Ilillis- I*'.'.';.!;. MICROSCOPIC HISTOCHEMISTRY Principles and Practice By GEORGE GOMORI, M.D. THE UNIVERSITY OF CHICAGO PRESS The original research work reported in this hook has been sup- ported by grants from the Douglas Smith Foundation for Medical Research of The University of Chicago and from the Pathology Study Section of the United States Public Health Service. THE UNIVERSITY OF CHICAGO COMMITTEE ON PUBLICATIONS IN BIOLOGY AND MEDICINE EMMET B. BAY • LOWELL T. COGGESHALL LESTER R. DRAGSTEDT • FRANKLIN C. McLEAN THOMAS PARK • WILLIAM H. TALIAFERRO The University of Chicago Press, Chicago 37 Cambridge University Press, London, N.W. 1, England W. J. Gage & Co., Limited, Toronto 2B, Canada Copyright 1952 by The University of Chicago. All rights reserved. Published 1952. Composed and printed by The Unfversity of CfflCAGO Press, Chicago, Illinois, U.S.A. No illustration or any part of the text may be reproduced without permission of The University of Chicago Press TABLE OF CONTENTS Introduction 1 PART I. HISTOCHEMICAL METHODS IN GENERAL I. The Nature of the Processes of Identification in Histochemistry 7 II. The Special Features of Histochemical Methods . . 10 III. The Histochemical Routine 14 IV. Controls To Prove Validity of Technique . .' . . 17 V. Quantitation in Histochemistry 20 PART II. SYSTEMATIC HISTOCHEMISTRY VI. Inorganic Substances 29 A. Metallic Elements 29 B. Nonmetallic Elements 46 VII. Organic Substances 50 A. Saccharides 50 General Principles of the Histochemical Demonstration of Saccharides 52 I. Substances Other than Nucleic Acids .... 61 II. Nucleic Acids 77 Appendix: Ascorbic Acid 90 B. Lipids 91 Histochemical Methods for Lipids 95 Physical Methods 95 Chemical Methods 99 C. Proteins, Amino Acids, and Products of Protein Me- tabolism Ill Proteins HI Amino Acid Components of Proteins 112 Antigens 116 Urea 117 Uric Acid US D. Prosthetic Groups 120 Phenolic Substances Especially Polyphenols . . . 120 V 67979 vi Table of Contents Adrenalin 1^^ The EnterochromaflRn Substance 124 Sulfhydryl (Thiol) Groups 128 E. Various Unclassified Organic Substances .... 130 F. Pigments 132 VIII. Enzymes 1^*^ Preparation of Tissues for Enzymatic Reactions . . . 138 Histochemical Reactions for Enzymes 141 1) Oxidative Enzymes 1^0 a) Dehydrogenases 1^0 b) Oxidases 1^^ c) Peroxidases 1"2 Appendix: Unna's Reduktionsorte and Sauerstofforte . 165 2) Hydrolytic Enzymes 167 Methods for Hydrolytic Enzymes in General . . . 167 Phosphatases 172 Alkaline Phosphatases • • 175 5-Nucleotidase 186 Lecithinase (Phospholipase) 187 Acid Phosphatase 189 Phosphamidase • • 1^4 Phosphorylase 197 Zymohexase (aldolase) 197 Sulfatase 198 Esterases 200 I. AHesterases 202 II. Cholinesterases 208 The Results of Histochemical Methods for Esterases . 212 /3-Glucuronidase 216 Carbonic Anhydrase 218 Urease 219 Appendix: Buffers for Use in Histochemistry . . . 219 ADDITIONAL REFERENCES Additional References 225 INDEXES Author Index 253 Subject Index 266 INTRODUCTION Histochemistry is a borderline field between histology and analytical chemistry or biochemistry. Its subject matter is the identification and localization of chemical substances in the tissues on a cytological scale. In the present book the term will be used in a more restricted sense to include only those methods in which the identifying chemical reaction is ob- served directly through the microscope, in tissues of which the architecture is not grossly altered. This definition will eliminate at once two other important ways of approach: ( 1 ) those in which certain morphological structures ( nuclei, mitochondria, etc.) are first separated by physical means, such as differential solubility or centrifugation, and are then analyzed chemically and (2) the ingenious statistical meth- ods developed by Linderstr0m-Lang^ and his school. "Cyto- chemistry" is often used as a synonym; however, it should be reserved for the study of the chemical organization of the cell in general.^ Histochemistry is a young science, although a few histo- chemical reactions have been known for over seventy years (iodine reaction for starch;^ Prussian blue reaction for iron*). Actually, probably most of the staining techniques described since the earliest days are based on some chemical or physi- cochemical interaction between dye and tissue; however, they cannot be called "histochemical" for two reasons: (1) the underlying chemical reactions are not understood, and (2) their results, valuable as they may be for the differenti- ation of various morphological structures, do not convey any information about their chemical constitution. 1. Lmderstr0m-Lang, K.: Bull. New York Acad. Med., 15:719, 1939. 2. Dick, A. T.: Australian Chem. Inst. J. & Proc, 15:294, 1948. 3. Caventou, J. B.: Ann. de chim. et phys., 31:337, 1826. 4. Perls, M.: Virchows Arch. f. path. Anat., 39:42, 1867. 2 Microscopic Histochemistry Although there had been a few surveys of histochemical methods pubhshed between 1920 and 1930, it was Lison^ in 1936 who made the first real effort to organize all existing knowledge into the science of histochemistry. By introducing a systematically critical attitude into the new discipline and by establishing the criteria of validity, he gave histochemis- try the standing of a science and succeeded in clearing away much pseudo-scientific rubbish from the path of advance- ment. He fully deserves to be called the "founder of histo- chemistry." Since the late 1930's histochemistry has undergone a re- markably rapid development. A large number of important new methods have been devised, and older methods have been subjected to critical analysis. Histochemical methods have become a routine in many laboratories of histology and pathology, and the number of papers reporting results ob- tained with their use has been increasing by leaps and bounds. The rate of development is best reflected in the number of histochemical review articles, chapters, and books printed since 1940. Not fewer than thirteen such compre- hensive works^'^ were published between 1941 and 1951, while the corresponding number for the period between 1912 and 1940 is only six,^"^ exclusive of reviews on microinciner- 5. Lison, L.: Histochimie animale (Paris: Gauthier-Villars, 1936). 6. Gersh, I.: Physiol. Rev., 21:242, 1941; Click, D.: Ann. Rev. Biochem., 13:705, 1944; Gomori, G.: J. Mt. Sinai Hosp., 11:317, 1945; Dempsey, E. W., and Wislocki, G. B.: Physiol. Rev., 26:1, 1946; GHck, D.: Tech- niques of histo- and cytochemistry (New York and London: Interscience Publishers, 1949); Glick, D.: Adv. Enzymol., 9:585, 1949; Holter, H.: Oesterreich. Chem. Ztschr., 50:204, 1949; Bradfield, J. R. G.: Biol. Rev. Cambridge Phil. Soc, 25:113, 1950; Bounce, A. L.: Cytochemical founda- tions of enzyme chemistry, in Sunrner and Myrback, The enzymes (New York: Academic Press, 1950); Gomori, G.: Microchemical tests for certain substances other than fats and chromatin, in BoUes Lee, The microtomist's vademecum (11th ed.; London: J. A. Churchill, 1950); Glick, D., Engstrom, A., and Malmstrom, B. G.: Science, 114:253, 1951; Gomori, G.: Histochemi- cal staining methods, in Methods of medical research (Chicago: Year Book Publishers, 1951); and Pearse, A. G. Everson: J. Clin. Path., 4:1, 1951. 7. Macallum, A. B.: Die Methoden der biologischen Mikrochemie, in Abderhalden's Handb. d. biol. Arbeitsmeth., V-2, 1099 (Berlin and Vienna: Introduction 3 ation. There is every indication that histochemistry is emerg- ing now as an independent discipHne with its own theoreti- cal background, methods, and special problems, just as was the case with biochemistry shortly after the turn of the century. Urban & Schwarzenberg, 1912); Prenant, A.: Rev. gen. d. sc, 32:581, 1921; Parat, M.: Biol. Rev., 2:285, 1927; Patzelt, V.: Animale Histochemie, in Klein and Strabinger's Fortschritte der Microchemie (Leipzig and Vienna: F. Deuticke, 1928); and Klein, G.: Praktikum der Histochemie (Berlin: J. Springer, 1929). PART I HISTOCHEMICAL METHODS IN GENERAL CHAPTER I THE NATURE OF THE PROCESSES OF IDENTIFICATION IN HISTOCHEMISTRY Most of the methods used in histochemistry have been bor- rowed from other fields of chemistry and partly modified for the purpose of apphcation to histological preparations. The methods can be divided into several classes, according to the nature of the procedures and phenomena utilized. 1. Chemical— Some of the reactions of this class are the same as those used in analytical chemistry or biochemistry (for Fe+ ++, Cl~, enzymes, etc.). However, modifications of the original technique are often necessary; e.g., determina- tion of melting point cannot be used in histochemistry; color- less or only slightly colored precipitates are transformed, whenever possible, into intensely colored ones for better visibility. 2. Semichemical—Re2iCtions in this class are more or less specific for certain chemically definable substances, but the nature of the reaction is poorly understood (Best's carmine for glycogen; mucicarmine for mucin ) . Mere staining with dyes of acid or basic character, al- though it does reveal something about the acid- or base-com- bining properties of the substance stained (basic dyes for nucleic acids; acid dyes for globins and histones), cannot be called a histochemical method in any true sense of the word. 3. Physical— This class can be subdivided into several sub- classes. a) Staining of fat —This is a purely physical phenomenon of solubility in oil, without any chemical reactions taking place. 8 Microscopic Histochemistry h) Fluorescence— K number of substances show various kinds of fluorescence under ultraviolet light. Some of the most typical are the fading green-blue fluorescence of vita- min A, the golden-yellow fluorescence of enterochromaffin granules, and the brown one of ceroid. Great care should be exercised in the evaluation of the results, and, if possible, the emitted fluorescent light should be analyzed spectroscopi- cally. Staining with fluorescent dyes is not a histochemical method, any more than staining with other, nonfluorescent, dyes. c) Ultraviolet spectrography— This was developed by Caspersson^ and his school. A number of compounds, such as nucleic acids and some amino acids, have characteristic ab- sorption spectra in the ultraviolet and can be identified by them accurately and in a quantitative fashion. A fairly com- plicated and expensive equipment is required, consisting of a suitable ultraviolet light source, a quartz optical system for the microscope, and a highly sensitive microphotometer. Areas of the size of 1 fx^ can be used for analysis. d) X-ray spectrography (Engstrom) .^— This is an even more complicated procedure, with a difficult theoretical background. However, it permits a highly accurate quanti- tation of almost any element (but not of compounds) in very small areas. e ) Spark spectrography.^— The emission spectrum of a part of a tissue section vaporized in a spark gap is analyzed. /) Tracer techniques^— These techniques utilize the emis- 1. Caspersson, T.: Arch. f. Physiol., Vol. 73, Suppl. 8, 1936, and J. Roy. Micr. Soc, 60:8, 1940. 2. Engstrom, A.: Acta radiol., Suppl. 63, 1946. 3. Gerlach, W., and Gerlach, W.: Die chemische Emissionsspektralana- lyse. II. Anwendung in Medizin, Chemie und Mineralogie (Leipzig: Voss, 1933); Policard, A., and Morel, A.: Bull, d'histol. appliq. a la physiol., 9:57, 1932; Policard, A.: Protoplasma, 19:602, 1933; and Scott, G. H., and Wil- liams, P. S.: Proc. Soc. Exper. Biol. & Med., 32:505, 1934. 4. Hamilton, J. G.: J. Appl. Physiol, 12:440, 1941; Simpson, W. L.: Radioactive isotopes, in Cowdry, Laboratory technique in biology and medi- Processes of Identification in Histochemistry 9 sion of energy-rich radiation by suitable isotopes. The tissue sections are mounted directly on a photographic emulsion, which, after a certain length of exposure, is developed. The unreacted silver halide is removed by hypo ( Na thiosulf ate ) , and the section can be stained by one of the conventional histological staining methods. Radioactive substances are revealed by local blackening of the emulsion. 4. Physicochemical.— In this class belong methods which attempt to obtain information about the dissociation con- stants of protein substances by staining them with dyes buffered at various pH levels. As will be shown, the validity of inferences drawn from the results of this method is open to doubt. 5. Microincineration.^— This technique cannot be fitted easily into any of the previously mentioned classes. Consider- able experience is required for the evaluation of spodograms; however, it seems that at least Fe and Si can be relatively easily recognized in the ash. The present book will be concerned only with techniques belonging in classes 1, 2, 3a, Sb, and 4. They require little or nothing in excess of the apparatus found in any reasonably well-equipped laboratory of histology. cine (2d ed.; Baltimore: Williams & Wilkins, 1948); and Kurbatov, J. D., and Pool, M. L.: Chem. Rev., 32:231, 1943. 5. Policard, A., and Okkels, H.: Anat. Rec., 44:349, 1930; Scott, G. H.: Protoplasma, 20:133, 1933, and Am. J. Anat., 53:243, 1933; Policard, A.: Compt. rend. Assoc, anat., 29:463, 1934; Gage, S. H.: Stain Technol., 13:25, 1938; and Scott, G. H.: Biol. Symp., 10:277, 1943. CHAPTER II THE SPECIAL FEATURES OF HISTO- CHEMICAL METHODS The usual requirements of a satisfactory reaction in ana- lytical chemistry are specificity and sensitivity. In histochem- istry, since one of the main objects is accurate localization, two more conditions must be fulfilled. First of all, the chemical substances to be identified must be immobilized at the sites they have occupied in the living tissue. This is no problem as long as substances such as cal- cium phosphate, hemosiderin, lipids, etc., insoluble in well- chosen fixatives, have to be demonstrated. It is a relatively minor problem in the case of large and poorly diflFusible molecules, like proteins and glycogen. These are, as a rule, precipitated by the fixative or emmeshed inextricably in a spongework of other co-precipitated substances before any gross displacement can take place. It must be remarked, however, that minor shifts on a cytological scale are not nec- essarily prevented. As the fixative penetrates into the interior of the tissue, the advancing front of a high-concentration gradient (especially when a fixative which acts partly by dehydration, such as alcohol or acetone, is used) may push certain substances ahead of itself until they are stopped by an impermeable barrier, such as a cell membrane. In this way artifacts of the type of the well-known "glycogen flight" may be produced. They are usually most marked near the surface of the tissue block, where the tissue is hit by a sudden high concentration of the fixative. In the case of easily soluble and highly diffusible sub- stances (ions, sugars, ascorbic acid, urea, etc.), the regular methods of fixation cannot effect an immobilization, even 10 Special Features of Histochemical Methods 11 if the fixative is, applied in the form of a gas (formaldehyde vapor) or if it contains a specific precipitant for the com- pound investigated. The spontaneous diffusion of small mole- cules, together with the violent drift of solutes, caused by the difference in osmotic pressure (global or specific) between tissue fluids and the fixative, will result in a more or less marked distortion in the pattern of distribution of these molecules within a short time. Applying the fixative by means of vascular perfusion may eliminate gross displace- ment but not the intracellular shift of solutes. Such a displacement of highly diffusible substances can be prevented by the use of the freezing-drying technique, origi- nally described by Altmann^ and perfected by Gersh.^ For a detailed description of the apparatus and its use the reader is referred to the bibliography;^— only the principles of the procedure will be given here. Small pieces of fresh tissue are dropped into liquid air (or, even better, into isopentane cooled by liquid air; tempera- ture lower than —150° C. ) where they are frozen solid almost instantly. Subsequently, they are dehydrated in vacuo at a temperature around —30° C. This, dehydration may take from, a few hours to weeks, depending on the temperature, the efficiency of the vacuum, and the size of the tissue, etc. The dehydrated tissue is then embedded in paraffin. The im- portant point is that there is no liquid phase present at any stage of the procedure; therefore, diffusion of solutes cannot take place. The sections may be floated directly on the re- agent (for instance, an alcoholic solution of silver nitrate 1. Altmann, R.: Die Elementarorganismen und ihre Beziehungen zu den Zellen (Leipzig: Veit & Co., 1890). 2. Gersh, L: Anat. Rec, 53:309, 1932, and Bull. Interaat. A. M. Mus., 28:179, 1948. 3. Hoerr, N. L.: Anat. Rec, 65:293, 1936; Hoerr, N. L., and Scott, G. H.: Frozen-dehydration method for histologic fixation, in Glasser, Medical phys- ics (Chicago: Year Book Publishers, 1944); Simpson, W. L.: Anat. Rec, 80:173, 1941; Packer, D. M., and Scott, G. H.: Bull. Intemat. A. M. Mus., 22:85, 1942; and Stowell, R. E.: Stain Technol., 26:105, 1951. 12 Microscopic Histochemistry for Cl~ ) or mounted on slides and processed after the re- moval of paraffin. Obviously, the freezing-drying method offers great advan- tages. Diffusion artifacts and shrinkage are largely elimi- nated; enzymes are very well preserved. However, it also has several disadvantages. The equipment required is bulky and rather expensive; only small tissue fragments can be used ff distortion by ice crystals is not to occur; sections of the em- bedded material are not easy to handle because they are quite sensitive to water. It should be added that, if frozen- dried sections are run down through xylene and alcohols to water in the same way that regular sections are, most of the proteins will not be fixed and will remain soluble. Unless the slides are coated with collodion, a considerable loss of pro- tein substances and glycogen is liable to occur on hydration. A short (2-3 minutes) bath in 70-80 per cent alcohol be- tween the second 95 per cent alcohol and water may be em- ployed to make proteins insoluble. In spite of its disadvantages, the freezing-drying method is a "must" in certain types of histochemical research, and its application has yielded most valuable information regarding the localization of mobile ions in the tissues. The second special condition for histochemical localiza- tion is that the chemical reaction or physical means utilized for identification should possess certain features which can be enumerated as follows: 1. It must be applicable to reactions in situ. Therefore, reactions taking place in solution only (e.g., the Carr-Price^ reaction for vitamin A; the Kober^ reaction for estrogens, etc.) are unsuitable. 2. It must preserve tissue structure. Methods utilizing con- centrated H2SO4 or KOH can have, at best, a very limited value in localization because tissue structure is badly dam- aged by these reagents. 4. Carr, F. H., and Price, E. A.: Biochem. J., 20:497, 1926. 5. Kober, S.: Biochem. Ztschr., 239:209, 1931. Special Features of Histochemical Methods 13 3. The end result produced must be reasonably stable. Soluble or fleeting colors, such as that of the rhodanate re- action^ for Fe+ + + or the change in the shade of indicators caused by acid liberated enzymatically from esters, do not localize with sufiicient accuracy. A reaction to be used in histochemistry must produce highly insoluble and, prefer- ably, intensely colored precipitates. 4. When a reaction produces a crystalline precipitate, the size of the crystals must be small enough to permit cytolog- ical localization. This condition is not fulfilled in the gypsum reaction for Ca^ or in the digitonine reaction for cholesterol.^ 5. In the case of soluble substances, the reaction must be prompt, of a speed not much inferior to that of ionic re- actions. Slow reactions are unusable. To give an extreme example, some otherwise excellent reactions for glucose (osazone formation; the reduction of Benedict's solution) are entirely unsuitable for histochemical purposes. Glucose would diffuse far from its original site before either of these reactions was completed. Failure to recognize these simple principles has resulted in a number of publications reporting the histochemical localization of substances for which suitable reactions are not known.^ 6. Schmelzer, W.: Ztschr. f. wissensch. Mikr., 50:99, 1933. 7. SchujeninofiF, S.: Ztschr. f. Heilk., 18:79, 1897. 8. LeuHer, A., and Noel, R.: Bull, d'histol. appliq. a la physiol., 3:316, 1926. 9. Seeger, P. G.: Arch. f. exper. Zellforsch., 21:308, 1938; Ztschr. f. mikr.- anat. Forsch., 48:181, 639, 1940, and 53:65, 1943. CHAPTER III THE HISTOCHEMICAL ROUTINE The first few steps in handling tissues for histochemical in- vestigations are essentially the same as those used in his- tology, except for minor differences due to chemical con- siderations. In a few cases fixation is not permissible. Certain sensitive enzymes, especially oxidative, are badly damaged by all known fixatives. In such cases the tissues must be used fresh, in an unfixed condition, either as smears or as frozen sections. Such tissues are not easy to handle; they are extremely frag- ile and are readily cytolyzed by many reagents. Further- more, a number of proteins will remain soluble and diffuse from their original sites unless the proper precautions are taken. The only way to keep proteins insoluble but undena- tured is to use strong ( half-saturated or better ) salt solutions [e.g., (NH4)2S04; NaCl; Na acetate] throughout the histo- chemical procedure, up to the last step, when the final pre- cipitate is produced. In some instances, when the enzyme is resistant to drying out and to moderate heat, the freezing- drying method may be the answer to the problem. However, in most cases fixation is possible and preferable. Tissues should be fixed promptly, although a few hours' de- lay, especially if the specimen is refrigerated, seldom causes noticeable changes. The choice of the right fixative is important. Fixatives giv- ing good cytological detail and a minimum of artifacts should be preferred whenever possible. Often, however, a compromise is necessary; one may have to sacrifice cytolog- ical excellence to the preservation of the chemical substance investigated. This applies especially to the preservation of 14 The Histochemical Routine 15 enzymes. The correct fixative or fixatives will be specified in the description of the individual techniques. The fixed tissue is either cut on the freezing microtome or embedded in paraffin or celloidin. Formalin-fixed tissues us- ually give excellent frozen sections, while, after alcohol or especially acetone fixation, frozen sections are very delicate and easily break to pieces. Occasionally, especially when an enzyme is not too sensitive to a relatively short fixation but is sensitive to a long exposure to dehydrating agents and/or heat, a sort of semi-embedding may be carried out as follows: thin slices of the tissue are dehydrated by several changes of absolute alcohol or acetone, a few hours each; subse- quently they are transferred to a mixture of equal parts of alcohol and ether for 2 hours, followed by about 4 per cent celloidin in alcohol-ether (Collodion, U.S. P.) for 12-24 hours, hardened in 70 per cent alcohol for a few hours, and finally transferred to water. The entire procedure is prefer- ably carried out at icebox temperature. Tissue blocks infil- trated with thin celloidin by this method can be cut on the freezing microtome; the sections have an excellent con- sistency and no tendency to break up. In most cases paraffin or celloidin embedding is possible, and both techniques have their advantages and disad- vantages. Celloidin embedding does not require the appli- cation of heat, and this may be an important advantage when dealing with heat-sensitive enzymes. However, dilute al- cohol, used in the storage of celloidin blocks, is very detri- mental to enz)Tiies, and most of them will be destroyed by it in a short time; therefore, it is imperative to cut and process the blocks promptly. Paraffin embedding usually causes considerable inacti- vation of enzymes, although in a completely anhydrous state many proteins will resist the denaturing eflFect of heat re- markably well. However, even traces of water left in the tissue will