in tissues. J.Histochem.Cytochem. 17:821--827
(1969).
Mai, J.K., Schmidt-Kastner, R., and Tefett, H.-B.: Use of acridine orange for histologic analysis
of the central nervous system. J.Histochem.Cytochem. 32:97-104 (1984).
Maniatis, T., Fritsch, E.F. and Sambrook, J.: Molecular cloning, a laboratory manual. Cold Spring
Harbour Lab. 2nd edit. (1989).
Marsha1I, P.N.: Thin-Iayer chromatography of Sudan dyes. J.Chromatogr. 136:353-357 (1977).
Marsha1I, P.N. and Galbraith, W.: Microspectrophotometric studies of Romanowsky stained blood
cells. III. The action of methylene blue and azure B. Stain Techno!. 59: 17-36 (1984a).
Marsha1I, P.N. and Galbraith, W.: The calculation of differences in the colours of histological
objects. HistochemJ. 16:211-218 (1984b).
Marsha1I, P.N. and Horobin, R.W.: The chemical nature of the gallocyanin-chrome alum staining
complex. Stain Technol. 47: 155-161 (1972).
Marsha1I, P.N., Galbraith, W. and Bacus, J.W.: Studies on Papanicolaou staining. 11. Quantitation
of dye components bound to cervical cells. Anal.QuantCytol. 1:169-178 (1979).
Marsha1I, P.N., Galbraith, W., Navarro, E.F. and Bacus, J.W.: Microspectrophotometric studies
of Romanowsky stained blood cells. 11. Comparison of the performance of two standardized
stains. J.Microsc. 124:197-210 (1981).
Masson, M P.: La glande endocrine de l'intestin chez l'homme. C.R.Acad.Sci. 158:59-61 (1914).
Mayall, B.H.: Deoxyribonucleic acid cytophotometty of stained human leukocytes. 1. Differences
among cell types. J.Histochem. Cytochem. 17:249-257 (1969).
Mayall, B.H. and Mendelsohn, M.L.: Errors in absorption cytophotometty: Some theoretical and
practical considerations, p.89. In: G.L. Wied and G.F. Bahr, Eds.: Introduction to quantitative
cytochemistty-I1. Academic Press, New York (1970).
Mayer, P.: Über das Färben mit Hämatoxylin. Min.Zoo!.StaLNeapeI10:170-186 (1891).
McAllister, H.A. and Rock, D.L.: Comparative usefulness of tissue fixatives for in situ viral
nucleic acid hybridization. J .Histochem.Cytochem. 33: 1026-1032 (1985).
McAuliffe. W.G.: A note on the purification of Alcian blue. Stain Techno!. 58:374-376 (1983).
McDonald, J.K.: An overview of protease specificity and catalytic mechanisms: aspects related
to nomenclature and classification. Histochem.J. 17:773-785 (1985).
McGadey, J.: A tetrazolium method for non-specific alkaline phosphatase. Histochemie 23:180-
184 (1970).
McGee-Russell, S.M.: Histochemical methods for calcium. J. Histochem.Cytochem. 6:22-42
(1958).
548 References
McKenzie, J.M. and Zakarija, M.: LATS in Grave's disease. Recent Progr.Horm.Res. 33:29-57
(1977).
McManus, J.FA.: Histological demonstration ofmucin after periodic acid. Nature 158:202 (1946).
McMillan, P.J.: Differential demonstration of muscle and heart type lactic dehydrogenase of rat
muscle and kidney. J.Histochem.Cytochem. 15:21-31 (1967).
Medawar, P.B.: The rate of penetration of fixatives. J.Roy.Micr. Soc. 61:46-57 (1941).
Meijer, A.E.F.H.: Semipermeable membrane techniques in quantitative enzyme histochemistry,
pp.103-120. In: Trends in enzyme histochemistry and cytochemistry. Ciba Foundation Sym-
posium 73. Excerpta Medica, Amsterdam (1980).
Mendelson, D., Tas, J., and James, J.: Cuprolinic Blue: a specific dye for single-stranded RNA
in the presence of magnesium chloride. ll. Practical applications for light microscopy. Histo-
chem.J. 15:1113-1121 (1983).
Menten, ML., Junge, J., and Green, M.H.: A coupling histochemical azo dye test foralkaline
phosphatase in kidney. J.Biol.Chem. 153:471-477 (1944).
MihaIy, A., KiraIy, E., N6gradi, A., and Bencsik, K.: A modified semipermeable membrane tech-
nique for carbonic anhydrase histochemistry of the nervous system. Hislochemistry 88:485-
487 (1988).
Mizutani, S.; Akiyama, H., Kurauchi, 0., Taira, H., Narita, O. and Tomoda, Y.: In vitro degradation
of angiotensin II (A-II) by human placental subcellular fractions, pregnancy sera and purified
placental aminopeptidases. Acta Endocrinol. 110:135-139 (1985a).
Mizutani, S., Sumi, S., Oka, K., Yamada, R., Kurauchi, 0., Taira, H., Narita, O. and Tomoda,
Y.: In vitro degradation of oxytocin by pregnancy serum, placental subcellular fractions and
purified placental aminopeptidases Exp.Clin.Endocrinol. 86:310-316 (1985b).
Montes, G.S., Krisztän, R.M., Shigihara, K.M., Tokoro, R., Mourao, and Junqueira, L.C.U.: His-
tochemical and morphologica1 characterization of reticular fibers. Histochemistry 65:131-141
(1980).
Morseit, AF.W.: Application and possibilities of cytophotometry for red cell hematology. Progr.
Histochem. Cytochem. 1113:1-42 (1978).
Morstyn, G., Pyke, K., Gardner, J., Ashcroft, R., De Fazio, A. and Bhathal, P.: Immunohisto-
chemical identification of proliferating cells in organ culture using bromodeoxyuridine and a
monoclonal antibody. J.Histochem.Cytochem. 34:697-701 (1986).
Mowry, R.W.: Aldehyde fuchsin staining, direct or after oxidation: Problems and remedies, with
special reference to human pancreatic B cells, pituitaries and elastic fibres. Stain Techno!.
53:141-154 (1978).
Mowry, R.W. and Kent, S.P.: Aldehyde pararosanilin (true aldehyde fuchsin) stains purified
insulin, oxidized or not, following adequate fixation with either formalin or Bouin's fluid.
Stain Techno!. 63:311-323 (1988).
Mullis, K.B.: The unusual origin of the polymerase chain reaction. SciAmer. 26214:36-43 (1990).
Murray, G.I., Burke, M.D., and Ewen, S.W.B.: Dehydrogenase enzyme histochemistry on freeze-
dried or fixed resin-embedded tissue. Histochem.J. 20:491-498 (1988).
Nadji, M. and Morales, A.R.: Immunoperoxidase techniques. A practica1 approach to tumor di-
agnosis. Raven Press, New York (1987).
Nauta, W.J.H. and Gygax, P.A.: Silver impregnation of degenerating axons terminals in the central
nervous system: 1) Technic, 2) Chemical notes. Stain Techno!. 26:5-11 (1951).
Neelsen, F.: Ein casuistischer Beitrag zur Lehre von der Tuberkulose. Centralb!.Med.Wiss. 21:497-
501 (1883).
News and Views. Histochem.J. 17:841-844 (1985).
News and Views. Histochem.J. 18:59-60 (1986).
Newshoime, E.A. and Start, c.: Regulation in metabolism, pp. 1-349. John Wiley & Sons, Chich-
ester (1976).