considerably accelerate the rate of inactivation by heat. That is why thorough dehydration of the tissue is so 16 Microscopic Histochemistry important. It is always advisable to avoid excessive heat; the temperature of the paraffin oven should not exceed 56°- 58 °C. Even at this temperature tissues should not be ex- posed to heat longer than absolutely necessary. The vacuum technique, to be described in the section on enzymes, will cut down considerably the time required for embedding. Paraffin sections should be floated on lukewarm water and attached to the slide either without any adhesive or with P. Mayer's egg-white-glycerol mixture. After complete drying, it is advisable to place the slides in the paraffin oven for a few minutes until they melt. The melted paraffin forms an excellent coating on the surface and protects the tissue from the injurious effect of atmospheric oxygen and moisture. The sections are dewaxed in xylene and carried through alcohols ( absolute and 95 per cent ) to water as usual. Often it is advisable to protect the tissue with a thin layer of col- lodion. This is done by flooding the slide after the last alcohol with a dilute (about /2 per cent) solution of celloidin in alcohol-ether, shaking off the excess and hardening the mem- brane in 80-95 per cent alcohol. The collodion membrane serves two purposes : ( 1 ) it prevents diffusion of large mole- cules not made insoluble by fixation, and (2) it facilitates the removal of nonspecffic surface precipitates, a by-prod- uct of some histochemical reactions. These precipitates will settle on the surface and can be washed off by dissolving the membrane in alcohol-ether or acetone. CHAPTER IV CONTROLS TO PROVE VALIDITY OF TECHNIQUE In a number of cases the reaction used is so highly specific that no control is necessary; a positive reaction indicates the presence of the substance searched for with absolute cer- tainty (for instance, the Prussian blue reaction for Fe"^ + + ). In other cases, however, "blanks" must be run, as in ana- lytical chemistry or biochemistry, to avoid confusing the genuine reaction with other similar reactions of a nonspecific nature. These blank runs are especially important in the identification of enzymes. The two main methods for the verification of the speci- ficity of enzymatic reactions are ( 1 ) the omission of essential ingredients (e.g., the substrate or Ca ions in the technique for alkaline phosphatase) and (2) the use of inactivators or inhibitors, such as excessive heat, strong acids, oxidants, fluoride, eserine, etc., depending on the nature of the enzyme investigated. Whatever reaction persists after such treatment cannot be due to enzymatic activity. One of the difiiculties of identification in histochemistry is the impracticability of applying reactions to purified sub- stances. The compounds investigated almost invariably occur in association with, and often adsorbed on, other compounds. The presence of these may profoundly modify the typical reactions, solubility, color, and other properties of the com- pound investigated as listed in textbooks of analytical chem- istry. In addition, fixatives may cause such significant changes in the reactive groups as to make them unrecogniz- able by the accepted identifying reactions. These are the main reasons why test-tube reactions, even if they otherwise 17 18 Microscopic Histochemistry meet the standards previously mentioned, may give rise to misleading results when applied to tissues. The applicabihty of procedures of analytic chemistry to histochemical research can be tested by model experiments. Such experiments attempt to carry out the identifying re- actions under conditions more or less similar to those pre- vailing in tissue sections. The first, rather primitive, model experiment is credited to Altmann,^ who investigated the differential staining reactions of various fatty substances im- bibed by a piece of tissue paper. In more accurate experi- ments the substance in question is dissolved or finely dis- persed in agar, gelatin, or some similar substance. The sus- pension can be smeared on shdes or allowed to gel, fixed and embedded like any tissue block. Another clever tech- nique has been devised by Coujard,- permitting the com- parison of a large number of substances on a single slide. The substances to be tested are dissolved in serum, dilute gelatin, or egg-white or some other freely flowing protein solution, and marks are made with a clean steel pen on a carefully cleaned shde, using the solutions as ink. The use of different symbols as marks for the different substances (e.g., abbreviations of their names, chemical formulas) will facilitate prompt and easy recognition of the marks. As soon as the shdes are dry, they can be processed as if they were smears. Coujard's method assures the chemical comparison of many different test substances under strictly identical conditions. Model experiments find a number of applications in histo- chemistry, of which a few will be mentioned. 1. Determining the chemical specificity of methods —Ex- amples of this will be mentioned in the sections on nucleic acids, lipids, and phenolic substances. Each simple histolog- ical staining method with no known chemical background 1. Altmann, R.: Die Elementarorganismen und ihre Beziehungen zu den Zellen (Leipzig: Veit & Co., 1890). 2. Coujard, R.: Bull, d'histol. appliq. a la physiol., 20:161, 1943. Controls To Prove Validity of Technique 19 can be tested. For instance, it can be shown that the staining of beta cell granules in the pancreatic islets by chrome hematoxylin^ is not due to their insulin content, since marks made with commercial insulin and fixed in Bouin's fluid just like a piece of pancreas do not stain. 2. Studying the effects of fixation and embedding.— Model slides (carrying, e.g., marks made with enzyme solutions) can be treated with any combination of fixatives, dehydrating and clearing agents, hot paraffin, etc., and the timing can be varied within wide limits. The effect of these procedures can be judged not only qualitatively but, to a certain extent, even quantitatively ( see next chapter ) by the outcome of the reaction. Such simple experiments may supplement or even replace those laborious studies in which tissue blocks are used and enzymatic activity, after various treatments, is de- termined chemically, by test-tube methods. 3. Quantitation of histochemical methods.— The important problem of quantitation in histochemistry and the role of model experiments in it will be discussed in the next chapter. 3. Gomori, G.: Am. J. Path., 15:497, 1939. CHAPTER V QUANTITATION IN HISTOCHEMISTRY The greatest advantage of biochemical over histochemical methods is the far superior abihty of the foiTner to quantitate results. However, as v^ill be show^n, a modest degree of quan- titation can be achieved also by purely histochemical tech- niques. Absorption colorimentry is the main tool of quantitation in analytical chemistry, biochemistry, and histochemistry. How^- ever, conditions prevailing in microscopic sections only rarely permit the theoretically correct application of color- imetry. One of the fundamental principles of absorption color- imetry is that the distribution of the absorbing material in the sample must be uniform. If this condition is fulfilled, quantitative evaluation of photometric readings obtained through the microscope is an entirely correct procedure, and it has given much valuable information especially in the hands of Caspersson^ and his school and a fev^ others.^ It should be remarked, how^ever, that areas sufiiciently uni- form in optical density are, as a rule, very small, usually occupying only a minute fraction of an oil-immersion field. The minimum diameter required for a measurement is about four times that of the w^ave length of the light used. Irregular distribution of the absorbing material vv^ill lead 1. Caspersson, T.: Arch. f. Physiol., Vol. 73, Suppl. 8, 1936, and J. Roy. Micr. Soc, 60:8, 1940. 2. Gersh, I., and Baker, R. F.: J. Cell. & Comp. Physiol., 21:213, 1943; Stowell, R. E.: J. Nat. Cancer Inst, 3:11, 1942, and Anat. Rec, 91:301, 1945; Ris, H., and Mirsky, A. E.: J. Gen. Physiol., 33:125, 1949; Swift, H. H.: Physiol. ZooL, 23:169, 1950, and Proc. Nat. Acad. Sc, 36:643, 1950; and Hoover, C. R., and Thomas, L. E.: J. Nat. Cancer Inst., 10:1375, 1950. 20 Quantitation in Histochemistry 21 to grave errors in quantitation, especially in the case of sub- stances with a high optical density, i.e., of dark shades. An extreme example will be given to shown this point. It will be taken for granted that the sections are of a very uniform thickness, a condition not easy to fufil. Let us assume that the field is occupied by a homogeneous colored layer the light transmission of which it 5 per cent. If the same amount of colored material is distributed in dis- continuous spots occupying only half the field, transmission of the entire area will go up to 50 + 2.5 = 52.5 per cent. If T=5% T=52.5% T=90.5% W^M ■ Fig. 1 it is distributed over only one-tenth of the field, transmission will rise to 90 + 0.5 = 90.5 per cent ( see Fig. 1 ) . In this, way the same amount of colored matter distributed over the same area in different ways may read 5 per cent, 52.5 per cent, or 90.5 per cent, which is a spread of eighteen fold in terms of transmission. The error decreases rapidly with increasing transparency of the color. If the transmission of the uniform layer is not 5 but 50 per cent, the corresponding readings will be 50, 75, and 95 per cent; and if the uniform transmission is 80 per cent, they will be 80, 90, and 98 per cent. However, since the absolute amount of colored matter is a function of the loga- rithm of transmission, the actual error in quantitation would be 30:1, 13.5:1, and 12:1, respectively. 22 Microscopic Histochemistry It follows that readings of transmission obtained through a variegated area have a very hmited quantitative signifi- cance, except in the case of reasonably uniform colored par- ticles of good transparency. Such relatively favorable condi- tions obtain, for instance, in thin sections stained by the Bauer-Feulgen method for glycogen,^ as could be shown by the fairly satisfactory agreement between colorimetric read- ings and the results of chemical analysis. The difiiculties and pitfalls of quantitation by absorption colorimetry were adeptly summarized by Click, Engstrom, and Malmstrom.* The same principle holds true for semiquantitative judg- ments arrived at from gross inspection. A well-known ex- ample of this is the change in color of the skin of some am- phibians and fish. The same fish may appear almost black or practically white, depending on the state of expansion or contraction of the chromatophores, although the amount of pigment per unit area remains unchanged. These simple facts are often ignored in histochemistry. In a semiquantitative way, a reaction is often called intense or the tissue is stated to contain large amounts of a substance if the section shows a widespread reaction. Quantitatively, transmission of areas grossly variegated in black and white is measured by photometers, and conclusions as to the con- centration of substances are drawn from the data.^ This practice is incorrect scientifically. For a semiquantitative (accuracy about ±50 per cent) evaluation of microscopic color reaction, model experiments can be used to great advantage. It is likely that under suit- able conditions even true photometric quantitation could be achieved. This has been attempted by Marza and Chiosa^ 3. Deane, H. W., Nesbett, F. B., and Hastings, A. B.: Proc. Soc. Exper. Biol. & Med., 63:401, 1946. 4. Click, D., Engstrom, A., and Malmstrom, B. G.: Science, 114:253, 1951. 5. Cleland, K. W.: Proc. Linnean Soc, N.S. Wales, 75:35, 1950. 6. Marza, V. D., and Chiosa, L.: Bull, d'histol. appliq. a la physiol., 12:58, 1935. Quantitation in Histochemistry 23 ( quantitative determination of potassium in tissue sections ) , but further investigations will be required before the value of this procedure is established. For such approximate estimations, model slides serve as standards, and the Coujard technique o£Fers a simple ap- proach. The principles of the method v^ill be illustrated by describing its application to the quantitation of histochem- ical reactions for enzymes.'^ First of all, a highly active enzyme preparation is made according to one of the accepted methods, and its activity is assayed accurately. Serial dilutions by a factor of 2 are made with a suitable diluent, such as a 1 per cent gelatin or gum acacia solution. Standard slides are prepared as follows: carefully cleaned microscopic slides are coated thinly with egg-white glycerol, just as for histological purposes. They are subsequently heated over a Bunsen flame until com- pletely dry. This pretreatment will prevent the running of ink when the marks are made. With a clean steel pen, marks are made on the slides, using the serial dilutions as ink. To avoid confusion, the dilution fraction can be used as a mark for each "ink." Every slide will carry the marks of the entire dilution series. The slides are dried, fixed for a few hours in alcohol or acetone, coated with thin (about 0.1 per cent) collodion, and washed. A number of them are incubated, to- gether with the same number of regular histological slides, in the substrate solution. At intervals (e.g., 5, 10, 20, 40 minutes, etc. ) one tissue slide and a corresponding standard are removed from the incubating mixture, and the color is developed. The pair of slides in :which the tissue structure in question first shows up in the shade chosen (for simple in- spection, preferably black; for colorimetric measurement, any shade lighter than black) is used for comparison. The mark made with the lowest dilution and showing in the shade chosen will have approximately the same activity per unit area as the histological detail in question. It is important 7. Gomori, G.: Exper. Cell Research, 1:33, 1950. 24 Microscopic Histochemistry that the time at which a tissue detail first shows up in black should not be missed, since estimation of further increase in color development beyond the level of black is impossible. Nothing can be blacker than black. The specific enzymatic activity (activity /ml of tissue) of a structural detail ( granule, fiber, brush border, etc. ) appear- ing in a uniform shade can be calculated on the basis of the 800 30 40 50 60 70 80 90 100 IIO Thickness of Lines, Micro Fig. 2 principle that the colorimetric density of a homogeneous layer is proportional to the absolute amount of colored ma- terial per unit of projection area. Therefore, the specific ac- tivity of any detail will be identical with that of the mark showing up in the same shade, multiplied by H/h, where H stands for the thickness of the fluid layer of the mark and h for the thickness of the tissue section, provided that the structure occupies the entire thickness of the section. The value for H can be obtained from the empirical curve given in Figure 2, after measuring the average width of the mark Quantitation in Histochemistry 25 lines with an eyepiece micrometer. For the derivation of the curve and for the possible sources of error inherent in this method the reader is referred to the original paper.*^ It should be stressed that, with methods of the type de- scribed, quantitation applies only to tissue details appearing in a uniform shade, and over-all values for the entire speci- men cannot be obtained. For such over-all values some method like that of Doyle,^ in which the reaction products are extracted from the entire section and their amount is determined chemically by test-tube assay, must be used. A new method,^ based on the introduction of a radioactive isotope into the precipitate obtained enzymatically, also ap- pears to be theoretically correct for over-all quantitation. With suitable minor modifications, model experiments can be employed in several of the quantitative aspects of histo- chemistry. 1. The sensitivity of a method, expressed in terms of mini- mal concentration of a substance still giving a recognizable reaction, can be determined. 2. The concentration of chemical constituents in tissue elements can be calculated. 3. The quantitative eflFect of procedures of fixation and embedding, especially loss in enzymatic activity, can be studied. 4. The kinetics of enzymatic reactions in tissue sections can be followed, and results obtained with Doyle's method can be checked and supplemented. 8. Doyle, W. L.: Science, 111:64, 1950; Doyle, W. L., Omoto, J. H., and Doyle, M. E.: Exper. Cell Research, 2:20, 1951; and Doyle, W. L.: Quanti- tative aspects of the histochemistry of phosphatases, in Symposium on cytol- ogy (East Lansing: Michigan State College Press, 1951). 9. Barka, T., Szalay, S., Posalaky, Z., and Kertesz, L.: Kiserletes orvostud., p. 1, 1951. PART II SYSTEMATIC HISTOCHEMISTRY The part of the book which follows contains the descrip- tion and critical evaluation of histochemical techniques for various substances. No claim is made to include all tech- niques ever described; in fact, for each substance only one or a few methods of proved value will be recommended, al- though some untested methods and methods of historical or theoretical interest will also be mentioned. The techniques as given in detail do not necessarily follow the exact specifi- cations of their authors; they may be slight modifications, found to be more satisfactory or simpler than the original procedures. CHAPTER VI INORGANIC SUBSTANCES Inorganic constituents occur in tissues in three forms: (1) soluble, diffusible, and, for all practical purposes, completely ionized; (2) insoluble but readily convertible into the solu- ble form; and (3) incorporated into complex, soluble, but poorly diffusible or insoluble organic molecules. The last form is often called "occult" or "masked" because the regular reactions of the inorganic part are not obtainable unless the organic matrix is first more or less completely destroyed. Diffusible substances can be localized only by the use of the freezing-drying method. Even so, localization is often only approximate on account of their high mobility, causing noticeable displacement of the solute before quantitative precipitation by the reagent can take place. Some substances of the second group are capable of giving direct ionic reactions without being dissolved first; others must be treated with acid to make them soluble. In the latter case the reagent must be present in the acid from the very beginning, to bind the ions as fast as they are formed in the course of solution. The demonstration of "masked" constituents presents considerable diflBculties. Unmasking agents, with a few ex- ceptions, are quite harsh chemicals (or intense heat) which may destroy tissue architecture beyond recognition. A. METALLIC ELEMENTS Sodium.— No method is available for the demonstration of this element. The demonstration of "sodium chloride" by the AgNOs technique^ is such a naive idea that it does not de- serve serious consideration. 1. Seeger, P. G.: Ztschr. f. mikr.-anat. Forsch., 53:65, 1943. 29 30 Microscopic Histochemistry Potassium— This element exists partly (probably to a small extent) in a poorly ionized, protein-bound, nondiffusible form, while the bulk of it is diffusible and ionized. Whether or not both fractions are demonstrable histochemically is not known. The reagent used is sodium cobaltinitrite, which forms with potassium a microcrystalhne precipitate of potas- sium cobaltinitrite, orange in color. For better visibility it is transformed in a second step into black cobalt sulfide. The reaction is quite specific for potassium. Originally it was believed that creatine could give a spurious reaction; more recent findings, however, indicate that the precipitate was caused by traces of potassium in the creatine sample used^ Pure creatine gives no precipitate. Localization is fair to good. Method (Macallurns,^ modified) Reagent -Solution A, dissolve 5 g. of cobalt nitrate in a mixture of 10 ml. of distilled water and 2.5 ml. of acetic acid; solution B, dissolve 15 g. of sodium nitrite in 25 ml. of dis- tilled water. For use, pour solutions A and B together, shake for a few minutes until the bulk of the nitrous fumes which develop on mixing has escaped. Chill the reagent, one dish of distilled water and three to four dishes of 50-70 per cent alcohol in an ice bath. Place small pieces of fresh tissues in the reagent for about 2 minutes. Rinse briefly in distilled water, followed by thorough rinses in the alcohols. Transfer to a dilute solution of yellow ammonium sulfide (about 1 drop to each 5 or 10 ml. of distilled water) for about 2 min- utes. Wash, counters tain as desired, dehydrate, and mount. A black, granular precipitate indicates the sites of potassium. Thorough rinsing of the tissue in dilute alcohol is very important if one wants to avoid a disturbing gray back- ground due to adsorbed cobalt. The mixed reagent can be kept for only a few days. 2. Macallum, A. B.: Australian J. Exper. Biol., 9:159, 1932. 3. MacaUum, A. B.: J. Physiol., 32:95, 1905. Inorganic Substances 31 Another reaction for potassium has been described by Carere-Comes/ based on the dipicrylamine test of Poluek- toff.^ Formalin fixation is recommended by the author. No pre- cautions are taken in the course of further handhng of the tissue. Under such conditions, only a minute fraction of potassium will remain at the original sites, and the technique would have little value even if the reaction were a satis- factory one, which it is not. If a drop of a dilute solution of KCl is mixed with a drop of the reagent on a slide, one can observe under the microscope that precipitation of the orange-red K-dipicrylamine crystals starts only after about 90 seconds, and complete precipitation requires several min- utes. The crystals formed are very coarse. The criticism of Claesson^ that this method is unsuitable for the identification and locahzation of potassium appears to be fully justified. The explanation of "positive reactions" is this: The reagent is a good plasma stain, antiquated now but much used during the pioneer period of histology under the name of "Aurantia." It will stain all acidophilic structures, some of which (muscle, red cells ) happen to be rich in potassium. Calcium— Soluble calcium could possibly be demonstrated by treating frozen-dried sections with a solution of ammo- nium oxalate, although there is no record of this ever having been attempted. RabP proposed the use of a fixative com- posed of formalin and oxalate; however, localization under such conditions would be far from accurate. The octahedric crystals of calcium oxalate are easily recognizable. For insoluble deposits of Ca salts it is important to use neutral fixatives because both Ca phosphate and carbonate are acid-soluble. Neutral formalin or alcohol or their mixtures are suitable fixatives. 4. Carere-Comes, O.: Ztsclir. f. wissensch. Mikr., 55:1, 1938. 5. Poluektoff, N. S.: Mikrochemie, 14:265, 1933. 6. Claesson, L.: Acta anat., 3:1, 1947. 7. Rabl, C. R. H.: Klin. V^chnschr., 2:1644, 1923. 32 Microscopic Histochemistry The methods used for the demonstration of insoluble Ca fall into two classes: ( 1 ) those which are specific for Ca itself (the gypsum, oxalate, and lake-dye reactions) and (2) those which demonstrate the anions of calcareous deposits (heavy- metal methods ) . The only absolutely specific reaction is Schujeninoff's.^ This is based on the formation of insoluble crystals of gypsum (CaS04) by the action of H2SO4 in a dilute alcoholic me- dium. Method Cover the section with 50 per cent alcohol; place a drop of 5-10 per cent H2SO4 on the under surface of a cover slip and put it on the section. Watch slide under the microscope; almost immediately typical rhombic crystals, very prone to aggregate in swallowtail or rosette-like formations, will appear. The sections cannot be counterstained, and the crystals are far too coarse to permit exact localization. Several hydroxyanthraquinone dyes ( alizarinsulf onic acid, purpurin,^ anthrapurpurin ) ^^ give intensely colored ( reddish or purple ) insoluble lakes with calcium and can be used for the demonstration of deposits of a medium particle size. Relatively large, dense structures such as bone spicules are poorly penetrated; in the case of fine, dustlike deposits the shades are not intense enough to be seen distinctly. This applies also to gallamin blue,^^ a lake dye of another chemi- cal group. Anthraquinone lake dyes can be used in the form of a 0.1-0.5 per cent solution in 50 per cent alcohol; gallamin blue in 0.1-0.2 per cent aqueous solution in an M/5 borate buffer of pH 7.6. Staining time is several hours. For counter- staining, one of the plasma stains ( anilin blue or light green 8. SchujeninofF, S.: Ztschr. f. Heilk. 