Newshoime, E.A., Crabtree, B. and Zammit, VA.: Use of enzyme activities as indices of max-
imum rates of fuel utilization. pp.245-258. In: Trends in enzyme histochemistry and cyto-
chemistry. Ciba Foundation Symposium 73. Excerpta Medica, Amsterdam (1980).
Nibbering, P.H., Marijnen, J.G.J., Raap, A.K., Leijh, P.C.J. and van Furth, R.: Quantitative study
of enzyme immunocytochemica1 reactions performed with enzyme conjugates immobilized
on nitrocellulose. Histochemistry 84:538-543 (1986).
References 549
Paschen, W.: Imaging of energy metabolites (ATP, glucose and lactate) in tissue sections: A
bioluminescent technique. Progr. Histochem.Cytochem. 20/4:1-122 (1990).
Patau, K.: Absorption microphotometry of irregular-shaped objects. Chromosome 5:341-362
(1952).
Patzelt, W.J.: Refiexionskontrast, eine neue lichtmikroscopische Technik. Mikrokosmos 3:78-81
(1977).
Pearse, A.D. and Marks, R.: Measurement of section thickness in quantitative microscopy with
special reference to enzyme histochemistry. J.Clin.PathoL 27:615-618 (1974).
Pearse, A.G.E.: Common cytochemical and ultrastructural characteristics of ceUs producing poly-
peptide hormones (the APUD series) and their relevance to thyroid and ultimobranchial C
cells and calcitonin. Proc.Roy.Soc. 170:71-80 (1968).
Pearse, A.G.E.: Histochemistry: Theoretical and applied. 3rd edil Vol. 11. Churchill-Livingstone,
Edinburgh (1972).
Pearse, A.G.E.: Histochemistry: Theoretical and applied. 4th edil Vol. I. Churchill Livingstone,
Edinburgh (1980).
Pearse, A.G.E.: Histochemistry: Theoretical and applied. 4th edit. Vol. 11. Churchill Livingstone,
Edinburgh (1985).
Pedersen, K.J.: Va:vsbiologi. Fra celler til va:v. FADL's Forlag, Copenhagen (1982).
Pellicciari, C. and Fraschini, A.: Methods of denaturation and renaturation of DNA in interphasic
chromatin: cytochemical quantitative analysis by Methyl Green staining. HistochemJ.I0:213-
222 (1978).
Pentz, S. and Schulle, H.: Darstellung des Adhäsionsverhaltens kultivierter Leberzellen an Glas-
oberflächen während der Mitose durch Refiexions-Kontrast-Mikroskopie. Zeiss Inf., Oberko-
chen 25:41-43 (1980).
Perrild, H., Frost, J., and Robertson, W.R.: The cytochemical assay of NAD+ kinase activity in
the follicular cells of guinea-pig thyroid gland. Histochem.J. 16:71-83 (1984).
Perrild, H., Loveridge, N., Reader, S.C.J. and Robertson, W.R. Acute stimulation of thyroidal
NAD+ kinase, NADPH reoxidation, and peroxidase activities by physiological concentrations
of thyroid stimulating hormone acting in vitro: A quantitative cytochemical study. Endocrinol-
ogy 123:2499-2505 (1988).
Perrild, H., Hlllyer, P.E., Loveridge, N., Reader, S.C.J. and Robertson, W.R.: Acute in vitro thy-
rotropin regulation of lysosomal enzyme activity in the thyroid follicular cello MoI.Cell.Endo-
crinol. 65:75-80 (1989).
Peters, A.: Staining of nervous tissue by protein-silver mixtures. Stain Technol. 33:47-53 (1958).
Petersen, O.W., Hlllyer, P.E., and van Deurs, B.: Effect of oxygen on the tetrazolium reaction for
glucose 6-phosphate dehydrogenase in cryosections of human breast carcinoma, fibrocystic
disease and normal breast tissue. Vrrchows Arch. [Cell Pathol.] 50:13-25 (1985).
Petersen, O.W., Hlllyer, P.E., and van Deurs, B.: Frequency and distribution of estrogen receptor-
positive cells in normal, nonlactating human breast tissue. Cancer Res. 47:5748-5751 (1987).
Pette, D. and Brandau, H.: Intracellular localization of glycolytic enzymes in cross-striated muscles
of Locusta migratoria. Biochem.Biophys.Res.Commun. 9:367-370 (1962).
Pette, D. and Wimmer, M.: Microphotometric determination of enzyme activities in cryostat
sections by the gel film technique, pp.121-134. In: Trends in Enzyme Histochemistry and
Cytochemistry, Ciba Foundation Symposium 73. Excerpta Medica, Amsterdam (1980).
PfUller, Y., Franz, H., and Preiss, A.: Sudan black B: Chemical structure and histochemistry of
the blue main components. Histochemistry 54:237-250 (1977).
Pilgrim, C. and Stumpf, W.E.: Applications of autoradiography in neurobiological research.
J.Histochem.Cytochem. 35:917-928 (1987).
PilIer, H.: Microscope photometry. Springer Verlag, Berlin (1977).
Ploem, J.S.: Standards for fiuorescence microscopy, pp.137-153. In: EJ. Holborrow, Ed.: Stan-
dardization in immunofiuorescence. Blackwell Scientific Publications, Oxford (1970).
Ploem, J.S.: Refiection-contrast microscopy as a tool for investigation of the attachment of liv-
ing cells to a glass surface, pp.405-421. In: R. van Furth, Ed.: Mononuclear phagocytes in
immunity, infection and pathology. Blackwell, Oxford (1975).
References 551
Ploem, J.S.: Appropriate technology for the quantitative assessment of the final reaction product
of histochemical techniques, pp.275-303. In: Trends in enzyme histochemistry and cytochem-
istry. Ciba Foundation Symposium 73. Excerpta Medica, Amsterdam (1980).
Ploem, J.S., de Sterke, J.A., Bonnet, J. and Wasmund, H.: A rnicrospectrofluorometer with epi-
illumination operated under computer control. J.Histochem.Cytochem. 22:668-677 (1974).
Poenie, M., Alderton, J., Steinhardt, R., and Tsien, R.: Calcium rises abruptly and briefly through-
out the cell at the onset of anaphase. Science 233:886-889 (1986).
Polak, J.M. and van Noorden, S., Eds.: lmmunocytochemistry. Practical applications in pathology
and biology. John Wright & Sons, Bristol (1983).
Polak, J.M. and van Noorden, S.: An introduction to irnmunocytochemistry: Current techniques
and problems. Oxford University Press, Oxford (1984).
Polak,J.M. and Varndell, I.M., Eds.: lmmunolabelling for electron microscopy. Elsevier, Amster-
dam (1984).
Polasaky, Z., Kiefer, G., and Sandritter, W.: Quantitative histochemische Untersuchungen über
die säurefeste Anfarbung des Spermienkopfes. Histochemie 4:312-321 (1964).
Pollister, A.W. and Ornstein, L.: The photometric chemical analysis of celfS, pp.431-518. In:
R.C. Mellors, Ed.: Analytical cytology, 2nd edit., McGraw-Hill Book Company, New York
(1959).
Poulter, L.W., Campbell, D.A., Munro, C. and Butcher, R.G.: The quantitation of HLA-DR ex-
pression on human cells using irnmunocytochemistry. J.lmmunol.Method. 98:227-234 (1987).
Pratt, W.K.: Digital image processing. Wiley, New York (1978).