18:79, 1897. 9. Grandis, V., and Mainini, C: Arch. ital. de biol., 34:73, 1900. 10. Salomon, H.: Jahrb. f. wissensch. Bot., 54:308, 1914. 11. Stock, A.: J. Roy. Micr. Soc, 69:20, 1949. Inorganic Substances 33 for the anthraquinone lakes, eosin for gallamin blue ) should be used, because basic dyes may be adsorbed to sites previ- ously stained by the lake, giving dark, murky shades. The method described by Cretin^- is quite specific and gives very sharp pictures; how^ever, the preparation of the highly unstable reagent is difficult. The method is so ca- pricious as to be a curiosity rather than a dependable test. Since in animal tissues almost all insoluble Ca is in the form of phosphate and carbonate and, conversely, practi- cally all insoluble phosphate and carbonate are Ca salts, any method which demonstrates these two anions is reasonably specific for Ca. With plant tissues, which may contain large amounts of Mg phosphate, this does not hold. The phosphates and carbonates of almost all heavier metals are insoluble, and many of them are convertible into intensely colored compounds. Thus there is a wide choice of reagents, as enumerated by Stoeltzner.^^ To mention a few: Ag-> metallic Ag (black); Co->CoS (black); Cu-^ Cu2Fe(CN) 6 (red-brown); Fe++ -^-Fe3++ (Fe+ + + [CN]6)2 (TurnbulFs blue); etc. Of these, only the silver technique is employed extensively, although at times the others may also be useful, especially when a shade other than black is desired. The silver technique ( Salge and Stoeltzner,^* Kossa^^ ) . Method Before using the silver solution, rinse the slide thoroughly in distilled water. Immerse the slide for 5-10 minutes in a 0.2-1 per cent solution of AgNOs. At the sites of Ca phos- phate-carbonate the corresponding silver salt, yellowish in color, will form. This can be reduced to metallic silver either by exposing the jar to direct sunlight (or to the light of an 12. Cretin, A.; Bull, d'liistol. appliq. a la physiol., 1:125, 1924. 13. Stoeltzner, W.: Virchows Arch. f. path. Anat., 180:362, 1905. 14. Salge, B., and Stoeltzner, W.: Berl. klin. V^^chnschr., 37:298, 1900. 15. Kossa, J. von: Beitr. z. path. Anat. u. z. allg. Path., 29:163, 1901. 34 Microscopic Histochemistry ultraviolet light source) until the precipitate appears black or by using photographic developers. Reduction of the silver salt by light has certain disadvan- tages: (1) a brownish halo may sometimes form around the black granules, or (2) the black precipitate may turn brown when the unreacted silver is removed. For these reasons, some workers prefer to use photographic developers. The sHdes are first thoroughly washed in many changes of dis- tilled water and then immersed in a dilute solution of hydro- quinone or pyrogallol (about 0.5 per cent; concentration not important) for about 2 minutes and rinsed again. After both methods of reduction, the unreacted silver must be removed by a short bath in a thiosulfate (hypo) solution (about 2 per cent); otherwise the shdes may darken later. After photo- graphic development the slide presents a very sharp contrast between black and white, no intermediate shades being pres- ent; however, failure to wash the slide very thoroughly be- fore the use of the developer will result in the deposition of a very fine, dustlike black precipitate all over the tissue. Slides prepared in either way can be counterstained as desired, de- hydrated, and mounted. The only source of error with the silver method is the pres- ence of massive deposits of uric acid and of its salts, which may stain very much like phosphate-carbonate. For differ- entiation between urate and phosphate-carbonate see the section on uric acid. The other heavy-metal techniques are performed in an analogous way. First, the section is immersed for %-l hour in the solution of a salt of the heavy metal (e.g., cobalt nitrate, ferrous sulfate, etc.), then washed thoroughly and treated with the reagent (ammonium sulfide, acidified po- tassium f erricyanide, etc. ) . The two main disadvantages of these techniques are that (1) the proteins may retain the heavy metal rather stubbornly and in this way give rise to a more or less intensely colored background and (2) the pene- tration of heavy-metal salts into dense granules or spicules Inorganic Substances 35 is very limited. The protein error is very marked in the case of native proteins (unfixed tissues), whereas after a good fix- ation it is usually negligible unless the slide is exposed to the heavy-metal salt for many hours or even days. The intense blue staining of calcified structures, such as bone matrix, by various hematoxylin lakes (hemalum, iron, and chrome hematoxylin) is not specific for any component of bone salt.^^ Neither Ca carbonate nor phosphate will stain with hematoxylin; on the other hand, bone matrix will stain even after complete removal of bone salt by a strong acid. The afiinity of bone matrix for hematoxylin lakes is due to the chemical constitution of the organic framework (pres- ence of mucopolysaccharides?), specifically, to its ability to bind the metallic component. Masked Ca (mainly in cell nuclei) can be demonstrated only by microincineration. Barium and strontium.— Wa.terhouse^'^ recommends a solu- tion of sodium rhodizonate in distilled water or in a buffer of pH 7 as a reagent for Ba and Sr. These two ions give red- brown precipitates with rhodizonate. According to the au- thor, the reaction with Ba can be prevented or abolished, respectively, by treating the section with a solution of sodium chromate; the reaction with Sr is not affected. The specificity of this test requires confirmation. Water- house finds that, under the conditions specified, Ca will not react. This is not entirely correct; both bone salt and freshly precipitated pure Ca phosphate stain in a distinct ochre shade with rhodizonate. It is questionable whether Ba and Sr can ever be recognized by the rhodizonate test in the presence of excess Ca. Furthermore, it is known^* that Pb 16. Schuscik, O.: Ztschr. f. wissensch. Mikr., 37:215, 1920; Cameron, G. R.: J. Path. & Bact., 193:929, 1930. 17. Waterhouse, D. F.: Nature, 167:358, 1951, and Australian J. Scient. Research, B, 4:145, 1951. 18. Feigl, F.: Chemistry of specific, selective, and sensitive reactions (New York: Academic Press, 1949). 36 Microscopic Histochemistry and Hg give a precipitate of a shade very similar to that obtained with Ba. Magnesium.— Quinsilizsinn, titan yellow, and a number of azo dyes have been suggested for the demonstration of this element. ^^ In model experiments some of them appear to be sufficiently specific; however, in sections of animal tissues no staining can be obtained with any of them. Presumably their sensitivity is relatively low. Iron— This metal occurs in two forms in animal tissue. In one form, represented by hemosiderin, it behaves like any poorly soluble inorganic ferric compound, such as ferric oxide, readily demonstrable by the common reagents of ana- lytical chemistry. In the other form (occult or masked iron), exemplified by hemoglobin, iron is a part of complex organic molecules and demonstrable only after destruction of the organic part. Iron appears to be present in the tissues ex- clusively in the ferric state, although there are a few reports on the finding of ferrous iron. For the demonstration of iron, tissues should be fixed in a neutral fixative, such as neutral formalin, alcohol, or a mix- ture of the two. Although hemosiderin is far less acid-soluble than Ca phosphate, it is attacked to a noticeable degree by acid-containing fixatives, such as Bouin's fluid. The result is not only blurring of the picture but also a false localization. Certain morphological structures, especially nuclei^^ and, to a lesser extent, coarse connective-tissue fibers, have an amazing affinity for the ferric ion and will adsorb it even from ex- tremely dilute solutions (Gilson,^^ Wiener^^) and hold it tenaciously. The oldest technique for the demonstration of iron is • 19. Broda, B.: Wiadomosci Farm., 63:6 and 15, 1936; and Okamoto, K., Seno, M., and Shibata, D.: Taishitzu Gaku Zasshi, 13:97, 1944. 20. Macallum, A. B.: Quart. J. Micr. Sc, 38:175, 1896, and Proc. Roy. Soc. London, 50:277, 1891. 21. Gilson, G.: Rep. British A. Adv. Sc, p. 778, 1892. 22. W^iener, A.: Biochem. Ztschr., 77:27, 1916. Inorganic Substances 37 Quincke's^^ iron sulfide test (A. Mayer, 1850). It is based on the formation of green-black ferrous sulfide by the action of ammonium sulfide. Method Dilute ammonium sulfide ( colorless or light yellow; darker samples do not work well) with 5-10 volumes of distilled water. Immerse sections for 20-30 minutes. Wash briefly under the tap; counterstain with a red nuclear stain (safra- nine, neutral red, lithium carmine); wash once more, dehy- drate, and mount. Sites of hemosiderin iron show up in a very dense green-black shade. This method is incompatible with the use of mercury-con- taining fixatives. Even after treatment with iodine, enough mercury may be left in the tissue to give a blackish precipi- tate with the reagent. Ferrous sulfide is quite sensitive to acids; even dilute hydrochloric acid will remove it in a mat- ter of minutes; this property will differentiate it from the acid-resistant sulfides of other metals ( lead, bismuth, etc. ) . Even an acidic dye solution used for counterstaining, such as alum carmine, may bleach the finest granules. It has been asserted that the ferrous sulfide method is more sensitive than the Prussian blue method (the description of which will follow) because it unmasks certain iron com- pounds which do not react with the latter. However, on care- ful comparison of consecutive serial sections stained accord- ing to the two techniques, it becomes clear that the efficiency of the two methods is the same, except for the fact that fer- rous sulfide is quite opaque, while Prussian blue is somewhat transparent. The most reliable reagent for the hemosiderin type of iron is an acidified solution of potassium ferrocyanide ( Perls ).^* Ferric ions, released by the action of the acid, are trapped by 23. Quincke, H.: Arch. f. klin. Med., 25:567, 1880, and Arch. f. exper. Path. u. Pharmakol, 37:183, 1896. 24. Perls, M.: Virchows Arch. f. path. Anat., 39:42, 1867. 38 Microscopic Histochemistry the ferrocyanide as fast as they are formed, provided that ferrocyanide is really present "on the premises." Since it dif- fuses much more slowly than the acid (and very poorly through collodion), it may arrive too late to bind the ferric ions where formed, and some diffusion of the latter may take place. For this reason, ferrocyanide must be given a head start. The following technique will invariably yield sharp pictures. Method Fixation in a neutral fluid. Celloidin must be removed from sections of celloidin-embedded material. Place slides (or frozen sections) in a fresh and filtered 5-10 per cent solution of potassium ferrocyanide. After 5 minutes add about M volume of a 10 per cent solution of hydrochloric acid. The latter must be of a good analytical grade and contain no iron (produce no visible greenish or bluish tinge when mixed with the ferrocyanide). Stir the mixture. Keep sections in it for about 20-30 minutes. Wash under the tap. Counterstain with a red nuclear dye (alum carmine, safranine, neutral red, but not lithium carmine, which will bleach Prussian blue), dehydrate, and mount. Hemosiderin iron will show up in an intense blue shade. Balsam of Canada may cause gradual fading of the stain over a period of months or years; the more modern synthetic mounting media are safe. Tirmann^^ suggested the combination of the preceding two methods, to utilize the advantages of the absolute specificity of the Prussian blue method and the purportedly higher sen- sitivity of the ferrous sulfide method. He proceeds as follows: The section is first treated with ammonium sulfide, washed, and then subjected to the action of an acidified solution of potassium ferricyanide, which will convert ferrous sulfide into TurnbulFs blue ( ferrous ferricyanide ) , of a shade indis- 25. Tirmann, J.: Gorbersdorfer Veroffentl., 2:101, 1898. Inorganic Substances 39 tinguishable from that of Prussian blue (ferric f errocyanide ) . This technique is not recommended because it invariably produces minor, and sometimes major, artifacts. ^^ Ferrous sulfide is attacked by the acid in an explosive way, resulting in the deformation of the solid granules into balloon- and burr like structures with a light center and dark outlines. It should be mentioned here that no reaction is ever ob- tained in animal tissues under either normal or pathological conditions if acidified ferricyanide is applied directly to the sections. ^*^ This shows absence of ferrous iron. A few other methods will be mentioned briefly. Macallum^ ' suggests the use of hematoxylin, which forms a blue-black lake with iron. It is easy to show that under strictly neutral conditions hematoxylin will not stain any iron in the tissues. Positive reactions are obtained only after treatment with acid (exposure to sulfuric acid-alcohol,-^ fixation in Bouin's fluid or even in unneutralized formalin ) . Such treatment is wrong in principle, as has been pointed out. In addition to pro- ducing diffusion artifacts, the specificity of the method is poor. Other m^etals such as lead and copper (under experi- mental and pathological conditions) will give reactions indis- tinguishable from that of iron. Moreover, dichromate fixation may cause unsaturated fatty substances (myelin, etc.) to stain very intensely.-^ The thiocyanate technique^^ is unsuitable for localization because the red coloration obtained with it is highly diffusi- ble. Schmelzer^^ recommends the application of HCNS gas to sections mounted in paraffin oil, and he claims good locali- zation. 26. Gomori, G.: Am. J. Path., 12:655, 1936. 27. Macallum, A. B.: J. Physiol., 22:92, 1897. 28. Dieterle, R. R.: Arch. Path., 10:740, 1930. 29. Miihlmann, M.: Virchows Arch. f. path Anat., 266:697, 1927. 30. Kockel, H.: Virchows Arch. f. path. Anat., 277:856, 1929. 31. Schmelzer, W.: Ztschr. f. wissensch. Mikr., 50:99, 1933. 40 Microscopic Histochemistry The 8-hydroxyquinoline^^ and the dinitroresorcinoP^ tech- niques ojBFer no advantages. Masked or occult iron.— A number of well-defined iron- containing compounds, such as hemoglobin, malaria pig- ment, formalin pigment, and possibly some other less well- known ones, do not show any reaction with the Prussian blue method; however, they can be made positive by destroying the organic part of the molecule ("unmasking" the iron). Ammonium sulfide and acids are sometimes referred to as unmasking agents, but wrongly so. They do not liberate demonstrable iron from any of the substances mentioned. A good demasking agent must spare tissue architecture and leave the iron at its original site. No strongly acidic sub- stance can satisfy the latter condition. The only reagents which have been used successfully are free chlorine, bro- mine,^^ and hydrogen peroxide.^^ Chlorine is applied either in the form of a gas ( Okamoto ) ^^ or in a nonaqueous solvent (Kockel).^^ These methods are cumbersome and not very reliable. Even if they do work, they will transform iron into highly hygroscopic and diffusible ferric chloride, which can- not be localized with any accuracy. On the other hand, hy- drogen peroxide produces insoluble ferric oxide. Whether all or only some of the iron-containing biological substances are unmasked by it and to what extent remains to be deter- mined; hemoglobin gives a fairly intense Prussian blue re- action after treatment with hydrogen peroxide. Method Apply a few drops of a 30 per cent solution of hydrogen peroxide (e.g., Superoxol), alkalized with some dilute am- monia or sodium carbonate, to the section and leave it on 32. Thomas, J. A., and LavoUey, J.: Bull, d'histol. appliq. a la physiol., 12:400, 1935. 33. Humphrey, A. A.: Arch. Path., 20:256, 1935. 34. Klein, G.: Praktikum der Histochemie (Berlin: J. Springer, 1929). 35. Brown, W. H.: J. Exper. Med., 13:477, 1911. 36. Okamoto, K.: Acta Scholae Med. Univ. Kioto, 20:413, 1937. Inorganic Substances 41 for about 30 minutes. Wash the shde under the tap and sub- ject it to the Prussian blue method. Pohcard^^ finds that the most sensitive method to demon- strate iron is microincineration. Copper.— According to Okamoto and co- workers, ^^ certain invertebrate tissues contain some kind of cupric analogue of hemosiderin. The form in which copper occurs in the tissues of higher species is not known, but it is almost invariably masked and not demonstrable directly, except when cupric salts have been administered parenterally. Okamoto and Utamura^^ suggest the use of two reagents for the demonstration of copper: rubeanic acid and p-di- methylamino-benzylidenerhodanine. The former was found in model experiments to be highly specific and sensitive, although, contrary to the findings of Okamoto,*^ positive re- sults were never obtained in human tissues. Perhaps an un- masking pretreatment with strong hydrogen peroxide would help to reveal copper even in mammalian organs. Method Fixation in absolute alcohol or in formalin. Incubate sections at 37° C. for 12-24 hours in a mixture consisting of 50 ml. of 10 per cent Na acetate and 1-3 ml. of a 0.1 per cent solution of rubeanic acid ( dithiooxamide ) in alcohol. Copper greenish black. Theoretically, copper should be demonstrated also by acid- ified potassium ferrocyanide, in the form of reddish-brown cupriferrocyanide. However, the color would be overlaid beyond recognition by the shade of Prussian blue, since the amount of iron present in the tissues far exceeds that of copper. 37. Policard, A.: Bull, d'histol. appHq. a la physiol., 11:216, 1934. 38. Okamoto, K., Utamura, M., and Mikami, G.: Acta Scholae Med. Univ. Kioto, 22:335, 1938-39. 39. Okamoto, K., and Utamura, M.: Acta Scholae Med. Univ. Kioto, 20:573, 1937-38. 40. Okamoto, K., Utamura, M., and Mikami, G.: Acta Scholae Med. Univ, Kioto, 22:348, 1938-39. 42 Microscopic Histochemistry Zinc— Fair amounts of this element are known to occur in the pancreatic islets ( as a component of insulin ) and in the red cells (as a component of carbonic anhydrase). Mendel and Bradley"^^ have described a method based on the insolubility of zinc nitroprusside. It has been applied only to tissues of mollusks, which contain very large amounts of zinc. Method Fixation not specified in original article; presumably alco- hol or formalin would do. Treat paraffin sections for 15 minutes at 50° C. with a 10 per cent solution of Na nitro- prusside. Wash under the tap for 15 minutes. Mount section in water and let a drop of dilute potassium sulfide solution run under the cover slip. An intense purple shade indicates the presence of zinc. Okamoto^" suggests diphenylthiocarbazide as a reagent. Method Alcohol fixation. Prepare a saturated solution of diphenyl- thiocarbazide in 60 per cent alcohol. Use solution when 2-4 days old. Add 1-3 ml. of this solution to 50 ml. of a borate buflPer of pH 8.4-9.2. Stain sections in the mixture for 2-3 hours at room temperature. Rinse briefly in water. Counter- stain lightly with hematoxylin; do not differentiate. Mount in glycerol-jelly. Zn purplish red. Mercury gives a more bluish-purple shade. This method has yielded good results in the rabbit pan- creas ;*^'^* equivocal or poor ones in the pancreases of other species.** Manganese.— GYund\a.nd and Bulliard*^ suggest the use of 41. Mendel, L. B., and Bradley, H. C: Am. J. Physiol., 14:313, 1905. 42. Okamoto, K.: Tr. Soc. Path. Jap., 32:99, 1942; and Okamoto, K., and Hashimoto, M.: Taishitzu Gaku Zasshi, 13:83, 1944. 43. Kadota, J.: J. Lab. & Clin. Med., 35:568, 1950. 44. Gomori, G.: Unpublished. 45. Grundland, I., and BuUiard, H.: Compt. rend. Soc. de biol., 142:201,^ 1948. Inorganic Substances 43 8-hydroxyquinoline for the demonstration of manganese. In an oxidizing medium it is supposed to stain manganese brown-black. The method has not been tested on animal material, but it appears unhkely that it will prove sensitive and selective enough, especially in the presence of much larger amounts of iron. Metals Occurring under Experimental or Pathological Conditions thallium Method of Barhaglia^^ Fix tissues in 95 per cent alcohol containing 2-5 per cent of iodine and 5-10 per cent of potassium iodide. Thallous iodide precipitates in the form of yellow crystals. This is an untested method. BERYLLIUM Method of Denz'^'^ Fix in formalin or formalin-alcohol. Mix equal volumes of a fresh 0.5 per cent solution of Naphthochrome green B and of phosphate buffer pH 5.0. Stain sections in this mixture for 30 minutes at 37° C. Wash in water, differentiate in absolute alcohol. Wash once more; counterstain with acridine red. Dehydrate and mount. Be stains apple green. Iron and alu- minum give somewhat different green shades. This is an un- tested method. Lead— The two methods described for the demonstration of this element are based on the insolubility of its yellow chromate and its brown-black sulfide, respectively. The chromate method"^^ is for use with tissues not re- quiring decalcification. The reagent (potassium dichromate) 46. Barbaglia, V.: Studi Sassari, 8:253, 1930. 47. Denz, F. A.: Quart. J. Micr. Sc, 90:317, 1949. 48. Cretin, A.: Compt. rend. A. anat., 16:241, 1929; Frankenberger, Z.: Compt. rend. A. anat., 16:241, 1921. 44 Microscopic Histochemistry can be incorporated in the fixative (in the form of Regaud's mixture) or appHed to sections of formaUn- or alcohol-fixed material. Lead is demonstrated in the form of yellow, rather opaque crystals, soluble in dilute nitric acid and blackened by ammonium sulfide. This method cannot be used with bones which require de- calcification, since lead chromate is soluble in acids. In bones, lead can be demonstrated by transforming it into brown- black PbS ( Sieber ) ,^^ which is resistant to the action of mod- erately strong acids. The pieces are fixed in formalin satu- rated with H2S, washed, and decalcified with 5-10 per cent trichloroacetic or formic acid. Lead will show up in a dark- brown or black shade, depending on the amount present. However, a number of heavy metals form very similar sul- fides, indistinguishable from that of lead without further identifying tests. Theoretically, it should be possible to localize lead by de- calcifying the pieces of bone with 5-10 per cent sulfuric acid containing about the same concentration of sodium or am- monium sulfate. Under such conditions lead would be trans- formed into white, insoluble lead sulfate, which could then be identified in the form of the sulfide. Other heavy metals would not interfere. Mercury.— The forms in which mercury may occur in the tissues are poorly known. Judging from the reactions de- scribed for its identification, it appears to be rather loosely bound and reactive. Brandino^^ suggests the use of diphenylcarbazide as a re- agent but does not give any detailed instructions. Okamoto^^ uses diphenylthiocarbazide or diphenylthiocarbazone (dithi- 49. Sieber, E.: Arch. f. exper. Path. u. Pharmakol., 181:273, 1939. 50. Brandino, G.: Studi Sassari, 5:85, 1927. 51. Okamoto, K., Seno, M., and Okumura, T.: Taishitzu Gaku Zasshi, 13:89, 1944. Inorganic Substances 45 zone; see method under "Zinc"). Hand and associates'^ re- duce mercurous mercury with 10 per cent thioglycoHc acid, and both the mercurous and mercuric form with stannous chloride. The metalHc mercury which appears is in the form of dark droplets. All the techniques mentioned require thorough testing for specificity. Bismuth, silver, and gold.— Alter being introduced into the animal organism, these metals are converted in a short time into dark-brownish or blackish, granular, rather unreactive compounds the chemical nature of which is not well under- stood (proteinate, sulfide, or reduced metal?). The identification of heavy metals in the tissues is a prob- lem of analytical chemistry rather than of histochemistry. In histochemical studies the identity of the metal to be dem- onstrated is usually known in advance (having been intro- duced artificially ) . Therefore, the task is not so much to dif- ferentiate between two or more metals as to estabhsh the metalhc nature of the pigment granules and to distinguish them from other colored granular substances such as soot, melanin, formalin pigment, pigment of wear and tear, etc. In most cases purely morphological criteria (type of cells in which the granules are seen) will do more than histo- chemical tests to decide this question. The following relatively simple tests are recommended: 1. Apply 30 per cent hydrogen peroxide to the section for 5 minutes. Silver and gold remain unchanged (the latter will actually darken; method of Elftman);'^ bismuth'* and most other pigments will be bleached. For the identification of bismuth, rinse the bleached shde thoroughly and proceed according to one of the following two methods: 52. Hand, W. C, Edwards, B. B., and Caley, E. R.: J. Lab. & Clin. Med., 28:1835, 1943. 53. Elftman, H., and Elftman, A. G.: Stain Technol., 20:59, 1945; Elft- man, H., Elftman, A. G., and Zwemer, R. L.: Anat. Rec, 96:341, 1946. 54. Wachstein, M., and Zak, F. G.: Am. J. Path., 22:603, 1946. 46 Microscopic Histochemistry a) Method of Komaya^^ Make up two solutions: solution A, 0.5 g. of quinine sulfate or hydrochloride in 25 ml. of 2-3 per cent H2SO4; solution B, a 5 per cent solution of potassium iodide. For use, mix equal parts of solutions A and B and pour the mixture over the slide. After 5-10 minutes, decant the reagent, counterstain with a dilute (about 0.05 per cent) solution of light green to which a few drops of the reagent are added. Blot section, dehydrate, and mount it in balsam. Bismuth orange-red. h ) Method of Castel^^ Same as that of Komaya, except that quinine is replaced by brucine. Results practically the same with both methods. 2. Treat section with a dilute solution of bromine (dis- solve a few crystals each of KBr and of KBrOs in a Coplin jarful of distilled water and add a few drops of concentrated HCl). Granules of gold and bismuth are completely dis- solved in 10-15 minutes; silver is bleached (formation of bromide). Wash the slide; pour over it any photographic developer; silver bromide will be reduced to black metallic silver. B. NONMETALLIC ELEMENTS Chlorine.— The chloride ion can be localized in frozen- dried material only.^^ The reagent is a dilute solution of AgNOs (0.1-2 per cent; concentration not important), pref- erably in 70-95 per cent alcohol. C1~,C03", and HPO^^will all be precipitated by it; however, dilute ( about 0.5 per cent ) nitric acid will remove phosphate and carbonate, and only chloride will remain undissolved. This can then be reduced to metallic silver, either by exposure to direct sunlight or by photographic developers ( see under "Calcium" ) . 55. Komaya, G.: Arch. f. Dermat. u. Syph., 149:277, 1925. 56. Castel, P.: Bull, d'histol. appliq. a la physiol., 13:290, 1936. 57. Gersh, I.: Proc. Soc. Exper. Biol. & Med., 38:70, 1938. Inorganic Substances 47 Active chlorine ( CI2 ) can be demonstrated with the meth- od of Ferguson and Silver. ^^ Method Fix small tissue fragments by dipping them in boiling for- malin for a few seconds. Cut frozen sections 25 ^ thick. Im- merse sections in the following reagent for 15-25 seconds: dissolve 0.1 g. of o-tolidine and 1 g. of citric acid in 20 ml. of water; fill up to 100 ml. Mount in glycerol. Greenish color indicates small amounts; yellow and orange, larger amounts of free CI2. Iodine —There are no good methods available either for the iodide ion (see Gersh and Stieglitz)^^ or for occult iodine ( e.g., in thyroxine ) . Phosphorus.— The demonstration of the phosphate ion was described in the section on calcium. It should be added here that, whereas the usual heavy-metal techniques do not differ- entiate between phosphate and carbonate, there is a theo- retical possibility of distinguishing between the two by the use of uranyl salts. These will give a precipitate with phos- phates but not with carbonates (uranyl carbonate being fair- ly soluble). Insoluble uranyl phosphate is convertible into reddish-brown uranyl ferrocyanide. Lihenfeld and Monti^^ have proposed the demonstration of organic phosphates by hydrolyzing them with nitric acid in the presence of ammonium molybdate and then reducing the phosphomolybdate precipitate in a second step to molyb- denum blue. This method is worthless histochemically. Even if inorganic phosphate were the first soluble phosphorus-con- taining compound liberated (and this is not in the least likely), the method could not possibly localize it. First of all, ammonium phosphomolybdate is not insoluble enough; 58. Ferguson, R. L., and Silver, S. D.: Am. J. Clin. Patli., 17:35, 1947. 59. Gersh, I., and Stieglitz, E. J.: Anat. Rec, 56:185, 1933. 60. Lilienfeld, L., and Monti, A.: Ztschr. f. wissensch. Mikr., 9:332, 1892. 48 Microscopic Histochemistry second, it does not precipitate fast enough; and, third, mo- lybdenum blue is very soluble and difiFusible and has a great aflSnity for protein substances.®^ Therefore, the best one could expect of this method would be a diffuse blue coloration around areas of very high concentration of organic phos- phate without any sharpness of localization. The same criti- cism applies to all other methods based on the molybdenum blue principle (Angeli,®^ Serra,®^ and Okamoto®*). On the other hand, it might well be possible to develop a histochemical adaptation of Mandel and Neuberg's idea®^ (destroying organic phosphates with strong H2O2). Sulfur.— MsLCsllum^^ demonstrates sulfates by treating frozen sections of fresh tissue with a dilute solution of lead acetate, washing them with dilute nitric acid to remove all lead precipitates except acid-insoluble sulfate, and convert- ing the latter with ammonium sulfide to brown-black PbS. Because of the great mobility of sulfate ions, this method should be applied to frozen-dried tissues. According to Klein,®^ a number of organic compounds of sulfur are converted to sulfate by strong H2O2. There is a possibility that a histochemical method could be developed on this basis. Arsenic— The methods proposed for the demonstration of inorganic arsenic are based either on the formation of As2S3®^ 61. Bensley, R. R.: Biol. BuU., 10:49, 1908. 62. Angeli, B.: Riv. di biol., 10:702, 1928. 63. Serra, J. A., and Queiroz Lopes, A.: Portugal. Acta biol., 1:111, 1945. 64. Okamoto, K., Seno, M., and Kato, A.: Taishitzu Gaku Zasshi, 13:97, 1944. 65. Mandel, J. A., and Neuberg, C: Biochem. Ztsclir., 71:196, 1915. 66. Macallum, A. B.: Die Methoden der biologischen Mikrochemie, in Abderhalden's Handb. d. biol. Arbeitsmeth,, V-2:1099 (Berlin and Vienna: Urban & Schwarzenberg, 1912). 67. Klein, G.: Oesterreich. bot. Ztschr., 76:15, 1927. 68. Brunauer, S. R.: Arch. f. Dermat. u. Syph., 129:186, 1921; Osborne, E. D.: Arch. Dermat., 12:773, 1925; Memmesheimer, A. M.: Dermat. Ztschr., 54:4, 1928. Inorganic Substances 49 by H2S or on the precipitation of arsenites and arsenates by copper sulfate in the form of the corresponding insoluble cupric salts. ^^ These methods lack all specificity. Tannenholz and Muir^^ have shown that the yellow precipitate obtained after treatment of tissues with H2S is neither AS2S3 nor any other compound of arsenic; as far as the copper method is concerned, it would give positive reactions also with fatty acids, phosphates, carbonates, and many proteins. 69. Castel, P.: Bull, d'histol. appliq. a la physiol., 13:106, 1936. 70. Tannenholz, H., and Muir, K. B.: Arch. Path., 15:789, 1933. CHAPTER VII ORGANIC SUBSTANCES A. SACCHARIDES Sugars in their free form cannot be localized histochemically, not even in frozen-dried material, because all known reac- tions for them are far too slow. Very recently, Okamoto, Ka- dota, and Aoyama^ have suggested a new method based on the insolubility of Ba-glucose and Ba-lactose in alcohol. Method Fix very thin slices of fresh tissue in methyl alcohol satu- rated with barium hydroxide for about 24 hours at icebox temperature. Dehydrate with two or three changes of abso- lute alcohol, carry tissue through two changes of chloroform, and embed in paraffin. Mount paraffin sections directly on the slide; use a minimum of egg-white-glycerol. Deparaffin- ize section with chloroform, rinse in absolute alcohol, and transfer to a 1 per cent solution of AgNOs in 80-90 per cent alcohol; expose jar to direct sunlight. The insoluble precipi- tate of Ba-sugar will be transformed into the corresponding Ag compound and the latter reduced to metallic silver. Rinse slide in distilled water, remove unreduced silver vvdth a dilute solution of Na thiosulfate; counters tain and mount. A slide rinsed thoroughly in distilled water before the silver bath can serve as a control; only the difference between the two slides is due to the presence of glucose or lactose. This is an untested method. It sounds promising, although the accuracy of localization is probably only approximate. Phosphates are very likely to give false positive reactions. The method should be tried also on frozen-dried material. 1. Okamoto, K., Kadota, I., and Aoyama, Z.: Taishitzu Gaku Zasshi, 14:35, 1948. 50 Organic Substances 51 Carbohydrates occur in four types of substances in a form insoluble enough to be demonstrated even by relatively slow reactions. The chemical nature and physiological role of these substances vary widely; the only reason for their being dealt with in one section is the fact that they all con- tain sugar and that they can be demonstrated by the re- actions of the sugar moiety. For the sake of brevity, they will be called "saccharides." Some of these substances have additional identifying reactions, due to other chemical or physicochemical features of their molecules. Classification of the Saccharides 1. Simple polysaccharides are built up from sugar mole- cules only. In animals the building stones are glucose ( glyco- gen) or galactose ( galactogen ) ; in plants they may be glu- cose (cellulose) or pentoses. Substances belonging in this group are either insoluble or soluble in a colloidal form only. Polymer linkages holding the individual molecules together go from Ci to C4 of the next molecule, although in some cases the linkages may connect to C3. This last type of compounds shows a histochemical behavior different from that of the other members of the group. 2. Mucoid substances are characterized by their content of aminosugar (glucosamine, acetylglucosamine, or acetyl- galactosamine) as the most typical component. K. Meyer^ divides this group into three classes. A. Mucopolysaccharides (protein-free) a) Neutral, containing no acid groups (example: chitin, composed of acetylglucosamine only) b) Acid 1. Simple; acid component, uronic acid (example: hyaluronic acid, composed of acetylglucosamine and glucuronic acid) 2. Complex; acid component, uronic acid and sulfmic acid or phosphoric acid (examples: corneal mucoid, hog gastric mucin, heparin, all composed of glucosamine or acetylglu- cosamine, some uronic acid, and sulfuric acid; chondroitin- sulfuric acid, composed of acetylgalactosamine, glucuronic 2. Meyer, K.: Adv. Protein Chem., 2:249, 1945. 52 Microscopic Histochemistry acid, and sulfuric acid; certain bacterial acid polysaccha- rides containing phosphoric acid) B. Mucoproteids, containing more than 4 per cent aminosugar a) Soluble neutral (gonadotrophins, etc.) b) Insoluble neutral (ovomucin) c) Acid (submaxillary mucoid) C. Glycoproteins, containing less than 4 per cent aminosugar (oval- bumin, serum albumin) 3. Glycolipids^ ( cerebrosides ) are complex substances which on hydrolysis yield a nitrogenous base ( sphingosine ) , a long-chained fatty acid, and a sugar ( usually galactose but in some cases glucose). The sugar is linked to the sphingo- sine molecule glycosidically, and the fatty acid is attached to the amino group of sphingosine. These substances are in- soluble in water, ether, and petroleum ether but are soluble in pyridine and hot alcohol. Another carbohydrate-containing lipid substance is liposi- tol, which contains inositol, galactose, fatty acids, phosphoric acid, and ethanolamine. 4. Nucleic acids are high polymer molecules, the indi- vidual building blocks of which are the nucleotides. The nucleotides are phosphoric esters of nucleosides, the latter being N-glucosides of purine and pyrimidine bases. It has been known for over fifty years that there are two different kinds of nucleic acids. One of them, thymonucleic acid, is the typical component of cell nuclei; its sugar moiety is desoxyribose (hence desoxyribose nucleic acid, DNA). The other one, formerly referred to as yeast nucleic acid, occurs in cytoplasmic structures; its sugar component is ribose (hence ribose nucleic acid, RNA). Both nucleic acids are bound to proteins in an insoluble form. General Principles of the Histochemical Demonstration of Saccharides Aldehyde reactions are used to demonstrate sugars histo- chemically. Besides carbohydrate substances, the only other 3. Everett, M. R.: Medical biochemistry (2d ed.; New York and London: Paul B. Hoeber, 1946). Organic Substances 53 known potential aldehydes occurring in the animal body are certain lipid substances and the unidentified aldehyde of elastic tissue. Whenever an aldehyde reaction is obtained under specific conditions in a tissue defatted with lipid sol- vents, it is attributed to the presence of carbohydrate. Aldehyde reactions are not given directly by any of the carbohydrate substances occurring in the tissues. The alde- hyde groups have to be "liberated" first by certain chemical agents, oxidative or hydrolytic. In the case of oxidants, "liber- ation" is a deceptive term to describe their action. True, all aldoses have a potential aldehyde group, masked by the pyranose or furanose ring formation. However, it is not this aldehyde group which is "set free" or "revealed" oxidatively but rather entirely new aldehyde groups which are created in the middle of the carbon chain of the sugar molecule. The action of one of the oxidants, periodic acid, is fairly well understood since the studies of Malaprade,^ Jackson and Hudson,^ Nicolet and Shinn,^ Hotchkiss,'^ and others. It con- sists of the breaking of the carbon chain through a glycoHc or HO-C-C-NH2 group and in the oxidizing of the broken ends to aldehyde groups according to the following schemes : H H H H R— C— C— Ri + HIO4 = R— C=0 + 0=C— Ri + HIO3 + H2O or O O H H H H H H R— C— C— Ri + HI04= R— C=0+0=C— Ri+ (NHJIOa O N H H2 4. Malaprade, L.: Bull. soc. chim. France, 43:683, 1928; Malaprade, L.: ibid., ser. 5, 1:833, 1934. 5. Jackson, E. L., and Hudson, C. S.: J. Am. Chem. Soc, 60:989, 1938. 6. Nicolet, B. H., and Shinn, L. A.: J. Am. Chem. Soc, 61:615, 1939. 7. Hotclikiss, R. D.: Arch. Biochem., 16:131, 1948. 54 Microscopic Histochemistry In the case of polysaccharides, treatment with periodic acid will not cause depolymerization of the molecule because the carbohydrate units remain connected by C1-C4 linkages, as indicated in the following scheme: HC: HCOH HOCH O HC HC- HC: HCOH HOCH O 0\ HC HC- HC: HC=0 > HC=0 O 0\ \ HC HC- HC: HC=0 HC=0 O 0\ HC HC- 0\ '^% CH2OH CHsOHi CH2OH W ' CH2OH The hydroxyl and/or amino groups must be free to be attacked by periodate; if they are tied down by substitution or by any kind of linkage, the compound will not be touched. The scheme is very general and applies to all sorts of com- pounds possessing vicinal -OH or -OH and -NH2 groups, whether or not they are of carbohydrate nature. Some of the noncarbohydrate substances oxidized to aldehydes are the amino acids serine, threonine, and hydroxylysine. The first two would not be attacked while incorporated within pep- tide chains unless they occupied a terminal position and were thus engaged through their carboxyl groups only. Some car- bohydrates are not attacked because they lack the vicinal -OH groups ( desoxyribof uranose ) ; in other cases the indi- vidual sugar molecules do have such groups, but they are tied down in polymeric linkages; (ribose nucleic acids; C3- Hnked polymers such as agar^ or heparin).^ But even sugars which seem to possess all the necessary prerequisites for 8. Jones, W. G. M., and Peat, S.: J. Chem. Soc., p. 225, 1942. 9. Meyer, K.: Ann. New York Acad. Sc, 52:943, 1950; Meyer, K., and Odier, M.: Experientia, 2:311, 1946. Organic Substances 55 being oxidized by periodate may prove exceptionally resist- ant^*^ (cellobiose, some methylglycosides ) ; the reasons for this irregular behavior are not clear. The amount of aldehyde formed from C4-linked polysac- charides does not depend on the degree of polymerism, be- cause of the availability of two vicinal OH groups (C2-C3) in every one of the carbohydrate units. On the other hand, in Cs-linked polysaccharides, only the temiinal units can be oxidized. For this reason, the relative yield of aldehyde w^ill increase with decreasing chain length. The mechanism of other oxidants has not been studied in sufficient detail, but it seems to be essentially similar to that of periodate, although there are minor quantitative or quali- tative differences. Periodate appears to be the most specific for glycolic linkages proper; the aldehyde groups are not further oxidized, even on prolonged exposure to the oxidant. Chromic acid, another popular oxidant, is less energetic as far as attacking all glycolic linkages is concerned; its action is limited almost exclusively to glycogen and mucin; glyco- proteins are quite resistant to its action. The aldehyde groups are destroyed after prolonged exposure and cannot be dem- onstrated after a certain time. Acidified permanganate, as used by Casella,^^ combines the properties of periodate and chromic acid in that it seems to attack all glycol linkages but, on prolonged action, destroys the aldehydes formed. In the case of nucleic acids, aldehydes cannot be produced by oxidative procedures. As mentioned before, desoxyribo- furanose, the carbohydrate component of DNA, does not possess a glycolic group. In ribofuranose, the carbohydrate moiety of RNA, phosphoric acid residues in position 3 block the reaction. If the latter are removed by mild hydrolysis, the depolymerized nucleic acid becomes diffusible and will be lost from the section. 10. Jeanloz, R.: Science, 111:289, 1950. 11. Casella, C: Anat. Anz., 93:289, 1942. 56 Microscopic Histochemistry Feulgen and Rossenbeck^^ in 1924 discovered that thymo- nucleic acid yields on hydrolysis with dilute mineral acids a substance which gives an intense aldehyde reaction, repro- ducible both in the test tube and in microscopic sections ("nucleal reaction"). The mechanism of this reaction has been considered a mystery until very recently. According to the latest reports of Overend and Stacey,^^ a large percentage of desoxyribose and a few other carbohydrates exist naturally in a noncyclic form which gives direct aldehyde tests (due to a terminal -CH=0 group). This feature distinguishes them sharply from the more common hexoses and pentoses which have no overt aldehydic character. By controlled acid hydrolysis it is possible to break just a sufficient number of glycosidic linkages to expose enough aldehyde groups for visualization of the reaction and at the same time retain much of the polymerism and insolubility of the acid on which good histochemical localization depends. Ribonucleic acid gives no reaction under such conditions. The free aldehyde groups can be demonstrated by a num- ber of specific aldehyde reagents. Theoretically, any alde- hyde reaction used in organic analysis could be employed; in practice, however, only a few have the desirable features of a histochemical test. The most important reagent for aldehydes is Schiff's re- agent, fuchsinsulfurous acid. The mechanism of the reaction has been clarified by Wie- land and Scheuing.^* In an acid solution, fuchsin is trans- formed by an excess of SO2 into a colorless N-sulfinic acid which forms highly colored addition complexes with alde- hydes. The addition complexes, of a purple shade different from that of fuchsin itself, have an excellent staining power and, being usually more insoluble than the corresponding 12. Feulgen, R., and Rossenbeck, H.: Ztschr. f. physiol. Chem., 135:203, 1924. 13. Overend, W. G., and Stacey, M.: Nature, 163:538, 1949. 14. Wieland, H., and Scheuing, G.: Ber. deutsch. chem. Gesellsch., B, 54:2527, 1921. 15 Organic Substances 57 aldehyde, cling very tenaciously to the structures stained. Schiff's reagent can be recolorized also by a different mecha- nism and with different end-products. The reagent is stable only in the presence of an excess of SO2 and at a high acid- ity. Anything that removes or oxidizes SO2 or reduces acidity will break the linkage between fuchsin and sulfurous acid and restore the original dye. This type of recolorization, not likely to take place under the conditions of a correctly per formed histochemical test, is called a "pseudo-reaction. It should not be confused with the genuine one, which is practically specific for aldehydes. Lison lists a few nonalde- hydic substances which may give colorations very similar to those occuning with aldehydes; however, they are almost all nonphysiological substances, not found in animal tissues. The red shade of the pseudo-reaction is readily extracted by acids and alcohol ( except in the case of acid-fast structures ) , while the aldehyde-regenerated dye is extremely resistant to both agents. It must be mentioned here that not only Schiff's reagent but also undecolorized fuchsin is able to combine with alde- hydes and yield purplish dyes which have staining properties different from those of fuchsin itself. Such dyes are used in Fite's modification^^ of the Ziehl-Neelsen stain, for the stain- ing of nuclei (DeLamater)^^ and in the aldehyde-fuchsin technique^^ for elastic fibers. In periodate-treated tissues a fairly good localized staining of the aldehydes may be ob- tained by fuchsin proper (Arzac);^^ however, the use of Schiff's reagent is always preferable. Another good analytical reagent for aldehydes is an alka- line solution of silver nitrate, first utilized for histochemical 15. Lison, L.: Bull, d'histol. appliq. a la physiol., 9:177, 1932. 16. Fite, G. L.: J. Lab. & Clin. Med., 25:743, 1940. 17. DeLamater, E. D.: Stain Technol., 23:161, 1948. 18. Gomori, G.: Am. J. Clin. Path., 20:665, 1950. 19. Arzac, J. P.: Analecta Med., 9:15, 1948; Arzac, J. P.: Stain Technol., 25:187, 1950; Arzac, J. P.: J. Nat. Cancer Inst, 10:1341, 1950. 58 Microscopic Histochemistry purposes by BignardP^ and Clara^^ and later described inde- pendently also by Gomori,^^ who was not then familiar with Bignardi's work. Two more silver techniques for glycogen and mucin were published by Mitchell and Wislocki^^ and by Pritchard;^^ however, these latter are silver-impregnation methods rather than histochemical tests. At this point the difference between the argentaffin reac- tion and simple impregnation by silver ( argyrophilia ) must be pointed out. Lison^^ has emphasized that some substances ( ascorbic acid, polyphenols, aldehydes, uric acid, etc. ) have the ability to reduce silver solutions under specific condi- tions. This is the "argentaffin reaction" which may be used for the histochemical characterization of reducing substances. It should be distinguished sharply from impregnation by silver such as is obtained in various techniques for nerve and reticulum fibers, etc. In such techniques the silver is not re- duced by substances contained in the tissue itself but by extraneous reducers (formalin, hydroquinone, etc.). Some of these methods are remarkably selective for various mor- phological structures but cannot be called "specific" in the chemical sense of the word. The silver technique for aldehydes can be used for glyco- gen and mucin-like substances only, not for DNA. The reason for this is not clear; it seems that acid hydrolysis makes DNA somewhat alkali-soluble, and it will diffuse out of the section before reduction of the silver can take place. PREPARA.TION OF THE ReAGENTS 1. Periodic acid— A 0.25-0.5 per cent solution in distilled water will serve all purposes. The use of alcoholic or buffered 20. Bignardi, C: Atti Soc. nat. e mat. Modena, 70:97, 1939; Bignardi, C: Boll. Soc. med.-chir. Pavia, 54:799, 1940; Bignardi, C: Atti Soc. ital. sc. nat, 79:23, 1940. 21. Clara, M.: Ztschr. f. mikr.-anat. Forsch., 47:183, 1940. 22. Gomori, G.: Am. J. Clin. Path., 10:177, 1946. 23. Mitchell, A. J., and Wislocki, G. B.: Anat. Rec, 90:261, 1944. 