Pren~, P.: Metals and phosphate in the chloragosomes of Lumbricus terrestris and their possible
physiological significance. Cell Tiss.Res. 196:123-134 (1979).
Pren~, P.: The effect of histochemical methylation on the phosphate groups of nucleic acids:
Interpretation of the absence of nuclear basophilia Histochem.J. 12:661-668 (1980).
Proctor, G.B. and Horobin, R.W.: The aging of Gomori's aldehyde-fuchsin: the nature of the
chemical changes and the chemical structures of the coloured components. Histochemistry
77:255-267 (1983).
Proctor, G.B. and Horobin, R.W.: Chemical structures and staining mechanisrns of Weigert's
resorcin-fuchsin and related elastic fiber stains. Stain Technol. 63: 101-111 (1988).
Proescher, F. and Arkush, A.S.: Metallic lakes of the oxazines (gallamin blue, gallocyanin and
celestin blue) as nuclear stain substitutes for haematoxylin. Stain Technol. 3:28-38 (1928).
Puchtler, H. and Meloan, S.N.: On the chemistry of formaldehyde fixation and its effects on
irnmunohistochemical reactions. Histochemistry 82:201-204 (1985).
Puchtler, H., Meloan, S.N., and Brewton, B.R.: On the binding of Schiff's reagent in the PAS
reaction: effects of molecular configurations in models. Histochemistry 40:291-299 (1974).
Puchtler, H., Meloan, S.N., and Spencer, M.: Current chemical concepts of acids and bases and
their application to anionic ("acid'') and cationic ("basic'') dyes. Histochemistry 82:301-306
(1985).
Puchtler, H., Meloan, S.N., and Waldrop, F.S.: Application of current chemical concepts to metal-
hematein and -brazilein stains. Histochemistry 85:353-364 (1986).
Puchtler, H., Meloan, S.N., and Waldrop, F.S.: Are picro-dye reactions for collagen quantitative?
Chemical and histochemical considerations. Histochemistry 88:243-256 (1988).
Raap, A.K.: Localization properties of fluorescence cytochernical enzyme procedures. Histochem-
istry 84:317-321 (1986).
Raap, A.K. and van Duijn, P.: Studies on the phenazine methosulphate tetrazolium salt capture
reaction in NAD(p)-dependent dehydrogenase cytochemistry. 11. A novel hypothesis for the
mode of action of PMS and the study of the properties of reduced PMS. Histochem.J. 15:881-
893 (1983).
Raja, S.: Histochemical demonstration of conjugated and unconjugated bilirubin using a modified
diazo-reagent Nature (London). 205:304-305 (1965).
Raju, B., Murphy, E., Levy, L.A., Hall, R.D., and London,RE.: A 1luorescent indicator for
measuring cytosolic free magnesium. AmerJ.Physiol. 256:C540-548 (1989).
Raspail, F.-V.: Nouveau syst~me de chirnie organique. Paris (1833). Cited in: Pearse, p.3 (1980).
Rees, L.H., Holdaway, I.M., Kramer, R., McNeilly, A.S. and Chard, T.: New bioassay for luteiniz-
ing hormone. Nature 244: 232-234 (1973).
552 References
Reid, P.E., Culling, C.FA, and Dunn, WL.: Saponification-induced increase in the periodic
acid-Schiff reaction in the gastrointestinal tract. Mechanism and distribution of the reactive
substance. J.Histochem.Cytochem. 21:473-482 (1973).
Reid, P.E., Volz, D., Cho, K.Y., and Owen, DA.: A new method for the histochemical demon-
stration of O-acyl sugars in human colonic epithelial glycoproteins. Histochem.J. 20:510-518
(1988).
Reid, P.E., Dunn, W.L., Ramey, C.W., Coret, E., Trueman, L., and Clay, M.G.: Histochemi-
cal studies of the mechanism of the periodic acid-phenylhydrazine-Schiff (pAPS) procedure.
Histochem.J. 16:641-649 (1984).
Reid, P.E., Owen, D.A., Fletcher. K., Rowan, R.E., Reimer, C.L., Rouse, G.J., and Park, C.M.:
The histochemical specificity of High Iron Diamine-Alcian Blue. Histochem.J. 21:501-504
(1989).
Rietveld, W.J.: The reliability of color perception in man, pp.73-S1. In: M.E. Boon and L.P.
Kok, Eds.: Standardization and quantitation of diagnostic staining in cytology. Coulomb Press
Leyden, Leiden (1986).
Rink, TJ., Tsien, R.Y., and Pozzan, T.: Cytoplasmic pH and free Mg2+ iii lymphocytes. J.Cell
Biol. 95:189-196 (1982).
Rio-Hortega, P. deI: Noticia de un neuvo y fäcH m~todo para la coloraci6n de la neuroglia y deI
tejido conjuntivo. Trab.Lab. Invest.Biol. 15:367-378 (1917).
Robertson, W.R.: A quantitative cytochemical method for the demonstration of ß S,3ß-hydroxy-
steroid dehydrogenase activity in unfixed tissue sections of rat ovary. Histochemistry 59:271-
285 (1979).
Robertson, W.R.: A quantitative study of N-acetyl-ß-glucosaminidase activity in unfixed tissue
sections of the guinea-pig thyroid gland. Histochem.J. 12:87-96 (1980).
Robertson, W.R., Earnshaw, RJ. and Larnbert, A.: The determination by microdensitometry of
the initial maximum velocity rate of rat ovarian 3ß-hydroxysteroid dehydrogenase activity.
Histochemistry 74:435-442 (1982a).
Robertson, W.R., Frost, J., HI/lyer, P .E., and Weinkove, C.: 20a-Hydroxysteroid dehydrogenase ac-
tivity in the rat corpus luteum; a quantitative cytochemical study. J.Steroid Biochem. 17:237-
243 (l982b).
Roelfzema, W.H., Tohyama, C., Nishimura, Nishimura, N. and Morselt, A,F.W.: Quantitative
immunohistochemistry of metallothionein in rat placenta. Histochemistry 90:365-369 (1989).
Rogers, A.W.: Techniques of autoradiography. 3rd edit. Elsevier and North-Holland, Amsterdam
(1979).
Roozemond, R.C.: The effect of calcium chloride and formaldehyde on the release and compo-
sition of phospholipids from cryostat sections of rat hypothalamus. J.Histochem.Cytochem.
17:273-279 (1969a).
Roozemond, R.C.: The effect of fixation with formaldehyde and glutaraldehyde on the composi-
tion of phospholipids extractable from rat hypothalamus. J.Histochem.Cytochem. 17:482-486
(1969b).
Ruhnke, M. and Gossrau, R.: Kinetic end-point microdensitometry (section biochemistry) of 'Y-
glutamyl transpeptidase and dipeptidyl peptidase IV in the mature mouse decidua and visceral
yolk sac. Acta histochem. 87:23-31 (1989a).
Ruhnke, M. and Gossrau, R.: Reaction rate measurements of proteases and glycosidases with
chromogenic methods. Histochem.J. 21:535-544 (1989b).
Ruijgrok, J.M., Boon, M.E. and de Wijn, J.R.: The effect of heating by microwave irradiation and
by conventional heating on the aldehyde concentration in aqueous glutaraldehyde solutions.
Histochem.J. 22:389-393 (1990).
Rüter, A., Wittekind, D., Gunzer, U., Aus, H.M. and Harms, H.: Comparative colorimetry in
cytophotometric measurements of azure B-eosin Y -stained and Giemsa-stained blood cell
smears. Anal.Quant.Cytol. 4:128-139 (1982).
Rys, P. and Zollinger, H.: The theory of coloration of textiles. The Dyers Company Publications
Trust (1975).
Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B. and
Erlich, HA: Primer-directed enzymatic amplification of DNA with a thermostable DNA
polymerase. Science 239:487-491 (1988).
References 553
Salthouse, T.N.: Luxol fast blue ARN: A new solvent azo dye with improved staining qualities
for myelin and phospholipids. Stain Technol. 37:313-316 (1962).
Sandritter, W., Kiefer, G., and Rick, W.: Über Stöchiometrie von Gallocyanin-chromalaun mit
Desoxyribonukleinsäure. Histochemie 3: 314-340 (1963).
Sannes, P.L.: The histochemical and cytochemical localization of proteases. Progr.Histochem.
Cytochem. 18/1:1-48 (1988).
Sarkar, G. and Sommer, S.S.: Shedding light on PCR contamination. Nature 343:27 (1990).
Schauenstein, E., Desoye, G. and Nöhammer, G.:Quantitative Proteinbestimmung in Einzelzellen
mit Amidschwarz. Histochemistry 68:75-90 (1980).
Schenk, E.: Note from the Biological Stain Commission. A newly certified dye-Alcian Blue 8GX.
Stain Technol. 56:129-131 (1981).
Scherini, E., Mares, V., Bemocchi, G. and Bami, S.: lntranuclear differences in the response of
Purkinje cell DNA of the rat cerebellwn to bleomycin. A microphotometric and autoradio-
graphic study. Histochemistry 89:227-230 (1988).
Schieck, R., Tauben, G. and Krug, H.: Dry mass, DNA and non-histone protein determinations
in lung cancer cells. Histochem.J. 19:504-508 (1987).
Schmidt, H.: Phenol oxidase (E.C. 1.14.18.1) a marker enzyme for defence cells. Progr.Histochem.
Cytochem. 17/3:1-194 (1988).
Schmorl, G.: Die Pathologisch-Histologische Untersuchungsmethoden. 15th ediL Vogel, Leipzig
(1928).
Schofield, K.P., Stone, P.C.W., Beddall, A.C., and Stuan, J.: Quantitative cytochemistry of the
toxic granulation blood neutrophil. Brit.J.Haematol. 53:15-22 (1983).
Schuben, M. and Hamerman, B.: Metachromasia; chemical theory and histochemical use. J.Histo-
chem.Cytochem. 4:159-189 (1956).
Schulte,E., Wittekind, D., and Kretschmer, V.: Victoria Bille B-a nuclear stain for cytology. A
cytophotometric study. Histochemistry 88:427-433 (1988).
Scopsi, L. and Larsson, L.-I.: lncreased sensitivity in peroxidase immunocytochemistry. Histo-
chemistry 84:221-230 (1986a).
Scopsi, L. and Larsson, LA.: Bodian's silver impregnation of endocrine cells. A tentative expla-
nation to the staining mechanism. Histochemistry 86:59-62 (1986b).
Scott, J.E.: On the mechanism of the Methyl Green-Pyronin stain for nucleic acids. Histochemie
9:30-47 (1967).
Scott, J.E.: Histochemistry and Alcian Blue. I. Metachromasia of Alcian Blue, Astrablau and
other cationic phthalocyanin dyes. Histochemie 21:227-285 (1970).
Scott, J.E.: The histochemistry of Alcian Blue. Note on the presence and removal of boric acid
as the major diluent in Alcian blue 8GX. Histochemie 29:129-133 (1972a).
Scott, J.E.: Amplification and staining by Alcian blue and similar ingrain dyes. J.Histochem.Cyto-
chem. 20:750-752 (1972b).
Scott, J.E.: Histochemistry of Alcian Blue. 1lI. The molecular biological basis of staining by
Alcian Blue 8GX and analogous phthalocyanins. Histochemie 32:191-212 (1972c).
Scott, J.E.: Alcian dyes: I.C.I. cease manufacture and release details of composition. Histochemie
37:379-380 (1973a).
Scott, J.E.: The second BDH-Iecture. Trans.Biochem.Soc. 1:787-806 (1973b).
Scott, J.E.: Collagen-proteoglycan interactions. Localization of proteoglycans in tendon by elec-
tron microscopy. Biochem.J. 187:887-891 (1980).
Scott, J.E. and Dorling, J.: Differential staining of acid solutions. Histochemie 5:221-233 (1965).
Scott, J.E. and Mowry, R.W.: Alcian Blue-a consumer's guide. J. Histochem.Cytochem. 18:842
(1970).
Seidler, E.: New nitro-monotetrazoliwn salts and their use in histochemistry. HistochemJ. 12:619-
630 (1980).
Seifen, G., ed.: Morphological tumor markers. General aspects and diagnostic relevance. Springer-
Verlag, New York (1987).
Seligman, A.M., Karnovsky, MJ., Wasserkrug, H.1. and HaRker, J.S.: Non-droplet ultrastruc-
tural demonstration of cytochrome oxidase activity with a polymerising osmiophilic reagent,
diaminobenzidine (DAB). J.Cell Biol. 38:1-14 (1968).
554 Referenees
Shea, J.R., Jr.: A method for in situ cytaphotometric estimation of absolute amount of ribonucleic
acid using azure B. J.Histochem.CytOChem. 18:143-152 (1970).
Shennawy, I.E., Gee, D.J., Horobin, R.W., and Aparicio, S.R.: Novel application of aniline blue.
Fluorescent staining of glycogen and some protein structures. Histochemistry 81:93-98 (1984).
Shibata, D.K., Arnheim, N. and Martin, WJ.: Detection of human papilloma virus in paraffin-
embedded tissue using the polymerase chain reaction. J.Exp.Med. 167:225-230 (1988).
Shires, T.K.: Tests for nucleic acid selectivity of Einarson's gallocyanin-chrome alum stain on
monolayers of strain L-929 mouse fibroblasts. Stain Technol. 42:67-74 (1967).
Sinapius, D.: Über Chromierungsmethoden zur Lipoiddarstellung. Histochemie 2:217-233 (1961).
Singer-Sam, J., Tanguay, R.L. and Riggs, A.D.: Use of chelex to improve the PCR signal from
a small number of cells. Amplifications-A forum for PCR users, Perkin EImer Cetus, 3:11
(1989).
Sippel, T.O.: The histochemistry of thiols and disulphides. I. The use of N-(4-aminophenyl)malei-
mide for demonstrating thiol groups. Histochem.J. 5:413-423 (1973).
Sippel, T.O.: New fiuorochromes for thiols: maleimide and iodoacetamide derivatives of a 3-
phenylcoumarin fiuorophore. J.Histochem.CytOChem. 29:314-316 (1981).
Sisken, J.E.: Fluorescent standards, pp.1l3-126. In: D.Lansing Taylor and Yu-Li Wang, Eds.:
Methods in cell biology, Vol. 30: Fluorescence microscopy of living cells in culture, Part B.
Quantitative fiuorescence microscopy-lmaging and spectroscopy. Academic Press, Inc., San
Diego (1989).
Sjögren, S.: Lactate dehydrogenase isozymes in developing rat oral mucosa. A comparative study
of LDH biochemistry and histochemistry. J.Histochem.Cytochem. 32:958-964 (1984).
Smith, H.: The leukaemias, pp.206-214. In: M.I. Filipe and B.D. Lake, Eds.: Histochemistry in
pathology. Churchill Livingstone, Edinburgh (1983).