24. Pritchard, J. J.: J. Anat, 83:30, 1949. 25. Lison. L.: Histochimie animale (Paris: Gauthier-Villars, 1936). Organic Substances 59 aqueous solutions oflFers on obvious advantages. If loss of water-soluble matter is feared, the slide should be protected with collodion. 2. Chromic acid.— A 4-5 per cent solution of chromium trioxide in distilled water is used. 3. N HC/.-Dilute 10 ml. of concentrated HCl (assay, 36- 39 per cent ) with 100 ml. of distilled water. 4. Schiff's reagent.— There are a number of formulas^^"^^ published which all give essentially identical results. SHght variations in the results are due to differences between vari- ous batches of dye and to the age of the solution rather than to minor departures from any of the formulas. The following procedure by Coleman,^^ recommended by Lillie, was found to be simple and reliable: Dissolve 1 g. of basic fuchsin (C.I. No. 677) in 200 ml. of boiling distilled water. Cool to about 50° C. Add 1 ml. of concentrated hydrochloric acid (assay, 36-39 per cent) and 2 g. of either sodium bisulfite (NaHSOs) or potassium meta- bisulfite (K2S2O5); shake the flask, stopper it tightly, and allow it to stand at room temperature for 24 hours. The dye will be decolorized to a straw yellow or light-brownish so- lution with or without some precipitate floating in it, depend- ing on the quality of the dye used. Add about 0.5 g. of ad- sorbent charcoal (e.g., Norit), shake, and filter the mixture. The filtrate must be crystal clear and colorless; should it have a hght pinkish tinge, add hydrochloric acid drop by drop until all trace of color is discharged. Keep reagent in the icebox. Some batches of reagent will remain perfectly usable for over a year; others, for reasons completely unknown, may de- teriorate in a matter of weeks. The source of the dye does not seem to make much difference; reagents of highly variable keeping properties may be prepared from the same commer- 26. Wermel, E.: Ztschr. f. Zellforsch. u. mikr. Anat., 5:400, 1927. 27. De Tomasi, J. A.: Stain Technol., 11:137, 1936. 28. Coleman, L. C: Stain Technol., 13:123, 1938. 29. Rafalko, J. S.: Stain Technol, 21:91, 1946; Lillie, R. D.: Stain Technol, 26:163, 1951. 60 Microscopic Histochemistry cial brand of fuchsin. There is nothing in the gross appear- ance of the reagent to indicate spoilage; this manifests itself only by unsatisfactory staining results. Entirely colorless, clear solutions may have lost all their staining power; grossly turbid or pinkish solutions, after filtration or the addition of some more bisulfite for decolorization, may work well. It is always advisable to keep a few slides, known to contain gly- cogen or mucin, handy and perform a test run before the main experiment. Wermel's solution, although more complicated to prepare than the simple Schiff's reagent, is distinctly more stable than the latter. 5. The silver sohition.— Either of the following two solu- tions can be used: A) Fontanas^^ silver solution.— To a 2-3 per cent solution of silver nitrate add, drop by drop and under continuous shaking, strong ammonia water until the initial brown-gray precipitate dissolves. To the clear mixture add some more silver solution, drop by drop and under continuous shaking. The turbidity resulting from the first few drops will disap- pear easily; continue to add silver solution until a minimal opalescence persists. The solution will keep in the refriger- ator for a few days. B) Methenamine-silver stock solution. ^^— Add 5 ml. of a 5 per cent AgNOs solution to 100 ml. of a 3 per cent methen- amine solution, shake until the initial heavy white precipi- tate disappears. This mixture will keep in the refrigerator for many months. Fontana's solution has a tendency to produce a fine dust- like precipitate all over the slide and a fairly intense back- ground staining. Methenamine-silver is practically free from these drawbacks; the pictures obtained with its use are very clear. 30. Fontana, A.: Dermat. Ztschr., 46:291, 1925-26. Organic Substances 61 Methods for saccharides I. SUBSTANCES OTHER THAN NUCLEIC ACIDS (GLYCOGEN, MUCIN, GLYCOPROTEIDS, ETC.) Fixation and embedding.— While for mucinoid substances almost any fixation will do, the correct fixative for glycogen has been the subject of numerous publications,^^' ^^~^^ and there is still no consensus. Most textbooks on microtechnique recommend alcohol or alcoholic mixtures on the principle that glycogen is insoluble in them while it is soluble in most aqueous fixatives. Theoretically, as applied to pure glycogen, this is a perfectly valid consideration; however, in tissues where glycogen is embedded in a complex mixture of pro- teins and lipids the situation is different. Any good protein precipitant will coat the glycogen particles with a protein membrane which is impermeable to the large molecules of glycogen, thus keeping them in situ. Lison^^ quotes the stud- ies of Pasteels and Leonard, who state that one of the best fixatives is Bouin's fluid. Lillie,^^ on the other hand, finds that Bouin's fluid is an unsuitable fixative for glycogen. The writer agrees with the French workers on the excellence of Bouin's fluid and with Lillie on the poor results obtained with fix- atives containing mercury salts. In summary, it may be said that any good histological fixative, with the exception of sub- limate-containing mixtures, can be used. The best are those which act fast and produce considerable hardening of the tissues ( Bouin's fluid, formalin-alcohol with or without acetic acid, etc. ) . However, the morphology of glycogen wfll vary somewhat with different methods of fixation. In general, alco- holic or acid-containing fixatives cause the aggregation of glycogen into fairly coarse droplets, while simple formalin or formalin-bichromate mixtures reveal a more unfform, fine- 31. Bensley, C. M.: Stain TechnoL, 14:47, 1939; Deane, H. W., Nesbett, F. B., and Hastings, A. B.: Proc. Soc. Exper. Biol. & Med., 63:401, 1946. 32. Lillie, R. D.: Bull. Intemat. A. M. Mus., 27:23, 1947; Vallance- Owen, J.: J. Path. & Bact., 60:325, 1948. 62 Microscopic Histochemistry ly granular distribution, resembling that seen in frozen-dried preparations.^- The temperature of fixation seems to be of considerable importance. Liver contains a powerful glyco- genolytic enzyme system, which, unless the tissue is chilled promptly, may cause a considerable loss of glycogen, espe- cially in the interior of thick blocks. This loss may be quite conspicuous in sections stained with Bauer's^^ or Best's^* methods, while it may not be noticeable at all if the Hotch- kiss-McManus"^' ^^ technique is used— an indication of a change in the constitution of the molecule rather than of actual disappearance. Although the glycogen found in other tissues is much more stable, it is, safer to perform fixation of all tissues in the refrigerator. Frozen sections, as a rule, cannot be used for the staining of glycogen. Celloidin- or parafiin-embedding are equally good. A) Methods specific for the group in general— These methods are based on aldehyde reactions after oxidative pretreatment. Choice of the oxidizing agent —The two oxidizers most often employed are chromic acid and periodic acid. Chromic acid is the agent of choice if a selective staining of glycogen and/or mucin is desired. Besides these two substances, starch, galactogen, cellulose, tunicin, chitin, and colloid of the thyroid follicles will also react more or less strongly after chromic acid oxidation. Glycoproteins of the connective tis- sue remain almost completely unstained, and for this reason the contrast between glycogen and /or mucin and the back- ground is quite sharp. It is a curious fact that Bauer^^ in his original paper on the chromic acid-Schi£F method explicitly states that mucin does not stain; all subsequent workers have found that it stains quite intensely. After periodate oxidation, all the substances previously mentioned will react, and, in 33. Bauer, H.: Ztschr. f. mikr.-anat. Forsch., 33:143, 1933. 34. Best, F.: Ztschr. f. wissensch. Mikr., 23:319, 1906. 35. McManus, J. F. A.: Nature, 158:202, 1946. Organic Substances 63 addition, a number of others, such a fibrin, hyaluronic acid, various undefined glycoproteids of connective tissue, poly- saccharides of bacteria, and fungi,^^ kerasin in Gaucher's dis- ease,^^ lipofuscin,^^ ceroid,^^ and an unidentified substance in the foamy cells of Whipple's disease. ^^ In general, the purple-red shade obtained after periodate oxidation is much more brilliant than that obtained after chromic acid. The permanganate method of Casella^^ is not standardized well enough to be recommended for routine use. Some mucoid substances which stain faintly, irregularly, or not at all by the aldehyde techniques are heparin ( mast cell granules; marked species differences), chondroitin-sulfuric acid (cartilage ground substance), and amyloid. Choice of the aldehyde reagent. —SchiS's reagent gives equally good results after any of the oxidants; silver solu- tions are less specific and less reliable after periodate and should be used only in combination with chromic acid. The deep black shade may be an advantage in microphotography. Methods Oxidation with chromic acid.— Tresit sections for 40-60 minutes with a 4-5 per cent solution of chromic acid. Wash under the tap for at least 5 minutes; the color of chromic acid must be removed completely. Oxidation with periodic acid.— Tresit sections for about 10 minutes with a 0.25-0.5 per cent solution of periodic acid; wash under the tap for 10 minutes. Hotchkiss^ originally rec- ommended a rinse in bisulfite after oxidation. The eflFect of this step will be a delayed and gradual development of color in the next step. This may be a desirable feature if the aim is a sharper differentiation between the more promptly re- 36. Kligman, A. M., and Mescon, H.: J. Bact., 60:415, 1950; Kligman, A. M., Mescon, H., and DeLamater, E. D.: Am. J. Clin. Path., 21:86, 1951. 37. Morrison, R. W., and Hack, M. H.: Am. J. Path., 25:597, 1949. 38. Lillie, R. D.: Anat. Rec, 108:239, 1950. 39. Lee, C. S.: J. Nat. Cancer Inst., 11:339, 1950. 40. Black-Schaffer, B.: Proc. Soc. Exper. Biol. & Med., 72:225, 1949. 64 Microscopic Histochemistry acting substances (glycogen and mucin) and carbohydrates of the connective tissue. Staining with Schiffs reagent —StSiin sections for about 10-15 minutes in SchiE's reagent (straight or diluted with an equal volume of water ) . Usually very little staining is seen during this step, especially after chromic acid oxidation. Rinse the reagent oflF; flood sections for 1 or 2 minutes with a 1-3 per cent solution of Na bisulfite. Wash under the tap for about 5-10 minutes. It is during this washing that the stain- ing of mucin, glycogen, etc., in an intense purplish-red shade becomes more apparent. Counterstain with hematoxylin (staining after chromic acid often rather poor). Dehydrate and mount. Staining with meihenamine-siher.—Rmse sHdes thoroughly in distilled water. Incubate them at a temperature of 37°- 50 °C. in a mixture consisting of 25 ml. each of methenamine stock solution and distilled water with a few ml. of an M/5 borate buffer of pH 8.3-9.2 added. The more alkaline the so- lution, the faster it will work but the more likely it is to pro- duce a brownish background, especially in the case of unduly long incubation. The reaction will show up in a yellowish- brown shade in about 1 hour. Inspect sections under the mi- croscope once every 30 or 60 minutes; as soon as glycogen, mucin, etc., appear in black or in a very dark purpHsh-brown shade while the background is still a very light tobacco brown, remove sections from the silver solution, rinse in dis- tilled water, and tone with a 0.1-0.2 per cent solution of gold chloride for about 10 minutes. This treatment usually com- pletely bleaches the background. Remove unreacted silver by a short rinse with dilute (1-3 per cent) Na thiosulfate. Wash and counterstain as desired (e.g., with hematoxylin and eosin). Dehydrate and mount. If the background is too dark, carefully differentiate the section (before gold toning) in a mixture containing about 0.2 per cent of ferric sulfate and 0.5 per cent of sulfuric acid Organic Substances 65 until the background is a pale yellow-brown. Overdifferentia- tion may weaken the reaction proper. Oxidation by chromic acid followed by SchiflF's reagent is Bauer's method;^^ oxidation by periodate followed by Schiff's reagent is the Hotchkiss-McManus method."^' ^^ Etcheverry and Mancini^^ have described a method for carbohydrates based on the fact that tungstic acid, on expo- sure to ultraviolet light, is reduced by carbohydrates to a blue oxide of unknown composition. The mechanism of the reaction is not clear. The authors believe that it involves the oxidation of glycols to aldehydes. The method is mentioned only as a curiosity; it cannot compete with other techniques which give a much better color contrast. Specificity of aldehyde reactions after oxidative procedures. —As mentioned before, Schiff's reagent applied to paraffin or celloidin sections following the oxidative procedures de- scribed is specific for vicinal OH or OH and NH2 groups, which occur only in carbohydrates and in a few amino acids. The same appHes to the silver reagent after chromic acid, since substances which would reduce it without previous treatment (uric acid, phenols, melanin and its precursors) are destroyed by oxidation. Two exceptions, however, must be made. First, under cer- tain conditions, both normal and pathological, unsaturated fatty acids of the tissues may undergo changes resulting in the formation of peculiar substances the exact nature of which is not clearly understood. They exhibit the character of aldehydes and/or peroxides; most, but not all, of them are removed by the procedures of dehydration and embedding. They may be fairly well preserved in paraffin sections after prolonged fixation in bichromate-containing mixtures and give a positive SchiflF reaction, to some extent even without 41. Etcheverry, M. A., and Mancini, R. E.: Rev. Soc. argent, de biol., 24:156, 1948. 66 Microscopic Histochemistry oxidative pretreatment (example: ceroid). Second, elastic fibers in certain locations (mainly in the inner elastic coat of small-caliber arteries) contain an unidentified nonlipid sub- stance which behaves as an aldehyde whether or not an oxi- dative pretreatment is employed. Differentiation between Schiff -positive substances —Al- though normally the localization and morphology of glycogen is quite different from that of mucin, the problem of distin- guishing between the two substances may come up once in a while. One of the oldest methods of differentiation is the saliva test. It is based on the fact that saliva contains a dia- static enzyme which will dissolve glycogen and starch but not mucin, amyloid and other related substances which may stain in a shade indistinguishable from that of glycogen. It may be appied to all staining methods, whether histochem- ical or not. Therefore, ff a substance stains without pretreat- ment but fails to do so after an exposure to saliva of about 30-60 minutes, it must be glycogen or starch. On the other hand, ff it persists in staining after the saliva test, it cannot be glycogen. At present, more appetizing and sanitary tests have replaced the time-honored saliva test. A number of highly active diastase (amylase) preparations are available which are specific and efficient. Being protein in nature, they will not diffuse through collodion; therefore, if they are to be used, sections should not be coated with collodion, and col- lodion should be removed from celloidin-embedded material. LiUie and Greco^^ recommend the use of a 1 per cent dilution of extract of malt, U.S.P., in a phosphate buffer of pH 6.8-7.4. A 1 per cent solution of malt or animal diastase (Nutritional Biochemicals) is even more effective. These enzymes will re- move glycogen and starch from sections not over 8 fjL thick within 15 minutes at room temperature. Prolonged exposure (24-48 hours) may remove all carbohydrate material. A number of other specific enzymes for the hydrolysis of various mucopolysaccharides have become available recently 42. LilHe, R. D., and Greco, J.: Stain TechnoL, 22:67, 1947. Organic Substances 67 43-45 (hyaluronidase, pectinase, polygalacturonidase, etc.). Hyaluronidase appears to be specific in abolishing some or all staining reactions (depending on the source of the en- zyme and on the length of exposure ) due to hyaluronic acid and to chondroitin;*^ the limits of specificity of the other enzymes remain to be established. B ) Methods for individual members of the group.— a) Glycogen and starch.— The oldest method for starch is the staining by iodine;*^ it is mentioned as early as 1825.^^ Glycogen will stain in a mahogany brown, native starch in dark blue. Iodine can be applied in the form of Lugol's solu- tion,^^ iodine vapor, or a solution of iodine in parafiin oil.^^ The iodine reaction is not absolutely specific; amyloid, some protein substances, and lecithin^^ will also stain. Galactogen does not stain.^^ The degree of contrast obtained is not too high, and the preparations are not stable. Although some workers still use it extensively, the iodine reaction can be said to have little more than a historical interest. Best's carmine. ^'^— This is a purely empirical but highly se- lective stain. Besides glycogen, it may stain mast cell gran- ules, mucin, and fibrin, but in so much lighter shades that confusion is unlikely. 43. Hale, C. W.: Nature, 157:802, 1946. 44. Grishman, E.: Bull. Intemat. A. M. Mus., 28:104, 1948. 45. Gersh, L, and Catchpole, H. R.: Am. J. Anat., 85:457, 1949; Stough- ton, R., and Wells, G.: J. Invest. Dermat., 14:37, 1950; McManus, J. F. A., and Saunders, J. C.: Science, 111:204, 1950; McManus, J. F. A.: Am. J. Path., 26:690, 1950; Bunting, H.: Ann. New York Acad. Sc, 52:977, 1950. 46. Mathews, M. B., Roseman, S., and Dorfman, A.: J. Biol. Chem., 188:327, 1951. 47. Mancini, R. E.: J. Nat. Cancer Inst., 10:1371, 1950; Gage, S. H.: Tr. Am. Micr. Soc, 28:203, 1906. 48. Caventou, J. B.: Ann. de chim. et de phys., 31:337, 1826. 49. Langhans, T.: Virchows Arch. f. path. Anat., 120:28, 1890. 50. Mancini, R. E., and Celani-Barry, R.: Rev. Soc. argent, de biol., 19:493, 1943; Mancini, R. E.: Medicina, 7:327, 1947; Mancini, R. E.: Anat. Rec, 101 : 149, 1948. 51. Romieu, M.: Compt. rend. Soc. de biol., 96:1230, 1927. 52. Grainger, J. N. R., and Shillitoe, A. J.: Stain Technol., 27:81, 1952, 68 Microscopic Histochemistry Method Use celloidin sections or protect paraffin sections by dip- ping them into dilute (0.5-1 per cent) collodion in alcohol- ether between the first and second alcohols. Prepare the following stock solution: Carmine 2 g. Potassium carbonate .... 1 g. Potassium chloride 5 g. Distilled water 60 ml. Simmer mixture gently for a few minutes; cool. Add 20 ml. of concentrated ammonia water. This solution will keep in the refrigerator for several months. For use, dilute 10 ml. with 15 ml. of concentrated ammonia water and 15. ml. of 95 per cent alcohol. This solution must be made fresh every time. Filter if not perfectly clear. Stain nuclei rather dark with hematoxylin; rinse slide. Stain for 5-10 minutes in the dilute mixture. Differentiate in 60 per cent alcohol to which a few drops of ammonia water are added. Dehydrate and mount. Glycogen brilliant red. b) Mucin.— Mayer s^^ mucicarmine stain— This is an empirical but specific stain for most types of mucus. Some varieties, how- ever, are not stained; species and organ differences are quite marked. Mucus staining with mucicarmine also exhibits metachromatic properties, whereas mucicarmine-negative mucus is not metachromatic (see under "Metachromasia"). Method {modification of SouthgateY^ Prepare the following stock solution: simmer gently a mix- ture of 1 g. of carmine, 0.5 g. of anhydrous aluminum chloride (or 0.9 g. of AICI3 -61120), 1 g. of aluminum hydroxide, and 100 ml. of 50 per cent alcohol until it turns into a deep ruby- red liquid. This usually takes a few minutes. Let stand for 24 hours and filter; keep filtrate in the icebox. 53. Mayer, P.: Mitt. zool. Stat. Neapel, 12:303, 1896. ^. Southgate, H. W.: J. Path. & Bact., 30:729, 1927. Organic Substances 69 Fix tissues in alcohol or formalin-alcohol. Other fixatives are also usable, but the background may be stained more or less intensely. Stain nuclei with hematoxylin. Rinse slide in water; stain in mucicarmine for about 15 minutes; rinse in water, dehydrate, and mount. Some kinds of mucus stain intensely red; others are much paler or may fail to stain altogether. If staining time is prolonged, hyaluronic acid and glycogen may become stained, although in a much paler shade. The Metachromatic Staining of Acid Polysaccharides The concept of metachromasia.— It was observed simulta- neously by Comil,^^ Jiirgens,^^ and HeschP^ in 1875 that methyl violet stains amyloid red and other tissue structures blue. This phenomenon— namely, the staining of certain tis- sue components in a color different from that of the dye so- lution itself— has been named "metachromasia" by Ehrlich.^^ According to his nomenclature, the shade of the dye solution is the orthochromatic shade; the different shade in which cer- tain tissue elements stain is the metachromatic shade; the substances which exhibit the property of staining in the metachromatic shade are the chromotropic ( or metachromat- ic) substances. It must be made clear that metachromasia applies only to the cases in which both shades are produced by a single dyestuff and not by a mixture of dyes. Some com- mercial dyes may contain substantial admixtures, as impuri- ties, of other related dyestuff s. The number of dyes exhibiting metachromatic properties is large. However, only a few of them have practical impor- tance. The list includes thionin, the azures, toluidin blue, cresyl violet, and methyl violet. Methyl green, used for the metachromatic staining of amyloid, owes its metachromatic properties to its content of methyl violet. The most typical 55. Comil, v.: Compt. rend. Acad. d. sc, 80:1288, 1875. 56. Jiirgens, R.: Virchows Arch. f. path. Anat., 65:189, 1875. 57. Heschl, R.: Wien. med. Wchnschr., 25:712, 1875. 58. Ehrlich, P.; Arch. f. mikr. Anat, 13:263, 1877. 70 Microscopic Histochemistry common property of these dyes is that their shade depends on the concentration of the solution (although this feature has not been investigated in the case of methyl violet ) . Dilute solutions are bluish, and, as the concentration is raised, the shade shifts progressively toward red. The absorption spectra of these dyes show two ( or three ) maxima. In the case of to- luidin blue, for instance, one of the peaks ( a ) is around 630 m^, the second one ( /5 ) around 590 m^u,, and the third one (y) around 560-70 m^u,. In dilute solutions the a peak is the tallest; the other two are barely noticeable. With increasing dye concentration the /3 peak will become more and more prominent and actually surpass the a peak. The y peak is observed only at rather high concentrations. Acidification, high temperature, and the addition of alcohol tend to sup- press the y8 peak; alkalinization, low temperature, and the addition of chromotropic substances have the opposite effect. The physicochemical nature of metachromasia is not com- pletely elucidated. According to Lison, ^^' ^^ it is due to a tautomeric form ( imino base ) of the dye, which is in a labile equilibrium with the orthochromatic form. This equilibrium is displaced toward the metachromatic form by chromo- tropic substances. Michaelis and Granick,^^ on the other hand, have formulated the theory that the metachromatic form of the dye is simply a dimer ( or polymer ) of the ortho- chromatic (monomer) form. In the absorption spectrum, the absorption maximum of the monomer is represented by the a peak; that of the dimer by the /3 peak; that of polymers by the rather flat plateau of the y peak. Factors favoring polymerization are high concentration, low temperature, rela- tively high pH values, an aqueous medium, and especially the presence of large molecules with suitably placed and spaced acidic groupings (-COOH, -OPO(OH)2, -OSO2OH) which can hold the polymer dye molecule together by its 59. Lison, L.: Acad. roy. Belgique, cl. de sc, 19:1332, 1933. 60. Lison, L.: Arch, de biol., 46:599, 1935. 