Smith, M.T., Redick, I.A. and Baron, J.: Quantitative immunohistochemistry: A eomparison of
mierodensitometric analysis of unlabeled antibody peroxidase-antiperoxidase staining and of
mierofiuorometric analysis of indirect fiuorescent antibody staining for nicotinamide adeno-
sine dinucleotide phosphate (NADPH)-cytOChrome-c (p-450) reductase in rat liver. I.Histo-
chem.Cytochem. 31:1183-1189 (1983).
Smith, M.T., Loveridge, N., WiIls, E.D. and Chayen, I.: The distribution of glutathione in the rat
liver lobule. Biochem.I. 182:103-108 (1979).
Society of Dyers and Colourists: The colour index. 3rd edit Bradford (England) (1971).
Solcia, E., Vassallo, G., and Capella, C.: Selective staining of endocrine cells by basic dyes after
aeid hydrolysis. Stain Technol. 43:257-263 )1968).
Spieer, S.S.: The use of various cationic reagents in histochemical differentiation of mucopolysac-
charides. Amer.I. Clin.Pathol. 36:393-409 (1961).
Spicer, S.S.: Diamine methods for differentiating mucosubstances histochemically. I.Histochem.
Cytochem. 13:211-234 (1965).
Stellmach, I. and Severin, E.: A fiuorescent redox dye. Infiuence of several substrates and electron
carriers on the tetrazolium salt-formazan reaction of Ehrlich ascites tumour eells. HistochemJ.
19:21-26 (1987).
Stemberger, L.A. and Sternberger, N.H.: The unlabeled antibody method: Comparison of peroxi-
dase-antiperoxidase with avidin-biotin complex by a new method of quantification. I.Histo-
ehem.Cytochem. 34:599-605 (1986).
Stemberger, N.H., Itoyama, Y., Kies, M.W. and Webster, H. de F.: Myelin basic protein demon-
strated immunocytochemically in oligodendroglia prior to myelin sheath formation. Proc.
NatlAcad.Sci.USA 75:2521-2524 (1978).
Stoward, P.I.: Criteria for the validation of quantitative histochemical enzyme techniques, pp.11-
31. In: Trends in enzyme histochemistry and cytochemistry. Ciba Foundation Symposium 73.
Excerpta Mediea, Amsterdam (1980).
Stoward, P.I. and AI-Sarraj, B.: Quantitative histochemical investigations of semipermeable mem-
brane techniques for the assay of acid phosphatase in skeletal muscle. I. Meijer's method.
Histochemistry 71:405-431 (1981a).
Stoward, P.I. and AI-Sarraj, B.: Quantitative histochemical investigations of semipermeable mem-
brane techniques for the assay of acid phosphatase in skeletal muscle. m. A modified simul-
taneous coupling technique. Histochemistry 71:599-607 (1981b).
References 555
Stoward, P.J. and Nakae, Y.: Simultaneous histochemical assay of two dehydrogenases in the
same cello Histochem.J. 20:610-616 (1988).
Stoward, P.J., AI-Sarraj, B., and Ireland, H.: Quantitative histochemical investigations of semiper-
meable membrane techniques for the assay of acid phosphatase in skeletal musele. 11. Non-
specific reactions in control sections. Histochemistry 71:585-598 (1981).
Stoward, P.J., Campbell, J.B., and Al-Sarraj, B.: Quantitative histochemical investigations of
semipermeable membrane techniques for the assay of acid phosphatase in skeletal musele.
IV. A post-coupling technique. Histochemistry 74:367-377 (1982).
Stoward, P.J., Meijer, A.E.F.H., Seidler, E. and Wohlrab, F.: Dehydrogenases. Chapter 22. In:
P J. Stoward, Ed.: Pearse, A.G.E.: Histochemistry: Theoretical and applied. 4th edit. Vol. m.
Churchill Livingstone, Edinburgh In press (1991).
Sumi, Y.: Histochemical staining of cadmium with thiazolylazophenol derivatives. Histochemistry
67:1-6 (1980).
Sumi, Y., Muraki, T., and Suzuki, T.: Histochemical staining of cadmium with benzothiazolyla-
zonaphthol derivatives. Histochemistry 73:481-486 (1982).
Suurmeijer, AJ.H., Boon, M.E. and Kok, L.P.: Notes on the application of microwaves in
histopathology. HistochemJ. 22:341-346 (1990).
Swift, H.H. and Rasch, E.: Microphotometry with visible light. In: G. Oster and AW. Pollister,
Eds.: Physical techniques in biological research, Vol 3. Academic Press, New York (1956).
Takamatsu, H.: Histologische und biochemische Studien über die Phosphatase. I. Mitteilung.
Histochemische Untersuchungsmethodik Phosphatase und deren Verteilung in verschiedenen
Organen und Geweben. Trans.Soc.Path.Jap. 29:492-498 (1939).
Tanasugarn, L., McNeil, P., Reynolds, G.T., and Lansing Taylor, D.: Microspectroftuorometry by
digital image processing: Measurement of cytoplasmic pH. J.CelI Biol. 98:717-724 (1984).
Tas, J.: The Alcian Blue and combined Alcian Blue-Safranin 0 staining of glycosaminoglycans
studied in a model system and in mast cells. Histochem.J. 9:205-230 (1977).
Tas, J. Mendelson, D., and Noorden, C.J.F.: Cuprolinic Blue: a specific dye for single-stranded
RNA in the presence of magnesium chloride. I. Fundamental aspects. HistochemJ. 15:801-
814 (1983).
Taylor, C.R.: Immunomicroscopy. W.B. Saunders Company, Philadelphia (1986).
Taylor, D.L. and Salmon, E.D.: Basic ftuorescence microscopy, pp.207-237. In: Y.-L. Wang and
Taylor, D.L., Eds.: Methods in cell biology, Vol. 29. Fluorescence microscopy of living cells
in culture. Part A Fluorescent analogs, labeling cells, and basic microscopy. Academic Press,
Inc., Boston (1989).
Thomas, RC.: Bicarbonate and pHi response. Nature 337:601 (1989).
Troyer, H., Fisher, O.T. and Rosenquist, T.H.: Bone alkaline phosphatase kinetics studied by a
new method. Histochemistry 50: 251-259 (1977).
Tsien, R.Y.: New calcium indicators and buffers with high selectivity against magnesium and
protons: Design, synthesis, and properties ofprototype structures. Biochemistry 19:2396-2404
(1980).
Tsien, RY.: A non-disruptive technique for loading calcium buffers and indicators into cells.
Nature 290:527-528 (1981).
Tsien, RY.: Fluorescence measurement and photochemical manipulation of cytosolic free calcium.
TINS 11:419-424 (1988).
Tsien, R Y.: Fluorescent indicators of ion concentrations. In: D Lansing Taylor and Y.-L. Wang,
Eds: Methods in cell biology, vol 30 part B. Fluorescence microscopy of living cells in culture.
pp.127-156. Academic Press, Inc. San Diego, Califomia (1989).
Tsien, Y., Pozzan, T., and Rink, T.J.: Calcium homeostasis in intact lymphocytes: Cytoplasmic
free calcium monitored with a new, intracellularly trapped ftuorescent indicator. J.Cell Biol.
94:325-334 (1982).
Tsien, RY., Rink, T.J., and Poenie, M.: Measurement of cytosolic free Ca2+ in individual small
cells using ftuorescence microscopy with dual excitation wavelengths. Cell Calcium 6:145-
157 (1985).
Tuckett, F. and Morriss-Kay, G.: Alcian Blue staining of glycosaminoglycans in embryonic ma-
terial: effect of different fixatives. HistochemJ. 20:174-182 (1988).
556 References
Tuqon, N.A. and Adams, C.W.M.: Histochemistry of myelin. I. Proteins and lipid-protein com-
plexes in the normal sheath. J. Neurochem. 6:327-333 (1961).
Umezawa, H., Aoyagi, T., Ogawa, K., Naganawa, H., Harnada, M. and Takeuchi, T.: Diprotins
A and B, inhibitors of dipeptidyl arninopeptidase IV, produced by bacteria. J.Antibiotics
37:422-425 (1984).
Uzman, LL.: Histochemical localization of copper with rubeanic acid. Lab.1nvest 5:299-305
(1956).
Van Bogaert, L.J., Quinones, J.A., and van Craynert, M.P.: Difficulties involved in diaminoben-
zidine histochemistry of endogenous peroxidase. Acta Histochem. 67:180-194 (1980).
Van de Kant, H.J.G., van Pell, A.M.M., Vergouwen, R.P.F.A. and de Rooij, D.G.: A rapid
immunogold-silver staining procedure for detection of bromodeoxyuridine in large numbers
of plastic sections, using microwave irradiation. Histochem.J. 22:321-326 (1990).
Van den Brink, W.J., Zijlmans, HJ.M.A.A., Kok, L.P., Bolhuis, P., Volkers, H.H., Boon, M.E.
and Houthoff, HJ.: Microwave irradiation in label-detection for diagnostic DNA-in situ hy-
bridization. Histochem.J. 22:327-334 (1990).
Van der Ploeg, M. and van Duijn, P.: 5,6-dihydroxy indole as a substrate in a histochemical
peroxidase reaction. J.R.microsc.Soc. 83:415-423 (1964).
Van der Ploeg, M. and van Duijn, P.: Cytophotomettic determination of alkaline phosphatase
activity of individual neutrophilic leukocytes with a biochemical calibrated model system.
J.Histochem.Cytochem. 16:693-706 (1968).
Van der Ploeg, M. and van Duijn, P.: Reflection versus fluorescence. A note on the physical
backgrounds of two types of light microscopy. Histochem. 62:227-232 (1979).
Van der Ploeg, M., van den Broek, K. and MilChell, J.P.: Dual wavelength scanning cytophotom-
etry (Bicoscan). Histochemistry 62:29-43 (1979).
Van Duijn, P.: Fundamental aspects of enzyme cytochemistry, pp.3-23. In: E. Wisse, W.Th.
Daems, I. Molenaar and P. van Duijn, Eds.: Electron microscopy and cytochemistry. North-
Holland, Amsterdam (1974).
Van Duijn, P.: Prospects for microscopical cytochemistry. Histochem.J. 8:653-676 (1976).
Van Duijn, P. and van der Ploeg, M.: Microscopic cytochemistry as Matrix chemistry, pp.209-229.
1n: Ciba Foundation symposium 73. Excerpta Medica, Amsterdam (1980).
Van Duijn, P. and van Noorden, CJ.F.: Plateau absorbance measurements: an alternative ap-
proach to enzyme activity determination illustrated by the example of alkaline phosphatase.
Histochem.J. 21:619-624 (1989).
Van Holde, K.E.: Physical biochemistry. Prentice-Hall, Englewood Cliffs, New Jersey (1971).
Van Kuppevelt, T.H.M.S.M., Cremers, F.P.M., Domen, J.G.W., and Kuyper, C.M.A.: Staining
of proteoglycans in mouse lung alveoli. H. Characterization of the Cuprolinic Blue-positive
anionic sites. Histochem.J. 16:671-686 (1984a).
Van Kuppevelt, T.H.M.S.M., Domen, J.G.W., Cremers, F.P.M., and Kuyper, C.M.A.: Staining
of proteoglycans in mouse lung alveoli. I. UltrastruclUral localization of anionic sites. His-
tochem.J. 16:657-669 (1984b).
Van Noorden, C.J.F.: Histochemistry and cytochemistry of glucose-6-phosphate dehydrogenase.
Progr.Histochem.Cytochem. 15/4:1-85 (1984).
Van Noorden, C.J.F.: Enzyme reaction rate studies in tissue sections. Proceedings Royal Micro-
scopical Society 23:93-97 (1988a).
Van Noorden, CJ.F.: On the role of oxygen in dehydrogenase reactions using tetrazolium salts.
HistochemJ. 20:587-593 (1988b).
Van Noorden, C.J.F.: Principles of cytophotometry in enzyme histochemistry and validity of the
reactions. Acta Histochem. 37 Suppl.:21-35 (1989).
Van Noorden, C.J.F. and Butcher, R.G.: Histochemicallocalization of NAD(p)-dependent dehy-
drogenase activity with four different tetrazolium salts. J.Histochem.Cytochem. 32: 998-1004
(1984).
Van Noorden, C.J.F. and BUICher, R.G.: A quantitative histochemical study of NADPH-ferrihemo-
protein reductase activity. Histochem.J. 18:364-370 (1986a).
Van Noorden, C.J.F. and Butcher, R.G.: The out-of-range error in microdensitometry. Histochem.J.
18:397-398 (1986b).
References 557
Van Noorden, C,J.F. and Butcher, R.G.: Quantitative enzyme histochemistty. Chapter 33. In:
P,J. Stoward, Ed.: Pearse, A.G.E.: Histochernistty: Theoretical and applied. 4th edit. Vol. IlI.
Churchill Livingstone, Edinburgh In press (1991).
Van Noorden, C.J.F. and Jonges, G.N.: Quantification of the histochernical reaction for alkaline
phosphatase activity using the indoxyl-tetranitro BT method. Histochern.J. 19:94-102 (1987).
Van Noorden, C.F.J. and Tas. J.: Advantages of I-methoxyPMS as electron carrier in dehydroge-
nase cytochernistty. Proc.R.microsc.Soc. 16:248 (1981a).
Van Noorden, C.FJ. and Tas, J.: Model film studies in enzyme histochernistty with special refer-
ence to glucose-6-phosphate dehydrogenase. Histochem.J. 13:187-206 (1981b).
Van Noorden, C,J.F. and Tas, J.: Model film studies in enzyme histochernistty with special refer-
ence to glucose-6-phosphate dehydrogenase. Histochem.J. 13:187-206 (1981c).
Van Noorden, C.F,J. and Tas, J.: Advantages of I-methoxyPMS as electron carrier in dehydroge-
nase cytochemistty. Histochern.J. 14:837-845 (1982a).
Van Noorden, C.F.J. and Tas, J.: The role of exogenous electron carriers in NAD(p)-dependent
dehydrogenase cytochernistty studied in vitro and with a model system of polyacrylarnide
films. J.Histochem.Cytochem. 30:12-20 (1982b).
Van Noorden, C.J.F. and Vogels, I.M.C.: Enzyme histochernical reactions in unfixed and unde-
calcified cryostat sections of mouse knee joints with special reference to arthritic lesions.
Histochernistty 86:127-133 (1986).
Van Noorden, C.J.F. and Vogels, I.M.C.: Polyvinyl alcohol and other tissue protectants in enzyme
histochemistty: a consumer's guide. Histochem.J. 21:373-379 (1989a).
Van Noorden, C,J.F. and Vogels, I.M.C.: Cytophotometric analysis of succinate and lactate dehy-
drogenase activity in rat liver, heart muscle and tracheal epithelium. Histochern.J. 21:575-583
(1989b).