61. Michaelis, L., and Granick, S.: J. Am. Chem. See, 67:1212, 1945. Organic Substances 71 amino groups. Such substances with multiple acidic group- ings are the chromotropic substances. Depending on the strength of the acidic groups, the degree of metachromasia they produce may vary considerably. Some types of meta- chromasia can exist only in an aqueous medium and at a neu- tral or slightly acid reaction ( desoxyribose nucleic acid); others will resist acid down to pH 3 and mounting in glycerol (some connective-tissue polysaccharides). The extreme de- gree is resistant both to acid and to dehydration by acetone or alcohol, although the latter agents almost invariably decrease the color contrast. Lison^^' ^^ has shown that the alcohol-re- sistant form of metachromasia ( "true" metachromasia ) is due to the presence of sulfuric esters of large molecular size. Until recently the only known substances occurring in animal tis- sues and exhibiting true metachromasia actually were such sulfuric esters (heparin, chondroitin-, and mucoitin-sulfuric acids ) . However, in 1940 Bignardi^^ showed that, if glycogen is treated with chromic acid until it becomes negative by the iodine stain and negative or weakly positive by Schiff's reac- tion (more than 3 hours' exposure to 4 per cent chromic acid ) , it will exhibit true metachromasia. Francini^^ was able to show the same change in the case of starch. Glycogen made metachromatic by long chromation also strains in- tensely with mucicarmine. The intestine of some species (e.g., the hedgehog) contains both mucous and mucoid cells ( the latter stain poorly with mucicarmine and are not meta- chromatic ) . It is a highly interesting fact that, if sections of such a tissue are subjected to prolonged chromation, the mucinous cells will lose their typical staining properties while the mucoid cells become metachromatic and muci- carmine-positive.^^ The explanation of these findings is not 62. Lison, L.: Compt. rend. Soc. de biol., 118:821, 1935. 63. Bignardi, C: Atti Soc. ital. sc. nat, 79:85, 1940. 64. Francini, E.: Nuovo gior. bot. ital., 47:531, 1940. 65. Bignardi, C: Boll, di zool, 10:219, 1939; Bignardi, C: Atti Soc. nat. e mat. Modena, 71:59, 1940. 72 Microscopic Histochemistry entirely clear, but it is reasonably safe to asume that the im- derlying chemical change is not the detachment or attach- ment, respectively, of sulfuric groups to a polysaccharide. It is much more likely that a maximum number of carboxylic groups favors metachromasia; these are present in some sub- stances (mucin, amyloid, etc.) and are destroyed by pro- longed chromation. They are absent in other substances (glycogen, starch, mucoid) but can be produced via the intermediate stage of aldehydes by the same oxidative treat- ment. In any case, overchromated glycogen appears to be an example of a nonsulfate polysaccharide exhibiting true meta- chromasia. The latest addition to the substances exhibiting true metachromasia is ribose nucleic acid. This substance was always considered to be nonmetachromatic ( in Lison s sense ) ; however. Flax and Himes^^ have shown that under certain conditions (specifications not clear) it will stain in a strongly metachromatic shade, resistant to alcohol. Popper, Gyorgy, and Goldblatt^"^ described a peculiar metachromasia of ceroid stained with methyl green (not stated whether or not free from methyl violet). The nature of this metachromasia has not been investigated; it may be related to that of myelin as seen with the techniques of Feyr- ter^^ and Chang. ^^ Wislocki and Singer ^^ believe that Feyr- ter's metachromasia is due to the presence of sulfatides. Metachromasia of the mucinoid substances, heparin and chondroitinsulfuric acid, on the one hand, and that of amy- loid, on the other, are somewhat diflFerent, although, accord- ing to recent investigations, amyloid (not included in Meyer's classification) is a polysaccharide sulfate, closely similar to 66. Flax, M. H., and Himes, M. H.: Anat. Rec, 108:529, 1950; Himes, M. H., and Flax, M. H.: Anat. Rec, 108:539, 1950. 67. Popper, H., Gyorgy, P., and Goldblatt, H.: Arch. Path., 37:161, 1944. 68. Feyrter, F.: Virchows Arch. f. path. Anat., 296:645, 1936; Feyrter, F.: Wien. Idin. Wchnschr., 55:461, 1942; Feyrter, F., and Pischinger, A.: Wien. IcHn. Wchnschr., 55:463, 1942. 69. Min-Chueh Chang: Anat. Rec, 65:437, 1938. 70. Wislocki, G. B., and Singer, M.: J. Comp. Neurol., 92:71, 1950. Organic Substances 73 chondroitinsuKate.'^^ While both the mucinoids and amyloid will stain metachromatically with all tlie dyes enumer- ated,^^' '^^ the metachromasia of mucinoids with methyl yiolet is very much less marked than with the thiazin and oxazin dyes, and the opposite holds for amyloid. This fact shows that the phenomenon of metachromasia is more complex than it would appear from the relatively simple theory of Michaelis and Granick.^^ The results of metachromatic staining of mucin vary greatly with the tissue, fixation, and other incalculable details of technique. Actually, these random and uncontrollable vari- ations appear to be more important than the quaUty of the dye used or the special method followed. Of course, it can- not be denied that a certain batch of dye may give unusually good or unusually bad results, or that one of the specific techniques may yield better pictures than other methods in a given case. However, there is no single method which can be depended upon to give optimal results with all types of material. In our hands, satisfactory results were usually ob- tained with pinacyanol, introduced by Sylven,"^* with toluidin blue and with celestin blue."^^ For the metachromatic stain- ing of mucin, the use of almost any type of fixative, with the exception of bichromate-containing mixtures, is permissible. Methods a ) Toluidin blue or pinacyanol.— Use a dilute ( about 0.02- 0.05 per cent ) solution of either pinacyanol in distilled water or toluidin blue in a citrate buffer of pH 3.5-4.5. Stain section for 10-15 minutes or until, on inspection under the micro- 71. Hass, G.: Arch. Path, 34:92, 1942. 72. Johansson, G. A, and Wahlgren, F.: Acta path, et microbiol. Scandi- nav., 15:358, 1938. 73. Bignardi, C., and Casella, C.: Boll. Soc. med.-chir. Pavia, 55:843, 1941. 74. Sylven, B.: Personal communication. 75. Lendrum, A. C.: In Recent advances in clinical pathology (London: J. & A. Chm-chiU, 1947), p. 457. 74 Microscopic Histochemistry scope, nuclei are blue and mucin intensely pink. Rinse thor- oughly in water, dehydrate rapidly in absolute acetone or alcohol, clear in xylene, and mount in balsam. Mucin, carti- lage ground substance, and mast cells appear in shades of purple-blue to purple-red; nuclei, clear blue. Metachromasia of interfibrillar substance of connective tissue is greatly weak- ened by dehydration; this substance must be studied in unstable, water-mounted sections. b) Celestin blue.— Dissolve 0.1 per cent celestin blue in a 5 per cent solution of ammonium ( or potassium ) aluminum sulfate. Stain sections in this solution for 2-3 hours. Rinse under the tap, dehydrate, and mount. The color contrast is less brilliant than with toluidin blue but much more resistant to alcohol. Hale's technique for acid polysaccharides.— Hale described a technique^^ for the demonstration of acid polysaccharides, based upon the adsorption of colloidal iron hydroxide on acidic tissue components. In a second step the adsorbed iron is converted into Prussian blue. While the method sometimes gives a most beautiful and selective staining of some types of mucin, it cannot by any means be considered a specific method for acid polysac- charides. Not even all types of mucin are stained by it, but only those which are metachromatic with toluidin blue; the nonmetachromatic mucus of the stomach and of Brunner's glands is left entirely unstained. According to Grishman,^^ even some types of metachromatic mucin are negative with Hale's stain. Mast cells, which contain a strongly acid poly- saccharide, also stain very poorly. For connective-tissue poly- saccharides, Hale's method is greatly inferior to the Mc- Manus stain in respect to both sharpness and uniformity of results, especially after fixations other than alcohol. It usually stains chromatin quite intensely; in addition, there is a dif- fuse light-blue tinge to the background. In summary, the specificity of Hale's method is not limited to any chemically defined substance. The method will be de- Organic Substances 75 scribed here mainly because it is useful for the demonstra- tion of some types of mucin. Method Prepare a colloidal solution of ferric hydroxide by the method of Rinehart and Abul-Haj.^*^ Dissolve 30 g, of ferric chloride (lumps) in 100 ml. of distilled water, add 40 ml. of glycerol and, in small portions and under continuous stirring, 22 ml. of concentrated ammonia water. Dialyze the mixture under the tap for about 48 hours ( caution: fill the dialyzing bag less than halfway because there will be about a 2.5-fold increase in volume). Filter the dialyzate; it will keep indefi- nitely. For use, mix about 10 parts of the dialyzate and 1 part of concentrated acetic acid. This mixture is stable for 2 days only. Stain slides in the mixture for about 10 minutes, rinse them in repeated changes of distilled water, and im- merse for 10 minutes in a fresh solution containing about 1 per cent each of hydrochloric acid and potassium ferrocya- nide. Counterstain with a red nuclear dye, dehydrate, and mount. The mannitol-FeCls method of Lillie and Mowry*^^ gives results almost identical with those of Hale's technique. Ritter and Oleson"^^ suggest the McManus stain as a coun- terstain. Some substances will appear in a blue, others in a red shade; the interpretation of this difference is not clear. c) Heparin (granules of mast cells) .—Holmgren and Wi- lander"^^ suggest primary fixation in a 4 per cent solution of basic lead acetate for 12-24 hours; this may be followed, for better cytological detail, by formalin. Lead acetate precipi- tates heparin in the form oiF an insoluble lead salt. After this 76. Rinehart, J. F., and Abul-Haj, S. K.: Arch. Path., 52:189, 1951. 77. Lillie, R. D., and Mowry, R. W.: Bull. Intemat. A. M. Mus., 30:91, 1949. 78. Ritter, H. B., and Oleson, J. J.: Am. J. Path., 26:639, 1950. 79. Holmgren, H., and Wilander, O.: Ztschr. f. mikr.-anat. Forsch., 42:242, 1937; Holmgren, H.: Ztschr. f. wissensch. Mikr., 55:419, 1938; Holmgren, H.: Ztschr. f. mikr.-anat. Forsch., 47:489, 1940. 76 Microscopic Histochemistry fixation, the mast cells will stain almost blue-black, with only a slight tinge of red. The shade is probably due, at least in part, to the adsorption of the dye on the lead compound ( tolu- idin blue is adsorbed strongly on various insoluble lead precipitates ) . d) Amyloid.— The term "amyloid" was coined by Vir- chow^^ to denote a homogeneous., somewhat translucent sub- stance which is deposited in connective tissue under certain pathologic conditions and is stained somewhat like starch (amylos equals starch) by iodine. Chemically, the characteristic component of amyloid ap- pears to be a sulfuric ester of a polysaccharide."^^ Depending on their age and other poorly understood fac- tors, the staining reactions of deposits of amyloid are rather variable. Recent deposits may not show some or even any of the typical staining properties, whereas old deposits usually stain according to textbook specifications. The two most important tinctorial features of amyloid are its stainability by iodine and its metachromasia. The iodine reaction.— This is very similar to that of glyco- gen, with the exception that the mahogany brown shade may turn into a dark gray-bluish or greenish one on the applica- tion of dilute mineral acid. Metachromasia of amyloid: Methods 1. Stain section in a dilute (0.1-0.2 per cent) solution of methyl violet (or crystal violet) for about 10 minutes. Dif- ferentiate in 1 per cent acetic acid until amyloid is purple- red and the background blue. Rinse, mount in glycerin jelly. Preparations not permanent. 2. This variant yields permanent preparations. Float the parafiin sections directly on the dye solution (warmed to about 37°-45°C.) for about 15-20 minutes. Refloat them on water to wash out excess dye, then on 1-2 per cent acetic 80. Virchow, R.: Virchows Arch. f. path. Anat., 6:416, 1854. Organic Substances 77 acid until sections are suflBciently differentiated. Refloat on water once more; mount on slides. Dry slides, remove paraffin with xylene, and mount section in balsam. This technique avoids all dehydrating agents and preserves metachromasia very well. II. NUCLEIC ACIDS The three components of the nucleic acids are ( 1 ) purine and pyrimidine bases; (2) pentose or desoxypentose sugar; and (3) phosphoric acid. The histochemical methods for their demonstration will accordingly be divided into three groups, depending on which of the components is identified. The specificity of the identifying reactions can be checked by extraction techniques, enzymatic or other, which remove nucleic acids in a selective way. 1. The purine and pyrimidine bases can be demonstrated by the ultraviolet absorption method only. Recently, Danielli^^ has proposed another method which may be called descrip- tively "double azo-coupling." It is based on the theory that purine and pyrimidine bases will couple with diazonium salts. The resulting azo dye is, however, only pale yellowish in shade and unsuitable for direct observation. For this reason, Danielli uses a tetrazonium compound ( of benzidine or dia- nisidine ) which attaches itself to the base by only one of its diazo groups; the other free, diazonium group can then be coupled with highly chromogenic naphthol, and an intensely colored (purplish) azo dye results. Tyrosine, histidine, and tryptophane give similar color reactions; benzoylation, how- ever, will abolish their coupling ability while it will leave that of the purine and pyrimidine bases intact. The chemical basis of this method does not appear to be sufficiently firm. First of all, chemical studies on the azo-couphng of purines 81. Danielli, J. F.: Symp. Soc. Exper. Biol., 1:101, 1947. 78 Microscopic Histochemistry (Burian,^^ Johnson and Clapp,^^ Steudel,^* H. Fischer,^^ and Hunter^^ ) all agree that this coupling takes place exclusively in the presence of a high concentration of caustic alkali ( car- bonates, e.g., will not do). It is also specified that purines substituted in the imidazole ring ( caflFeine, theobromine, nu- cleotides ) will not react at all.^^' ^^ Therefore, if azo-coupling in slides is attempted according to the suggestion of Mitch- ell,^"^ as modified by Danielli^^ (at pH 9), no reaction can be expected to take place. Actually, in Coujard shdes, carry- ing marks made with gelatin alone and with suspensions of guanine, adenine, xanthine, uracil, uric acid, and RNA, no differentiation of any kind can be obtained. Second, all diazonium compounds are quite labile, espe- cially at an alkaline reaction. On standing (even at 0° C), they form dark-colored decomposition products which will stain almost anything (e.g., a strip of filter paper or even a collodion membrane ) and cannot be washed out completely. When pieces of such stained material, after thorough wash- ing in water and alcohol, are immersed in an alkaline solu- tion of ^-naphthol, they will stain intensely purple. Tissues treated by this method are stained quite diffusely, without much differentiation of morphologic detail. The enterochro- maffin granules are the only exception to this. They stain selectively and quite intensely both after the primary and after the secondary azo-coupling (the latter first recom- mended by Clara and Canal in 1932).^^ There can be no doubt that after acylation there is a marked decrease in the staining of all tissue structures, but the contrast is not much improved. It is remarkable that reactivity of the enter o- 82. Burian, R.: Ztschr. f. physiol. Chem., 51:425, 1907. 83. Johnson, T. B., and Clapp, S. H.: J. Biol. Chem., 5:163, 1908. 84. Steudel, H.: Ztschr. f. physiol. Chem., 48:425, 1906. 85. Fischer, H.: Ztschr. f. physiol. Chem., 60:69, 1909. 86. Hunter, G.: Biochem. J., 30:745, 1936. 87. Mitchell, J. S.: Brit. J. Exper. Path., 23:296, 1942. 88. Clara, M., and Canal, F.: Ztschr. f. Zellforsch. u. mikr. Anat., 15:801, 1932. Organic Substances 79 chromaffin phenol is not completely abolished even after 24 hours' acylation. For the reasons mentioned, great caution is warranted in the evaluation of the results of DanielH's technique. Its mech- anism is not entirely clear, and the possibility that the re- sults are largely nonspecific cannot be ruled out. 2. The sugar moiety is identified by aldehyde reactions after acid hydrolysis. The most commonly used aldehyde re- agent is SchiE's, which will demonstrate only desoxyribose (Feulgen's nucleal reaction ).^^ Schiff's reagent can be re- placed by other chromogenic carbonyl reagents, such as 2- hydroxy-3-naphthoic acid hydrazide followed by azo-coup- ling,^^ with essentially similar results. Turchini's reagent (9-phenyl-2,3,7-trihydroxy-6-fluorone) will condense with both ribose and desoxyribose and, in fact, even with hex- oses.^^ The condensation products with these sugars have different colors (ribose, yellowish pink; desoxyribose, bluish purple; hexoses, reddish purple ) . RNA must be protected by fixation in dichromate-containing mixtures (which make nu- cleic acids more resistant to hydrolysis; Hillary );^^ otherwise it may be lost in the procedure. No reports on the reliability of this method are available so far except by Turchini and his co-workers. 89. Feulgen, R.: Die Nuclealfarbung. In Abderhalden's Handb. d. biol. Arbeitsmeth., ¥2-2:1054, 1932. 90. Pearse, A. G. Everson: J. Clin. Path., 4:1, 1951. 91. Turcliini, J., Castel, P., and Kien, K.: Trav. soc. chim. biol., 35:1329, 1943; Turchini, J., Castel, P., and Kien, K.: Montpellier med., 23-24:599, 1943; Turchini, J., Castel, P., and Kien, K.: ibid., 25-26:396, 1944; Tur- chini, J., Castel, P., and Kien, K.: Bull, d'bistol. appHq. a la physiol., 21:124, 1944; Turchini, J., and Gosselin de Beaumont, L. A.: Compt. rend. Soc. de biol., 139:584, 1945; Kien, K., and Sentein, P.: Compt. rend. A. anat., p. 264, 1947; Turchini, J., Castel, P., and Kien, K.: Compt. rend. A. anat., p. 456, 1947; Turchini, J., and Kien, K.: Compt. rend. A. anat., p. 391, 1948; Turchini, J., Castel, P., and Kien, K.: Compt. rend. Soc. de biol., 142:1277, 1948; Turchini, J.: Exper. Cell Research, Suppl., 1:105, 1949; Turchini, J., and Kien, K.: Xllle Congres internat. de zool. Paris, p. 207, 1949. 92. Hillary, B. B.: Bot. Gaz., 101:276, 1939. 80 Microscopic Histochemistry The specificity of the Feulgen nucleal reaction has been the subject of a hvely controversy until very recently. The following arguments have been voiced against it: 1. There need be no chemical interaction between nucleic acid and the reagent. The active dye may be fuchsin itself, adsorbed by the nuclei, just as, e.g., alumina can adsorb fuchsin from Schiff's reagent. The effect of hydrolysis may consist simply in a dissolution of the cytoplasm while the nucleus is relatively resistant; in this way the contrast be- tween the nucleus and the cytoplasm is enhanced. The fact that an excess of bisulfite (which should block aldehyde groups) does not prevent staining of the nuclei shows that it is not due to an aldehyde reaction.^^ 2. Recolorization of SchifF's reagent may be due to non- aldehydic substances; pyridine and various purine bases will give color reactions more or less readily.^^ 3. It is admitted that the dye produced by the interaction of Schiff's reagent and hydrolyzed DNA is a true aldehyde addition product; however, it is soluble in water and a good stain for chromosomes,^^ especially for chromosomin and his- tone.^^ Therefore, DNA only contributes to the dye which then stains something else.^^- ^^ The weak points of these arguments can be easily shown: 1. The specificity of the nuclear staining is not a matter of increase in contrast; without hydrolysis nothing whatsoever will stain in a properly performed Feulgen test. As for the blocking of aldehyde groups by an excess of SO2, it must be remembered that aldehyde-bisulfite addition compounds (especially those of aldehydes of higher molecular weight) are very unstable and prone to break down.^^ 2. Recolorization of fuchsin by nonaldehydic compounds 93. Carr, J. G.: Nature, 156:143, 1945. 94. Semmens, C. S.: Nature, 146:130, 1940. 95. Choudhuri, H. C: Nature, 152:475, 1943. 96. Sibatani, A.: Nature, 166:355, 1950. 97. Stedman, E., and Stedman, E.: Symp. Soc. Exper. Biol., 1:232, 1947. 98. Dodson, E. O.: Stain Technol., 21:103, 1946. Organic Substances 81 is a pseudo-reaction, not resistant to acid and alcohol. Also, the recolorization of Schiff's reagent by purines could not be confirmed.^^ That the active dye is really an aldehyde ad- dition product can be proved by blocking the reaction with specific carbonyl reagents. ^^^ Cyanide is especially eflFec- tive.^^^ The successful substitution of Schiff's reagent by other reagents has been mentioned. The aldehyde compound must come from DNA; if the latter is removed by specific enzymes, the Feulgen reaction is abolished. 3. The dye prepared in the test tube from hydrolyzed DNA and Schiff's reagent is soluble (although not dialyz- able).^^^ However, this does not apply to tissue sections in which nucleic acid is present in firm chemical union with proteins and in a completely insoluble state. It is not made soluble by mild acid hydrolysis. ^^^ The nonnal HCl from a Coplin jar in which nine slides, carrying several large tissue sections each, were hydrolyzed for 12 minutes did not show the slightest trace of color with Schiff's reagent. There is no reason to assume that a degraded but insoluble DNA would form a soluble addition compound with Schiff's reagent; in fact, there is evidence to the contrary. ^^^ In Coujard slides, marks made with DNA of high purity and fixed in formalin- alcohol gave an intense and sharp reaction, with absolutely no indication of diffusion; marks made with RNA and several proteins were completely negative. Of course, it is conceiv- able that traces of moderately soluble fragments may be lib- erated during hydrolysis and converted into dye which may stain structures in the immediate proximity. However, it is difficult to explain why the infinitesimal concentration of dye 99. Barber, H. N., and Price, J. R.: Nature, 146:335, 1940. 100. Lessler, M. A., and Kopac, M. J.: Anat. Rec, 108:531, 1950; Less- ler, M. A., and Kopac, M. J.: ibid., 108:578, 1950; Lhotka, J. F., and Davenport, H. A.: Stain Technol., 26:35, 1951. 101. Gomori, G.: Unpublished. 102. Ely, J. O., and Ross, M. H.: Anat. Rec, 104:103, 1949. 103. Brachet, J.: Experientia, 2:142, 1946. 104. Li, C., and Stacey, M.: Nature, 163:538, 1949. 82 Microscopic Histochemistry present locally at any given moment (both the hydrolyzed DNA and the dyestuff are assumed to be soluble and mobile ) should stain the chromosomes intensely and sharply in a matter of a few minutes when a strong solution of the dye prepared in the test tube from hydrolyzed DNA and Schiff's reagent will give only a blurred and weak staining, even on prolonged application. Additional proofs for the specificity of the Feulgen stain and of the correctness of its localization are the staining of isolated chromatin threads^^^ and the excellent agreement between the Feulgen reaction and the results of localizing ultraviolet spectrophotometry ( Caspersson ) }^^ In summary, it may be said that the Feulgen reaction is a reliable method for the specific localization of DNA, pro- vided that adequate unhydrolyzed and hydrolyzed but alde- hyde-blocked controls are used in the case of doubt. The nature of the positive Feulgen reaction in nervous elements (in unembedded tissue only), reported by Liang^^"^ and Chu,^^^ is not clear. The Feulgen reaction can be performed in tissue blocks;^^^ however, the section technique is much more preferable. The optimal duration of hydrolysis depends on the type of fixation. In the case of dichromate-free fixatives, the most intense reactions are obtained between 8 and 12 minutes (N HCl, 60°C.). Overhydrolysis, will gradually abohsh the reaction. There is a difference between the resistance of nuclei of various tissues, thymus nuclei being the most re- sistant (up to 30 minutes ).^^^ After dichromate-containing 105. Claude, A., and Potter, J. S.: J. Exper. Med., 77:345, 1943; Barber, H. N., and Callan, H. G.: Nature, 153:109, 1944. 