Van Noorden, C.J.F., Vogels, I.M.C. and van Wering, E.R.: Enzyme cytochernistry of unfixed
leukocytes and bone marrow ceIls using polyvinyl alcohol for the diagnosis of leukemia.
Histochernistry 92:313-318 (1989).
Van Noorden, C.J.F., Kooij, A., Vogels, I.M.C., and Frederiks, W.M.: On the nature of the 'nothing
dehydrogenase' reaction. Histochem.J.17:1111-1118 (1985).
Van Noorden, C,J.F., Vogels, I.M.C., Everts, V. and Beertsen, W.: Localization of cathepsin B
activity in fibroblasts and chondrocytes by continuous monitoring of the formation of a final
fiuorescent reaction product using 5-nitrosalicylaldehyde. Histochem.J. 19:483-487 (1987).
Van Oostveldt, P. and Bauwens, S.: Quantitative fiuorescence in confocal microscopy. The effect
of the detection pinhole aperture on the re-absorption and inner filter phenomena. J.Microsc.
158:121-132 (1990).
Van Prooijen-Knegt, A.C., Raap, A.K., van der Burg, M.J.M., Vrolijk, J. and van der Ploeg, M.:
Spreading and staining of human metaphase chromosomes on arninoalkylsilane-treated glass
slides. Histochem.J. 14:333-344 (1982).
Vejlsted, H.: The infiuence of light on gallocyanin-chrome alum stained sections. J.Histochem.
Cytochem. 17:545-546 (1972).
Volz, D., Reid, P .E., Park, C.M., Owen, D.A., Dunn, W.L., and Rarney, C.W.: Can "mild"
periodate oxidation be used for the specific histochemical identification of sialic acid residues?
Histochem.J. 18:579-582 (1986).
Volz, D., Reid, P.E., Park, C.M., Owen, D.A., and Dunn, WL.: Histochernical procedures for the
simultaneous visualization of neutral sugars and either sialic acid and its side chain O-acyl
variants or O-sulphate ester. I. methods based upon the selective periodate oxidation of sialic
acids. Histochem.J. 19:249-256 (1987a).
Volz, D., Reid, P .E., Park, C.M., Owen, D.A., and Dunn, W L.: A new histochemical method for
the selective periodate oxidation of total tissue sialic acids. Histochem,J. 19:311-318 (1987b).
Von Bülow, F.A. and Hjilyer, P .E.:Use of epoxy resin slides for handling unfixed cryostat sections
intended for histochemistty at the ultrastructurallevel. Histochem,J. 15:825-827 (1983).
Wachsrnuth, E.D.: Assessrnent of immunocytochemical techniques with particular reference to
the mixed-aggregation irnmuno-cytochernical technique, pp.135-159. Trends in enzyme histo-
chemistty and cytochemistty. Ciba Foundation Symposium 73, Excerpta Medica, Arnsterdarn
(1980).
558 References
Wachsmuth, E.D.: A comparison of the highly selective fluorescence staining of fungi in tissue
sections with Uvitex 2B and Calcofluor White M2R. HistochemJ. 20:215-221 (1988).
Wachsmuth, E.D., Thöner, M., and Pfleiderer, G.: The cellular distribution of aldolase isozymes
in rat kidney and brain determined in tissue sections by the irnmuno-histochemical method.
Histochemistry 45:143-161 (1975).
Wachstein, M. and Meisel, E.: Histochemistry of substrate specific phosphatases at a physiological
pH. J.Histochem.Cytochem. 4:424-425 (1956).
Warford, A.: In situ hybridisation: a new tool in pathology. Med.Lab.Sci. 45:381-394 (1988).
Waterhouse, D.F.: Histochemical detection of barium and strontium. Nature, London 167:358
(1951).
Waters,S.E. and Butcher, R.G.: Studies on the Gomori acid phosphatase reaction: the preparation
of the incubation medium. HistochemJ. 12:191-200 (1980).
Weakley, B.S.: A beginner's handbook in biological transmission electron mieroscopy. 2nd edit.
Churchill-Livingstone, New York (1981).
Weber, P., Harrison, F.W., and Hof, L.: Histochemical application of dansylhydrazine as a fluo-
rescent labelling reagent for sialic acid in cellular glycoconjugates. Histochemie 45:271-277
(1975).
Weibel, E.R.: Stereologieal Methods: Practical methods for biologieal morphometry. Vol.l. Aca-
demic Press, London, (1979).
Weibel, E.R.: Stereologieal Methods: Theoretical foundations. Vol.2. Academic Press, London,
(1980).
Weier, H.-U.G., Segraves, R., Pinkel, D. and Gray, J.W.: Synthesis of Y chromosome-specific la-
beled DNA probes by in vitro DNA amplification. J.Histochem.Cytochem. 38:421-426 (1990).
WeIch, J.P., MitchelI, R.B. and Davis, J.C.: Cytophotometric measurements of DNA, RNA and
basic protein in FSH treated rat Sertoli cells in culture. Biol.Reprod. 21:69-74 (1979).
White, T.J., Arnheim, N. and Erlich, H.A.: The polymerase chain reaction. Trends Genetics
5:185-188 (1989).
Whittaker, J.R.: Quantitative measurement by microdensitometry of tyrosinase (dopa oxidase)
development in whole small ascidian embryos. Histochemistry 71:349-359 (1981).
Wieland, H. and Scheuing, G.: Die Fuchsin-Schweflige Säure und ihre Farbreaktion mit Aldehy-
den. Ber.Dtsch.Chem.Ges. 54:2527-2555 (1921).
Williams, G. and Jackson, D.S.: Two organic fixatives for acid mucopolysaccharides. Stain Tech-
nol. 31:189-191 (1956).
Williams, M.A.: Autoradiography and irnmunocytochemistry (part 1). Quantitative methods in
biology (part 2). In: Glauert, A.M., Ed.: Practical methods in electron microscopy. Vol. 6.
North-Holland Publishing Company, Amsterdam (1977).
Winzek, C., Plieninger, P. and Baumgärtel, H.: An improved method to investigate staining
kinetics in single cells. Histochemistry 86:421-426 (1987).
Wittekind, D.: Standardization of dyes and stains for automated cell pattern recognition. Ana-
lyt.Quant.Cytol.Histol. 7:6-30 (1985).
Wittekind, D. and Schulte. E.: Die Bedeutung der Standardisierung der Zell- und Gewebspräpara-
tion für bildanalytische Operationen. In: Eins, S., Ed.: Quantitative und strukturelle Bildanalyse
in der Medizin. GIT-Verlag, Darmstadt. (1987).
Wittekind, D., Hilgarth, M., and Kretschmer, v.: Die einfache und reproduzierbare Papanicolaou-
Färbung. Geburtshilfe Frauenheilkd. 39:969-972 (1979).
Wittekind, D., Hilgarth, M., and Kretschmer, v.: The new and reproducible Papanicolaou stain.
Morphologie and spectrophotometric observations on the influence and stain composition on
staining results. Anal.Quant.Cytol. 4:286-294 (1982).
Wittekind, D.H.: On the nature of Romanowsky-Giemsa staining and its significance for cyto-
chemistry and histochemistry: an overall view. HistochemJ. 15:1029-1047 (1983).
Wittekind, D.H., Reinhardt, E.R., Kretschmer, V., and Zipfel, E.: Influence of staining on fast
automated cell segmentation, feature extraction and cell image analysis. Anal.Quant.Cytol.
5:55-60 (1983).
Wohlrab, F., Seidler, E. and Kunze, K.D.: Histo- und Zytochemie dehydrierender Enzyme. Grund-
lagen und Problematik. J.A. Barth, Leipzig (1979).
References 559
Wolman, M. and Kasten, F.H.: Polarized Hght microscopy in the study of the molecular structure
of collagen and reticulin. Histochemistry 85:41-49 (1986).
Wolters, G.H.J., Pasma, A., Wiegman, J.B. and Konijnendijk, W.: Changes in histochemically
detectable calcium and zinc during tolbutamide-induced degranulation and subsequent regran-
ulation of rat pancreatic islets. Histochemistry 78:325-338 (1983).
World Health Organization, Expert Committee Biological Standardization, 26th Report. WHO
Techn.Rep.Ser. 565:5-72 (1975).
Wright, J.R., Calkins, E., and Humphrey, R.L.: Potassium permanganate reaction in amyloidosis.
A histologic method 10 assist in differentiating forms of this disease. Lab.Invest. 36:274-281
(1977).
Wolman, M. and Kasten, F.H.: Polarized light microscopy in the study of the molecular structure
of collagen and reticulin. Histochemistry 85:41-49 (1986).
Yamabayashi, S.: Periodic acid-Schiff-Alcian Blue: a method for the differential staining of gly-
coproteins. Histochem.J. 19:565-571 (1987).
Yasuma, A. and Itchikawa, T.:Ninhydrin-Schiff and alloxan-Schiff staining. A 1!Cw histochemical
staining method for protein. J.Lab. Clin.Med. 41:296-299 (1953).
Yoffe, J.R.: A comparison of two cytochemical methods used to determine lysosomal function in
cultured endothelial cells. HistochemJ. 12:537-542 (1980).
Ziehl, F.: Zur Färbung des Tuberkelbacillus. Dtsch.Med. Wochenschr. 8:451 (1882).
Ziehl, F.: Ueber die Färbung des Tuberkelbacillus. Dtsch.Med. Wschr. 9:247-249 (1883).
Zimmennann, H.: Physikalisch-chemische Grundlagen der Färbung für manuelle und apparative
Zytodiagnostik. In: Wittekind, D., Ed.: Manual and automated diagnosis in cytology. Mi-
crosc.Acta [Suppl]6:45-58 (1983).
Zipfel, E., Grezes, J.-R., Naujok, A., Seifert, W., Wittekind, D.H. and Zimmermann, H.W.: Über
Romanowsky-Farbstoffe und den Romanowsky-Giemsa-Effekt. 3. Mitteilung: Mikrospektral-
photometrische Untersuchung der Romanowsky-Giemsa-Färbung. Spektroskopischer Nach-
weis eines DNA-Azur B-Eosin Y-Komplexes, der den Romanowsky-Giemsa-Effekt verur-
sacht. Histochemistry 81:337-351 (1984).
Zoller, L.C., Mulhern, P. and Hulik, D.: Quantitative and qualitative cytochemical analysis of
changes in polysaccharides-proteins in rat ovulable and atretic ovarian follicles during the
estrous cycle. Histochemistry 73:339-351 (1981).
Subject Index
Please note: the numbers in the indices refer to section numbers in the text
of endocrine glands, 32.5.2 Van der Waal bonds, see Van der Waal forces
nervous system, 32.5.6 forces, 2.4.4, 4.5.4, 6.1.1, 21.5
ovary, 32.5.3 Vasculitis, 32.4.1
pancreas, 32.5.7 VCD, see Vinylcyclohexene dioxide
prostate gland, 32.5.7 Vertebrate, 17.7.1, 17.7.3
skin, 32.5.7 Vestopal W, see Resin, Epoxy
testis, 32.5.3 Vibratome, 10,27.1.4
Surface coelom epithelium, 32.5.3 Video camera, 28.2.3, 28.8.10, App. 8.6.1
Soft tissue, 31.7.2,31.10.6 Vimentin, 2.2.9
Undifferentiated, 32.5.1 Vinylcyclohexene dioxide, 14.4.2
Yolk sac, 32.5.3 Volume, 28.2
Turanose, 24.6.2 Volutin granule, 17.7.12
Tween, 24.6.1
Tyrosinase, 18.1.3
Tyrosine, 18.1.3 Warts, 31.3.1
Washing, 16.2
Water, 2.1.1,3.2,3.2.1,4:3.1, 14.4.1
Ubiquinone-10, 25.4.1 Demineralized, 16.2
Ulcerative colitis, see Colitis, Chronic ulcerative Distilled, 16.2
Ultracryotomy, 27.1.4 quality, 3.1.1
Ultrastructural cytochemistry, 27 as a solvent, 3.2.1
Problems, 27.1 structure, 3.2.1
Reaction types, 27.2 Wavelength, 3.3.2
immunocytochemistry, 27.3 Waxes, 2.3.2
Ultraviolet irradiation, 9.1.4
Urate, 6.1, 17.7.1
Urea, 4.4, 6.1.1, 25.1.3 Xylene, 3.2.3, 4.3, 16.4.2
Urobilin, 18.1.1 X-ray, 15.3, 17.3, 28
Urobilinogen, 18.1.1 microanalysis, 28.8.10
Uric acid, 6.1, 17.7.1 structural examination, 15.8
U rticaria, 32.4.1
Please note: the numbers in the indices refer to section numbers in the text
Please note: the numbers in the indices refer to section numbers in the text
Quin-2, 28.8.10
Quinacrine, see Atabrine Tetramethylrhodamine Isothiocyanate, 26.2.1
Mustard, see Atabrine Mustard Tetranitro blue tetrazolium, see TNBT
Quinone, 3.3.8 Tetrazolium salt, 3.3.8, 25.1.2
Quinoneimine, 3.3.8 Texas Red, 26.2.1
Thiazin, 16.3
Thiazol Yellow G, 17.7.2
Reactive, 3.3.8 Thioflavine TCN, 21.7, 30.3.2
Rhodamine B, 31.4.1 TG, 31.4.1
Rhodizonate, 17.7.6-7 Thionin, 3.3.9, 6.3.2, 31.2.3
Romanowsky stain, 6.3.2 Titan Yellow, see Thiazol Yellow G
Rosanilin, 3.3.9 TMRITC, see Tetramethylrhodamine Isothio-
Rubeanic acid, 17.7.3, 19.6.3 cyanate
Ruthenium Red, 27.2.1 TNBT, 25.1.2,25.1.3,25.1.4
o-Tolidine, 17.7.3
Toluidine Blue, 6.1.1,21.7,31.3.1
Safranin 0, 3.3.8, 6.1.1, 6.1.6, 28.8.6 Turnbull Blue, see Prussian Blue
SBFI, 28.8.10
Schiffs reagent, 3.3.9,4.5.2, 5.1.6, 19.6.4, 19.6.8,
28.8.6, 30.3.4, 31.4.2 Uvitex 2B, 31.4.2
Fluorescent, 28.8.6, 30.3.4, 31.8.5
Serva Blue G, see Brilliant Indocyanine G
Silver, 19.6.6 Van Gieson, 28.8.5
methenamine, 27.2.1 Victoria Blue B, 6.1.4, 6.1.6
Sirius Red 4B, 21.6
Sirius Red F3B, 21.6, 21.7
SNAFL-1, 28.8.10 Wright stain, 6.3.2
Index of Histochemical and Histological Methods
Please note: the numbers in the indices refer to section numbers in the text