106. Caspersson, T.: Nature, 153:499, 1944. 107. Liang, H. M.: Anat. Rec., 99:511, 1947. 108. Chu, C. H. U.: Science, 106:70, 1947. 109. Lhotka, J. F., and Davenport, H. A.: Stain Technol., 22:139, 1947; Lhotka, J. F., and Davenport, H. A.: ibid., 24:127, 1949. 110. DeLamater, E. D., Mescon, H., and Barger, J. D.: J. Invest. Dermat., 14:133, 1950. Organic Substances 83 fixatives the optimum duration of hydrolysis is much less critical (10-30 minutes ).^^ SchifF's reagent itself may cause some hydrolysis on long exposure. ^^^ This is especially important in the case of stain- ing unhydrolyzed control sections because of the possibility of obtaining false positive reactions. The following procedure is recommended: Method Hydrolyze tissue sections in N HCl preheated to 58°-62°C. for 8-10 minutes. Wash them under the tap and stain in Schiff's reagent as described in the section on polysaccha- rides (p. 59). Naturally, omit counterstaining with hema- toxylin; a yellow or green acid counterstain (picric acid, orange G, or light green) is permissible and may actually be advantageous. 3. The acid moiety. There are a few methods published for the specific identification of phosphoric acid in the nucleo- tides. ^^^ The reliability of these methods, especially as far as correct localization is concerned, is questionable (see under "Phosphorus" ) . Other methods demonstrate simply the presence of acidic groups of any kind. The use of indicators cannot be depended upon to yield accurate information because the color con- trasts are not sharp enough. The only methods to be used extensively are based on basophilia, the afiinity for basic dyes, which is a property of any substance possessing acidic groups. To obtain con- firmatory evidence for nucleotides as the cause of basophilia, specific extraction procedures must be resorted to in many cases because of the presence of basophilic substances other than nucleic acids (mucopolysaccharides) in the tissues. The basic dyes most often used are methyl green, meth- 111. Serra, J. A.: Bol. Soc. Broteriana, 17:203, 1943. 112. Lilienfeld, L., and Monti, A.: Ztschr. f. wissensch. Mikr., 9:332, 1892; Serra, J. A., and Queiroz Lopes, A.: Port, acta bioL, 1:111, 1945. 84 Microscopic Histochemistry ylene blue, thionin, toluidin blue, pyronin, and safranin. Methyl green stands in a class by itself on account of its unique property of staining some, but not all, of the baso- philic substances. It will stain high-polymer DNA as it occurs in the nuclei or as obtained by gentle extraction procedures, and also sulfate-polysaccharides, but not depolymerized DNA or RNA in any form. The reasons for this curious be- havior are not well understood. ^^^' ^^^ In practice, however, a combination of methyl green and pyronin (Pappenheim and Unna)^^^ is a most useful dye mixture, giving an excel- lent contrast between green-staining DNA of nuclei and red- staining RNA. All other dyes mentioned will stain both the nucleotides and also acid polysaccharides; the latter may be stained in a metachromatic shade. It is important to use the dyes at a neutral or slightly acid ( pH 5-6 ) reaction, because otherwise they may be taken up even by nonbasophilic struc- tures. The nature of the binding of basic dyes by basophilic sub- stances is not completely understood. There are good indi- cations that in a solution there is a strict stoichiometric re- lationship between the acidic groups and the amount of dye bound.^^^ However, it is very questionable whether this re- lationship holds for the insoluble nucleoproteins of the tissues where an undetermined proportion of the acidic groups may be tied down to strongly basic proteins in a stable, nonre- active form. It is quite likely that in sections ionic forces, im- 113. Pollister, A. W., and Leuchtenberger, C: Proc. Nat. Acad. Sc, 35:111, 1949. 114. Vendrely, C, and Vendrely, R.: Compt. rend. Soc. de biol., 143:1388, 1949; Kumick, N. B.: J. Gen. Physiol., 33:243, 1950; Kumick, N. B., and Mirsky, A. E.: J. Gen. Physiol., 33:265, 1950; Kumick, N. B.: J. Nat. Cancer Inst., 10:1345, 1950; Taft, E. B.: Exper. Cell Research, 2:312, 1951. 115. Unna, P.: Plasmazellen. In Enzykl. mikr. Technik (2d ed.; 1910), p. 744. 116. Chapman, L. M., Greenberg, D. M., and Schmidt, C. L. A.: J. Biol. Chem., 72:707, 1927. Organic Substances 85 portant as they may be, are not the only factors at work.^^^ Other structural properties (the presence of polar groups; steric configuration), both of the adsorbing nucleoprotein and of the adsorbed dye, may be more important than acid- base aflfinity, just as in the case of cation exchangers^^^ (resins and other substances). For instance, a compound as little acidic as silica (pK of silicic acid, ±9.7) will quantitatively remove methylene blue from a solution in 0.1 N sulfuric acid. Many attempts have been made in the past to stain tissue sections with basic dyes at graded pH levels and to draw meaningful conclusions from the more or less abrupt changes in staining intensity. ^^^"^^ The pH at which such sudden changes occur has variously been called the isoelectric point of the substance stained^^®' ^^^' ^^^' ^"^' ^^^ or the acidic dissoci- ation constant of its basophilic group.^^*' ^^^ The fundamental assumptions behind such experiments are that (a) acidic substances will stain with increasing intensity above their isoelectric point ( or pK of their acidic groups ) , while below this point their staining intensity will decrease rapidly; and 117. Michaelis, L.: Arch. f. mikr. Anat., 94:580, 1920. 118. W^hitehom, J. C: J. Biol. Chem., 56:751, 1923; Walton, H. F.: Ion exchange equilibria. In Ion exchange, theory, and application, ed. F. C. Nachod (New York: Academic Press, 1949). 119. Pischinger, A.: Ztschr. f. Zellforsch. u. mikr. Anat., 3:169, 1925-26. 120. Pischinger, A.: ibid., 5:347, 1927. 121. Hammarsten, E., Hammarsten, G., and TeoreU, T.: Acta med. Scan- dinav., 68:219, 1928; Zeiger, K.: Ztschr. f. ZeUforsch. u. mikr. Anat., 10:481, 1930. 122. Fautrez, J.: Bull, d'histol. appUq. a la physiol., 13:202, 1936. 123. Levine, N. D.: Stain Technol., 15:91, 1940. 124. Kelley, E. G.: J. Biol. Chem., 127:55, 1939. 125. Kelley, E. G.: ibid., 127:73, 1939. 126. Seki, M., and Kohashi, Y.: Okajima's FoHa anat. Jap., 19:47, 1940. 127. McCaUa, T. M.: Stain Technol., 16:27, 1941. 128. Hyden, H.: Acta physiol. Scandinav., Vol. 6, Suppl. 17, 1943. 129. Dempsey, E. W., Bunting, H., Singer, M., and Wislocki, G. B.: Anat. Rec, 98:417, 1947; Singer, M., and Morrison, P. R.: J. Biol. Chem., 175:133, 1948. 86 Microscopic Histochemistry (b) the part of dye which is adsorbed in excess of the ion exchange mechanism can be removed with alcohol/^^ or that adsorption can actually be prevented by the use of deter- gents incorporated in the staining solution. ^^^ Extreme caution is warranted in the interpretation of the results of such experiments. First of all, the results depend within wide limits on minor variations in technique ( the dye used, its concentration, time of staining, washing, and differ- entiation). Especially differentiation is a very delicate step. While acid in an aqueous medium has a relatively moderate effect on the staining by basic dyes, traces of it transferred to the differentiating alcohol (either from the dye solution or from the buffer used to wash out the excess dye ) may de- colorize the section extensively during the first few seconds of contact, even at pH values as high as 4-4.5. On the other hand, tissues stained at a pH as low as, 1.2 or lower (and they do stain with considerable intensity; even Gabbett's^^^ meth- ylene blue, which contains 25 per cent sulfuric acid, will stain nuclei), differentiated in a buffer of the same pH, and sub- sequently washed thoroughly to remove all traces of acid may retain a beautfful nuclear staining even after prolonged differentiation in alcohol. ^^^ In summary, it may be said that, other things being equal, substances possessing acidic groups are more likely to stain with basic dyes, especially at pH levels below neutrality, than substances without such groups. However, basophilia is the result of many factors, most of which are very poorly under- stood. It does not permit the drawing of even approximately quantitative conclusions as to isoelectric points or the values of dissociation constants. Extraction procedures Strictly speaking, "extraction" is a misnomer if applied to the whole group of procedures to be described here, because 130. Michaelis, L.: Cold Spring Harbor Symp. Quant. Biol., 12:131, 1947. 131. Gabbett, H. S.: Lancet, 1887-1, p. 757. 132. Gomori, G.: Unpublished. Organic Substances 87 some of them do not actually extract nucleotides but rather alter them sufficiently to abolish their normal staining char- acteristics. Most of the others, however, do break them down to small, soluble fragments which will diflFuse out from the section. The efficiency of all extraction procedures greatly depends on the type of fixation. Alcohol-, formalin-, and dichromate- fixed tissues are attacked with increasing difficulty in the or- der mentioned. Resistance to extraction also varies, with the type of tissue. The procedures will be divided in two groups : nonenzymatic and enzymatic. a) Nonenzymatic procedures.— Hot water will abolish the staining of DNA by methyl green, while the Feulgen reac- tion remains unchanged. ^^^ Exposure to 5 per cent citric acid at room temperature for 12 hours has the same eflFect.^^^ Ac- cording to Mirsky and Pollister,^^^ nucleoproteins can be ex- tracted from tissues with M NaCl; the use of this method has not been attempted histochemically. Strong acids, extract RNA first, and, on longer exposure, DNA is also removed. For instance, the Feulgen type of hy- drolysis (N HCl; 60° C.) removes RNA in about 3-10 min- utes ;^^^ DNA is not completely removed up to 50 minutes. ^*^^ Five to 10 per cent perchloric acid removes RNA in 4-18 hours at 5° C. and in about 2 hours at 25° C.^^^ DNA is not completely extracted at the latter temperature in 22 hours; at 70° C. both DNA and RNA are extracted in 20 minutes.^^^ Five per cent trichloroacetic acid also extracts both nucleo- 133. Brachet, J.: Arch, de biol., 53:207, 1942. 134. Mirsky, A. E., and Pollister, A. W.: Proc. Nat. Acad. Sc, 28:344, 1942. 135. Deane, H. V^.: Am. J. Anat., 78:227, 1946; Vendrely, R., and Lipardy, J.: Compt. rend. Acad, sc, 223:342, 1946; Vendrely-Randavel, C: Compt. rend. Soc. de biol., 143:294, 1949. 136. Ogur, M., and Rosen, G.: Fed. Proc, 8:234, 1949. 137. Erickson, R. O., Sax, K. B., and Ogur, M.: Science, 110:472, 1949; Seshachar, B. R., and Flick, E. W.: Science, 110:659, 1949; Sulkin, N. M., and Kuntz, A.: Proc. Soc Exper. Biol. & Med., 73:413, 1950; Koenig, H.: J. Nat. Cancer Inst., 10:1346, 1950. 88 Microscopic Histochemistry tides at about the same rate/^^ Extraction of RNA with 0.1 N KOH is recommended by Sulkin.^^^ Henry and Stacey^^^ have presented excellent evidence to show that the Gram-positive staining of bacteria is due to an RNA-Mg complex, extractable with a 2 per cent bile salt solution at 60° C. This bacterial RNA is obviously different from the ordinary type of RNA, which is Gram-negative and insoluble in bile salt solutions. b) Enzymatic procedures.— EnzymsLtic hydrolysis of nu- cleic acids follows a course different from that seen in acid hydrolysis. The first stage is depolymerization of a variable degree; the next stage is the removal of PO4 groups; and the last one is the breakdown of glycosidic linkages, which are the first ones to be attacked by acids. ^'^^ It is very likely that every stage has its own enzyme or enzymes ( depolymerase, phosphatase, nucleosidase) and that many of the enzyme preparations used in histochemistry are mixtures of several components. In fact, they are often contaminated by pro- teases of various kinds. This explains the different results obtained by the use of enzymes prepared in different ways. The depolymerases of RNA and DNA are definitely spe- cific; a good preparation of desoxyribonuclease will not at- tack RNA and vice versa. According to Danielli,^^ the specificity of the nucleases is open to doubt. He also questions whether the histochemical application of enzymes can ever yield quantitative results, since the penetration of the large enzyme molecules into the interior of the section would be blocked by even a monolayer of protein. However, there is ample experimental evidence 138. Koenig, H., and Stahlecker, H.: J. Nat. Cancer Inst., 12:237, 1951; Schneider, W. C: J. Biol. Chem., 161:293, 1945. 139. Sulkin, N. M.: Proc. Soc. Exper. Biol. & Med., 78:32, 1951. 140. Henry, H., and Stacey, M.: Nature, 151:671, 1943; Henry, H., Stacey, M., and Teece, E. G.: Nature, 156:720, 1945; Henry, H., and Stacey, M.: Proc. Roy. Soc. London, B, 133:391, 1946. 141. Catcheside, D. G., and Holmes, B.: Symp. Soc. Exper. Biol., 1:225, 1947. Organic Substances 89 both for the specificity of at least some of the nucleases and for the quantitative eiBFect of enzymes (for instance, diastase and ribonuclease ) . The preparation of highly active ribonuclease of excellent specificity is simple and v^ill be given here (Brachet's method ).^^^ The step of boiling the crude enzyme solution destroys all enzymes except the heat-resistant ribonuclease. Method Grind beef pancreas to a smooth pulp and suspend it for 24 hours at 37° C. in 1-2 volumes of 0.1 N acetic acid. Boil it for 10 minutes and filter. Neutralize the filtrate to about pH 6.9-7.5 and filter it once more. This solution can be kept in the icebox, with some camphor or thymol added, for sev- eral months. Brachet recommends dialysis as a last step, but it is not necessary. Incubate sections in the enzyme solution for 1 hour at 65°-70° C. Treat a control section with a buffer of the same pH. The enzyme-treated section will show a complete loss of all basophiha due to RNA^^^ (but not of that due to muco- polysaccharides ) . The Feulgen reaction of the nuclei is not affected. Of the desoxyribonuclease preparations, McCarty's^^* ap- pears to be the most specific. Its preparation is not easy; in most cases it will be simpler to purchase a commercial prep- aration. Desoxyribonuclease is not resistant to heat; it should be used around 37° C. A good preparation will abolish the Feulgen reaction of the nuclei but leave cytoplasmic baso- philia intact. According to Mazia,^^^ intestinal phosphatase is also effec- tive in removing RNA. Acid phosphatase has not been used histochemically. 142. Brachet, J.: Compt. rend. Soc. de biol., 133:88, 1940. 143. Brachet, J., and Shaver, J. R.: Stain Technol., 23:177, 1948; Deane, H. W.: Am. J. Anat, 78:227, 1946. 144. McCarty, M.: J. Gen. Physiol., 29:123, 1946. 145. Mazia, D.: Cold Spring Harbor Symp. Quant. Biol, 9:40, 1941. 90 Microscopic Histochemistry APPENDIX Ascorbic Acid Ascorbic acid has the unique property of reducing an acidi- fied solution of silver nitrate almost instantly. This unusual reducing power can be utilized for its histochemical identi- fication. The first histochemical experiments were made by Szent-Gyorgyi^*® even before the chemical nature of ascorbic acid was known. While the specificity of the reaction (if performed cor- rectly) is reasonably certain, the correctness of localization is much less so. Ascorbic acid is a highly diffusible substance which cannot be expected to stay at its original sites unless the tissue is frozen-dried. Its displacement cannot be avoided even by fixation in formaldehyde vapor^^"^ or by perfusing the tissue with reagent,^^^ although localization (in the gross histological, but not cytological, sense) may be improved by these maneuvers. The experiments of Barnett and Fisher^*^ clearly show the importance of physical factors, such as lipid-water interfaces, on the localization of the reaction. According to Huszak,^^^ Giroud and Leblond,^^^ and Bar- nett and Boume,^^^ a negative reaction does not rule out the presence of ascorbic acid because some tissues seem to con- tain a substance which prevents the reduction of silver ni- trate. The last-named authors believe that this inhibitor is glutathion. This substance actually slows down the reaction considerably in test-tube experiments. Method^^^ The reagent is an aqueous or alcoholic 1-5 per cent solu- tion of silver nitrate acidified with some acetic acid; concen- 146. Szent-Gyorgyi, A.: Biochem. J., 22:1387, 1928. 147. Bourne, G.: Australian J. Exper. Biol., 11:261, 1933. 148. Giroud, A., and Leblond, C. P.: Arch, d'anat. micr., 30:105, 1934. 149. Bamett, S. A., and Fisher, R. B.: J. Exper. Biol., 20:14, 1944. 150. Huszak, S.: Ztschr. f. physiol. Chem., 222:229, 1933. 151. Giroud, A., and Leblond, C. P.: Nature, 138:247, 1936. 152. Bamett, S. A., and Bourne, G.: J. Anat., 75:251, 1941. 153. Giroud, A., and Leblond, C. P.: Bull, d'histol. appliq. a la physiol., 11:375,1934. Organic Substances 91 trations are of little importance. Small pieces of tissue are fixed directly in this fluid. In case of intravascular injection it is advisable to perfuse the tissue with an isotonic (about 5 per cent) solution of glucose first, because otherwise silver chloride and silver protein precipitates may clog the vessels. The reagent is allowed to act for not much longer than neces- sary for complete penetration of the tissue (5-20 minutes, depending on its size ) ; the excess of silver is removed by re- peated changes of 1 per cent Na thiosulfate, and the latter by distilled water. The tissue may be frozen-cut or dehy- drated and embedded. A black granular precipitate indicates the presence of ascorbic acid. Unless exposure to the silver solution is unduly prolonged (in which case urates may re- act), false positive reactions need not be feared. Gold chloride^^^ can be used instead of silver nitrate; the results are claimed to be identical. To the writer's knowledge, the demonstration of ascorbic acid has not been tried in sections of frozen-dried material. It is not even certain that the reduction of silver nitrate is prompt enough for a sharp localization. However, in attempts at using frozen-dried tissues it should be borne in mind tliat ascorbic acid is soluble in alcohol; therefore, the reagent should be applied to deparaJBfinized, air-dried sections. B. LIPIDS "Lipids" (synonyms: "lipoids," "fatty substances") is a term which will be used to denote a large group of miscel- laneous chemical substances classified together, for histo- chemical purposes only, by their solubility properties. These properties are insolubihty in water and solubiUty in several or all of the so-called "fat solvents" (alcohol, ether, chloro- form, benzene, pyridine, acetone, etc.). The histochemistry of hpids has been reviewed very thor- oughly by Cain,^ and the reader interested in the topic is urged to consult this excellent paper for detailed inf onnation. An immense variety of lipid substances occurs in the ani- 1. Cain, A. J.: Biol. Rev. Cambridge Phil. Soc, 25:73, 1950. 92 Microscopic Histochemistry mal and plant kingdoms. The classification which follows does not claim completeness; it includes only such lipids as are likely to be met with in normal and pathological tissues of man and of the common laboratory animals. Classification of Lipids A. Paraffins (petrolatum). B. Isoprene derivatives (carotenoids). This gi-oup includes carotene, vitamin A, visual purple, and some of the pigments of crustaceans. C. Fatty acids and their derivatives. a) Fatty acids. b) Soaps, especially those of Ca. Although soaps are insoluble in fat solvents, they are included in this group for reasons of close chemical relationship. c) Triglycerids (neutral fats). d) Waxes (long-chained alcohol esters of fatty acids). e) Phosphatids.2 1. Lecithins (glycerine esterified with 2 molecules of fatty acid and 1 molecule of phosphorylcholine ) . 2. Cephalins (glycerine esterified with 2 molecules of fatty acid and 1 molecule of either phosphorylcolamine or phos- phorylserine ) . Lipositol is a complicated cephalin-like sub- stance with inositol as an additional component. 3. Plasmalogens (glycerophosphorylcolamine [or choline] in a cyclic acetal linkage with 1 molecule of fatty aldehyde) . 4. Sphingomyelin (sphingosine esterified with 1 molecule of phosphorylcholine and in acylamide linkage with 1 molecule of fatty acid). /) Cerebrosides.2 Sphingosine galactoside or glucoside in acylamide linkage with 1 molecule of fatty acid (examples: kerasin, phren- osin) are the simplest representatives of this group; there are also more complicated ones containing several units of various sugars and unidentified amino acid components. A special group of cerebrosides contains sulfuric acid (see the review of Blix^ on sulfur-containing lipids) . D. Lipid peroxides. E. Steroids. a) Cholesterol. 2. Page, J. H.: Chemistry of the brain (Springfield, 111., and Baltimore, Md.: C. C. Thomas, 1937). 3. Blix, C: Ztschr. f. physiol. Chem., 219:82, 1933. Organic Substances 93 b) Cholesterol esters. c) Steroid hormones. F. Group of chemically unidentified lipid pigments, to be dealt with in greater detail in the section on pigments (lipofuscin, ceroid). These substances appear to be complicated polymerization products of unsaturated fatty acids. Their solubility in lipid solvents is variable; some of them will resist even embedding in paraffin. It should be made clear that almost every group of sub- stances mentioned includes a vast array of individual com- pounds. The theoretically possible number of glycerids, and phosphatids, according to the laws of permutation, runs into the thousands because of the large variety of fatty acids which can combine. In addition, practically none of the in- dividual groups occurs singly within morphological struc- tures (droplets, mitochondria, etc.) but almost always as admixtures with other groups and in highly variable ratios. The physical properties of pure substances may be markedly modified by admixtures. This is a very important point because some of the most widely used histochemical methods for the demonstration of lipids, depend on the purely physical phenomenon of the solubility of certain dyes in them. The two most important features of lipid substances affect- ing their stainability by dyes are their melting points and the presence or absence of hydrophilic groups ( especially if the dye is applied in an aqueous solution ) . It is obvious that in the case of a solid, hydrophobic lipid a partition equilib- rium between the lipid and the aqueous ( or dilute alcoholic ) phases can never be established, simply because the dye can- not penetrate the lipid particle to any appreciable depth. On the other hand, a lipid of oily consistency or one possessing hydrophihc groups may be stained through and through. All transitions between these two extremes are possible. It is easy to see that admixtures lowering the melting point or im- parting a certain degree of hydrophilia will enhance the staining of the particles in bulk, whereas coating of other- 94 Microscopic Histochemistry wise tingible substances by even a thin layer of water-repel- lent lipid may block the contact between the interior of the particle and the dye solution. Owing to the similarity in the physical properties of the lipids as a group, sharp separation of the individual sub- groups is a formidable task even in conventional preparative chemistry, and an impossibility in histochemistry. Some de- gree of selective removal or preservation of certain compo- nents can be accomphshed, but the results are not even re- motely quantitative. Another complicating factor is the pro- tective action of proteins. For instance, kerasin is extractable from ground-up tissue by hot methyl alcohol or chloroform, but it is almost insoluble (especially after formahn fixation) while combined with protein within intact cells. ^ This complexity of the situation makes the histochemical analysis of lipids very difficult and often illusory. Even the results of model experiments, which have been done in fairly large numbers^' ^' '^' ^ since Altmann's original attempt in 1890, cannot be accepted at face value. The "pure" sub- stances used were mostly biological concentrates the purity of which is highly questionable. If model experiments are to possess the force of proof, they must be performed with syn- thetic compounds the components of which (especially the fatty acid moiety) are exactly known. So far, such experi- ments have not been done on a comprehensive scale. 4. Morrison, R. W., and Hack, M. H.: Am. J. Path., 25:597, 1949; Pearse, A. G. Everson: J. Clin. Path., 2:81, 1949. 5. Altmann, R.: Die Elementarorganismen und ihre Beziehungen zu den Zellen (Leipzig: Veit & Co., 1890); Mulon, P.: BibHog. anat., 3:208, 1904; Camus, J., and Pagniez, P.: Compt. rend. Soc. de bioL, 59:701, 1905. 6. Smith, J. L., and Muir, W.: J. Path. & Bact, 13:14, 1909. 7. Escher, H. H.: Corresp.-Blatt f. schweiz. Aerzte, 49:1609, 1919; Kauf- mann, C, and Lehmann, E.: Zentralbl. f. allg. Path. u. path. Anat., 37:145, 1926; Kaufmann, C, and Lehmann, E.: Virchows Arch. f. path. Anat., 261:623, 1926; Kaufmann, C, and Lehmann, E.: ibid., 270:360, 1928; Baker, J. R.: Quart. J. Micr. Sc, 88:463, 1947. 8. Cain, A. J.: Quart. J. Micr. Sc, 89:429, 1948. Organic Substances 95 HiSTOCHEMICAL METHODS FOR LiPIDS Fixation.— In some cases (especially when true chemical reactions are employed) it may be preferable to use fresh, unfixed tissues, because fixation may modify the reactive groups. In other cases such modifying effects are used de- liberately ( see p. 101 ) . As a routine fixative, formalin is the best and simplest, es- pecially when about 1 per cent CaCL is added to render phospholipids insoluble ( Baker ).^ Millot and Giberton^^ re- port that prolonged fixation in formalin will lead to a pro- gressive decrease in the amount of total fat and also to an increase in the proportion of free fatty acid. The description of their experiments is not clear enough to permit a judgment as to the validity of their conclusions. Of course, embedding either in celloidin or in paraffin will remove lipids more or less quantitatively unless they are made insoluble first (by special treatments, such as pro- longed chromation). Embedding in carbowax (Blank,^^ Firminger^^ ) , on the other hand, appears to be a usable pro- cedure. For most purposes it is simplest to use frozen sec- tions. They should not be exposed to concentrations of alco- hol higher than 70 per cent, and even this may cause the solution of the finest droplets. The actual procedures of demonstration will be divided into two groups— physical and chemical. Physical Methods A) Staining with oil-soluble dyes.— There is a large num- ber of oil-soluble dyes known, many of which are suitable for 9. Baker, J. R.: Quart. J. Micr. Sc, 87:441, 1946. 10. Millot, J., and Giberton, A.: Compt. rend. Soc. de biol., 97:1674, 1927. 11. Blank, H.: J. Invest. Dermat., 12:95, 1949. 12. Firminger, H. J.: Stain Technol, 25:121, 1950. 96 Microscopic Histochemistry histological purposes (Sudan III and IV; Sudan black B;^^* ^* Oil red O;^' Blue B.Z.L.;^^ Nile blue; etc.). With the excep- tion of Sudan black B, which is an excellent stain for phos- phatides, all the dyes mentioned stain triglycerides and fatty acids in a much darker shade than phospholipids. The latter may actually remain practically unstained. According to Gerard,^^ Sudan black B can be used to differentiate between petrolatum and animal lipids. The former will stain in a clear violet, the latter blue-black. Oil dyes (with the exception of Nile blue) are usually made up in alcohol ( 50-70 per cent ) or some similar solvent, such as ethylene or propylene glycoP^' ^^ or pyridine.^^ Solu- tions in 70 per cent alcohol stain fast, but they may remove some of the lipid. Lower concentrations of alcohol are safer, but their solvent power for the dyes is lower, and optimal staining time is greatly prolonged. Solutions in glycols are reported to be safe and fast in action; ^^ however, in view of the known solvent power of glycols (especially of propylene glycol) for many water-insoluble organic substances, the method should be re-examined critically. Sixty per cent iso- propyl alcohol -^ and 50 per cent diacetin-^ have also been recommended as solvents. The writer finds that one of the good solvents for oil dyes is 60 per cent triethylphosphate. It is entirely harmless to lipids, and the staining power of the solution is not much inferior to that of alcohohc solutions. In addition, the loW volatihty of triethylphosphate prevents the precipitation of 13. Lison, L.: Compt. rend. Soc. de biol., 115:202, 1934. 14. Lison, L., and Dagnelie, J.: Bull, d'histol. appliq. a la physiol., 12:85, 1935. 15. Lillie, R. D.: Stain Technol., 19:55, 1944. 16. Gerard, P.: Bull, d'histol. appliq. a la physiol., 12:92, 1935. 17. Hartman, T. L.: Stain Technol., 15:23, 1940. 18. Chiffelle, T. L., and Putt, F. A.: Stain Technol., 26:51, 1951. 19. Proescher, F.: Stain Technol., 2:60, 1927. 20. LilHe, R. D., and Ashbum, L. L.: Arch. Path., 36:432, 1943. 21. Gross, W.: Ztschr. f. wissensch. Mikr., 47:64, 1930. Organic Substances 97 the dye caused by solvent evaporation, an occurrence often experienced with alcohoHc solutions. Method Carry frozen sections in 50 per cent alcohol for a few min- utes. Stain in a saturated and filtered solution of any of the dyes mentioned in 70 per cent alcohol or 60 per cent triethyl- phosphate for 5-20 minutes. Differentiate in 50 per cent al- cohol for about 1 minute. Counterstain as desired. Mount in glycerin-gelatin or some similar medium. Pure kerasin stains intensely with Sudan III, while the kerasin-protein complex of Gaucher cells does not. Franco and Wolman-^ find that boiling the sections at pH 4 for 30-60 seconds will break the complex and make the cells intensely sudanophilic. Nile blue is different from the other dyes in several re- spects. First of all, it is water-soluble; second, it has two com- ponents: a blue oxazin and a red oxazone dye;^^"^^ third, be- sides being an oil dye, it is also a regular basic dye, staining nuclei blue and mucin metachromatically pink.^^ The value of Nile blue has been a subject of much debate ever since its introduction into histological technique by Lorrain Smith.^^ It appears, however, that the investigations of Cain-^ have settled the argument once and for all. Tri- glycerids, whether saturated or not, are colored red by the oxazone, provided that the temperature of staining is not be- low their melting points. All acidic lipids, (including fatty acids and phospholipids ) are stained blue because they bind the oxazin base (pink itself) in the form of blue salts. In mixtures, intermediate shades will be obtained. 22. Franco, S., and Wolman, M.: Schweiz. Ztschr. f. Path. u. Bakt., 10:49, 1947. 23. Smith, J. L.: J. Path. & Bact, 12:1, 1908. 24. Lison, L.: Bull, d'histol. apphq. a la physiol., 12:279, 1935. 25. Cain, A. J.: Quart. J. Micr. Sc, 88:383, 1947. 26. Baker, J. R.: Quart. J. Micr. Sc, 85:1, 1944. 98 Microscopic Histochemistry When using Nile blue, it should be remembered that it is not a specific lipid stain. While the Sudans stain lipids and nothing else, not all structures stained either red or blue by Nile blue are of a lipid nature. B) Fluorescence microscopy .—Msiny lipid substances flu- oresce in ultraviolet light (oxidation products of cholesterol and of various unsaturated fatty acids, vitamin A, ceroid, etc. ) , but only the fleeting green fluorescence of vitamin A is characteristic enough to be useful in histochemistry. For the technical details of the demonstration of vitamin A the reader is referred to the articles of Popper.^"^ A few attempts at the localization of vitamin A by chemical reactions, such as the Carr-Price test, appear to have been gross violations of the fundamental principles of histochem- istry.^* Tissues were dehydrated in alcohol and treated with a solution of antimony trichloride in chloroform. Of course, such a treatment would remove all lipids, vitamin A included, almost quantitatively. It is difiicult to see how the positive reactions reported were obtained. C) Polarization microscopy.— This used to be considered a valuable means for the distinction of doubly refractile cholesterol from other lipids. However, Lison^^ has shown that its results cannot be interpreted in a chemically mean- ingful way. In the case of lipids, birefringence appears to depend largely on factors other than chemical constitution ( such as state of aggregation, supercooling, the nature of the mounting medium, etc. ) even in the case of pure compounds; the behavior of mixtures is unpredictable. The sensitivity of the method is poor; lipid mixtures containing less than 5 per 27. Popper, H.: Proc. Soc. Exper. Biol. & Med., 43:133, 1940; Popper, H.: Arch. Path., 31:766, 1941; Greenberg, R., and Popper, H.: J. Cell. & Comp. Physiol., 18:269, 1941. 28. Bourne, G.: Australian J. Exper. Biol., 13:239, 1935; Joyet-Lavergne, P.: Protoplasma, 28:131, 1937; Joyet-Lavergne, P.: Compt. rend. Soc. de biol., 126:650, 1937; Jones, O. P.: J. Lab. & Clin. Med., 32:700, 1947. 29. Lison, L.: Bull, d'histol. appHq. a la physioL, 10:237, 1933. Organic Substances 99 cent cholesterol show no birefringence.^^ Even a positive re- sult is not necessarily evidence of the presence of choles- terol.^^ Chemical Methods A) For fatty acids and insoluble soaps (Ca).— These sub- stances can be demonstrated by Fischler's method.^" Method Mordant frozen sections in a half-saturated solution of cu- pric acetate for 3-12 hours around 37° C; rinse sections thoroughly in repeated changes of distilled water and stain them in a 0.5 per cent solution of hematoxylin in 50 per cent alcohol for 12-24 hours. Differentiate in Weigert's borax- ferricyanide mixture until the background (nuclei included) is decolorized. The cupric soaps are very resistant to decol- orization and will remain almost black. The reaction is not too specific; red cells, muscle, and cal- careous deposits of any sort are also intensely stained. Two control sections can be used: one from which lipids are re- moved with warm methyl alcohol or chloroform (preferably with about 10 per cent acetic acid added to decompose the soaps) and another one treated with a citrate buffer of pH 4.5-5 to remove other calcium deposits. The use of Nile blue has been mentioned. Faure-Fremiet^^ has reported that unsaturated fatty acids stain metachromat- ically with methyl green and hght green; his findings could not be confirmed. B) For cholesterol and its esters.— a) Schultz's method^* is the histochemical apphcation of SO. Okey, R.: J. Biol. Chem., 156:179, 1944. 31. Yoffey, J. M., and Baxter, J. S.: J. Anat, 81:335, 1947. 32. Fischler, F.: Zentralbl. f. allg. Path. u. path. Anat., 15:913, 1904. 33. Faure-Fremiet, Mayer A., and SchaeflFer, G.: Arch, d'anat. micr., 12:19, 1910. 34. Schultz, A.: Zentralbl. f. allg. Path. u. path. Anat., 35:314, 1924-25; Schultz, A., and Lohr, G.: Zentralbl. f. allg. Path. u. path. Anat., 36:529, 1925. 100 Microscopic Histochemistry the Liebermann-Burchardt test, which is positive with all unsaturated sterols,^^ whether esterified or not. For all prac- tical purposes, it may be considered a specific test for choles- terol and its esters. Method Mordant frozen sections in 2 per cent ferric alum for 24 hours. This step is essential, although its chemical back- ground is not well understood. Rinse sections in distilled water. Mount them on slides, blot them dry. Place a few drops of a mixture of equal parts of glacial acetic acid and concentrated sulfuric acid (caution: cool test tube while mixing the acids!) on the section and cover it with a cover shp. A change of colors from purple-red through dark blue to blue-green will take place within about 1 minute. Only the last shade mentioned is diagnostic for cholesterol. h) The digitonin reaction^*^ is specific for unesterified 3-cis-OH sterols, such as cholesterol, the vitamin D com- pounds, isoandrosterone, etc.; 3-trans-OH compounds (an- drosterone, bile acids) do not react. Method Immerse frozen sections for a few hours in a 0.5 per cent solution of digitonin in 50 per cent alcohol. Wash in 50 per cent alcohol and in water. Mount in glycerin- jelly. Under the polarizing microscope typical groups of fine needle-shaped crystals are seen; they are birefringent. C ) For car otenoids.— One of the carotenoids, vitamin A, has been mentioned under "Fluorescence microscopy." Ca- rotenoid pigments will be considered in the chapter on "Pigments." 35. Sobotka, H.: The chemistry of the sterids (Baltimore: Williams & WilHns, 1938), p. 158. 36. Brunswik, H.: Ztschr. f. wissensch. Mikr., 39:316, 1922; Leulier, A., and Noel, R.: Bull, d'histol. appliq. a la physiol., 3:316, 1926; Leulier, A., and Revol, L.: Bull, d'histol. appliq. a la physiol., 7:241, 1930; Lison, L.: Histochimie animale (Paris: Gauthier-Villars, 1936). Organic Substances 101 D) For unsaturated fatty acicfs.— Unsaturated fats are slowly oxidized even when injected into the tissues; more readily on exposure to atmospheric oxygen; and very rapidly in the presence of various oxidants and catalysts. The prod- ucts of oxidation may include a large variety of com- pounds^^"^^ such as peroxides, epoxides, aldehydes, possibly ketones, hydroxyacids, various fragments and polymers. Some of the latter may be colored and/or insoluble in lipid solvents. The insoluble fraction may exhibit surprising stain- ing reactions, such as acid-fastness and aflBnity to orcein and resorcinol-fuchsin.^^ The chemistry of the oxidative process is poorly understood; the factors which influence the rate of oxidation and the nature of the end-products are obscure. Methods for unsaturated fatty acids are relatively specific for phospholipids and cholesterol esters because of the marked unsaturation of these compounds. a) Osmium tetroxide (osmic acid),— This substance has been employed for the demonstration of triolein and oleic acid since 1895.^^ While oxidizing the ethylenic double bonds, it is reduced to a black substance (probably a mixture of lower oxides ) . It is of some value only if used according to the specifications of Cain^ (well- washed frozen sections fixed briefly in formalin-CaCL and treated with a solution of osmic acid for about 1 hour; avoidance of dichromate and alcohol). Even so, phospholipids containing oleic acid may react feebly or not at all. b ) Dichromate methods.— Fotassiuin dichromate has been 37. Cummings, M. J., and Mattill, H. A.: J. Nutrition, 3:421, 1930-31. 38. Bloor, W. R.: Biochemistry of the fatty acids (New York: Reinhold Pub. Co., 1943). 39. Markley, K. S.: Fatty acids (New York and London: Interscience Publishers, 1947); Hilditch, T. P.: The chemical constitution of natural fats (New York: John Wiley & Sons, 1947). 40. Ralston, A. W.: Fatty acids and their derivatives (New York: John Wiley & Sons, 1948). 41. Hass, G. M.: Arch. Path., 27:15, 1939; Hass, G. M.: ibid., 28:177, 1939. 42. Starke, J.: Arch. f. Physiol., p. 70, 1895. 102 Microscopic Histochemistry used in microscopic technique as a "hardener," especially of brain, for about one hundred years. Its effect on lipids con- sists in their oxidation to a miscellany of compounds, some of which are so insoluble as to be demonstrable by Sudan dyes even in paraffin sections (Ciaccio).*^ Bichromate itself is reduced at the sites, of oxidation to an insoluble compound (Cr203?) which can combine with hematoxylin to form a blue-black lake, quite resistant to differentiation. The rate at which unsaturated lipids chromate varies con- siderably. It appears that the presence of hydrophilic groups enhances chromation; e.g., phosphohpids chromate readily,** and triolein chromates much faster if it contains a small amount of cholesterol than in a pure state.*^ This is not sur- prising, since potassium dichromate is lipid-insoluble. The optimal conditions for the chromation of phospho- lipids have been studied carefully by Baker.^ As a result of his studies, he succeeded in devising a sensitive and specific modification of Lorrain Smith's method.^ The claims of specificity have been confirmed by Cain.*^ Method 1. Fix in formalin containing 1 per cent CaCl2. 2. Transfer pieces directly to a 5 per cent solution of po- tassium dichromate containing 1 per cent of CaCb and keep them in it for 18 hours. 3. Continue mordanting in the same solution for 24 hours at 60° C. 4. Cut frozen sections, preferably after gelatin-embedding. 5. Mordant sections in the dichromate-calcium solution for 1 hour at 60° C. 6. Wash sections in water and stain them in the following, freshly prepared, solution for 5 hours at 37° C: 43. Ciaccio, C: Zentralbl. f. allg. Path. u. path. Anat., 20:385, 1904. 44. Ciaccio, C: Compt. rend. A. anat., 25:87, 1930. 45. Dietrich, A.: Verh. d. deutsch. path. Gesellsch., 14:263, 1910. 46. Cain, A. J.: Quart. J. Micr. Sc, 88:467, 1947. Organic Substances 103 Dissolve 50 mg. of hematoxylin in 48 ml. of distilled water, add 1 ml. of a 1 per cent solution of potassium iodate, and heat it to a boil. Cool, add 1 ml. of acetic acid. 7. Rinse sections, differentiate in a 0.25 per cent solution each of borax and of potassium ferricyanide for 18 hours at 37° C. 8. Wash and mount in glycerin-jelly. Lecithin, cephalin, and sphingomyelin stain in an intense blue-black shade. Only this shade is diagnostic; other un- saturated lipids may stain in shades of brown and gray. A few nonlipid substances, such as mucin and hemoglobin, will stain in the same shade as the phospholipids. As a control. Baker suggests the extraction of an adjacent tissue block, fixed in Bouin's solution, with pyridine at 60° C, and run- ning it through the same schedule. Since this treatment re- moves all hpid, whatever is stained is nonlipid. The differ- ence between the two blocks is due to phosopholipid only. Valade^^ recommends a different scheme of extraction and chromation; its specificity for phopholipids, has not been tested. Chromated tissues can also be embedded in paraffin and stained with Sudan dyes, preferably Sudan black B. This method does not differentiate clearly between phospholipids and other unsaturated fats, although the former (since they chromate faster and are rendered insoluble more readily) will stain more intensely. Quantitative preservation of any type of lipid cannot be expected. c) Alsterberg^^ proposes the demonstration of phospho- lipids by a different principle. The reagent consists of cyan- ogen iodide and silver nitrate ^ ( or chlorate ) . According to the author, choline and colamiae decompose cyanogen iodide and cause the precipitation of a mixture of silver cyanide and iodide which can be visualized in a second step. This method requires further investigation; neither the soundness 47. Valade, P.: BuU. Acad. vet. France, 22:77, 1949. 48. Alsterberg, G.: Ztschr. f. Zellforsch. u. mikr. Anat., 31:364, 1940-41. 104 Microscopic Histochemistry of the chemical background nor the specificity of the results has been tested so far. E) Lipid aldehydes. -hi 1924 Feulgen and Voit,^^ while studying the nucleal reaction in HgCl2-fixed tissues, noticed a widespread staining of elastic membranes and various cy- toplasmic structures by SchifF's reagent, even without pre- ceding acid hydrolysis. Staining could be prevented by pre- treatment with phenylhydrazine. Obviously, the reaction was due to some hitherto undescribed aldehyde to which they gave the name of "plasmal," and the unknown compound from which it is set free by HgCl2 they called "plasmalogen." Further studies by Feulgen and his group^^ gradually suc- ceeded in unraveling the nature of these mysterious sub- stances. Plasmalogen turned out to be a new type of com- pound, an acetalphosphatide. Acetalphosphatides resemble lecithin or cephalin except for the fact that they contain only one molecule of fatty aldehyde per molecule of glycerol, and the linkage between the fatty aldehyde and glycerol is of the cyclic acetal type. alpha-Lecithin H2C— 0— O— C— R HC— O— O— C— Ri HoC — 0 — P — choUne (or colamine) OH O (R and Ri, fatty acid radicals) alpha-Acetalphosphatide H2C— 0\^ :CHR HC— O. HoC — O — P — choline / N^ (or colamine) OH O 49. Feulgen, R., and Voit, K.: Arch. f. d. ges. Physiol., 206:389, 1924. 50. Feulgen, R., and Imhauser, K.: Biochem. Ztschr., 181:30, 1927; Im- hauser, K.: Biochem. Ztschr., 186:360, 1927; Voss, H.: Ztschr. f. mikr.-anat. Forsch., 10:583, 1927; Behrens, M.: Ztschr. f. physiol. Chem., 191:183, 1930; Feulgen, R., and Behrens, M.: Ztschr. f. physiol. Chem., 256:15, 1938; Feulgen, R., and Bersin, T.: Ztschr. f. physiol. Chem., 260:217, 1939; Voss, H.: Ztschr. f. Zellforsch. u. mikr. Anat., 31:43, 1940-41; Bersin, T., Moldtman, H. C, Nafziger, H., Marchand, B., and Leopold, W.: Ztschr. f. physiol. Chem., 269:241, 1941. Organic Substances 105 The acetal linkage is slowly hydrolyzed by acids and oxi- dizing agents but promptly by heavy-metal salts such as HgCl2, with liberation of the corresponding aldehyde. Ac- tually, plasmal has been shown to be a mixture of several long-chained aldehydes, mainly stearic and palmitic, with the admixture of unidentified unsaturated ones. Plasmal will react with all aldehyde reagents such as Schiff's, phenylhydrazine, semicarbazide, naphthoic hydra- zide, etc. Highly colored reaction products (with 2,4-dini- trophenylhydrazine, SchifiF's reagent, or 2-hydroxy-3-naphtho- ic hydrazide followed by azo-coupling^^~^^ ) can be used for its histochemical identification. It should be mentioned, however, that plasmal is not the only lipid aldehyde in the tissues. ^^^^ Substances of aldehy- dic nature are formed in large amounts by the oxidation of unsaturated fatty acids. ^^^^ The presence of such nonplasmal aldehydes can be demonstrated (some of them at specific sites; Chu)^^ even in completely fresh tissues. Their amount increases and their distribution becomes more widespread on storage in either formalin or water. The eflPect is probably due to atmospheric oxygen, since it is especially marked in sections kept in a shallow layer of fluid. Intense reactions are obtained at all sites where cholesterol or phospholipids are present. This is not surprising, in view of the fact that the fatty acids of cholesterol esters^^ and of phopholipids^^ show 51. Camber, B.: Nature, 163:285, 1949. 52. Seligman, A. M., and Ashbel, R.: BuU. New England M. Center, 11:85, 1949. 53. Seligman, A. M., Friedman, O. M., and Herz, J. E.: Endocrinology, 44:584, 1949. 54. Cain, A. J.: Quart. J. Micr. Sc, 90:75, 1949. 55. Danielli, J. F.: Quart. J. Micr. Sc, 90:67, 1949. 56. Hayes, E. R.: Stain Technol., 24:19, 1949. 57. Chu, C. H. U.: J. Nat. Cancer Inst, 10:1344, 1950; Chu, C. H. U.: Anat Rec, 108:723, 1950. 58. Kelsey, F. E., and Longenecker, H. E.: J. Biol. Chem., 139:727, 1941. 59. Page, I. H., and Rudy, H.: Ztschr. f. Physiol. Chem., 205:115, 1932, 106 Microscopic Histochemistry a high degree of unsaturation. On prolonged storage, even depot fat will develop more or less intense aldehyde reac- tions. HgCL has httle effect on the formation of aldehyde from unsaturated fatty acids. Periodic acid, on the other hand, produces large amounts of aldehyde.^^ Plasmalogen as a source of aldehydes is probably much less important quan- titatively than unsaturated fatty acids. The question is vv^hether there may be any nonaldehydic substances in the tissues which could give positive reactions with the reagents mentioned. The first group of compounds to be considered are the ke- tones. Some of them may result form the oxidation of un- saturated fatty acids (especially by the rearrangement of epoxides^^), but there is no positive proof for their existence in animal tissues. On the other hand, ketosteroids are known to be present in a number of tissues. They give no color with Schiff's reagent,^^' ^^ which is fairly specific for aldehydes; but there can be no doubt that they would give good positive reactions with most other carbonyl reagents. Actually, the reaction obtained with phenylhydrazine^* or with 2-hydroxy- 3-naphthoic hydrazide^^ has been assumed to be indicative of the presence of ketosteroids. However, the evidence adduced is of only a circumstantial nature. Dempsey^^ says that, while not one of the physical and chemical features (carbonyl reac- tions, Schultz test, birefringence, fluorescence, solubility in 60. Wolman, H.: Proc. Soc. Exper. Biol. & Med., 75:583, 1950. 61. Foumeau, E., and Tiffeneau, M.: Compt. rend. Acad, sc, 144:662, 1905. 62. Albert, S., and Leblond, C. P.: Endocrinology, 39:386, 1946. 63. Oster, K. A., and Oster, J. G.: J. Pharmacol. & Exper. Therap., 87:306, 1946. 64. Bennett, H. S.: Proc. Soc. Exper. Biol. & Med., 42:786, 1939; Ben- nett, H. S.: Am. J. Anat., 67:151, 1940; Dempsey, E. W., and Bassett, D. L.: Endocrinology, 33:384, 1943; Deane, H. W., and McKibbin, J. M.: Endocrinology, 38:385, 1946; Deane, H. W., and Creep, R. O.: Am. J. Anat., 79:117, 1946; Creep, R. O., and Deane, H. W.: Anat. Rec, 97:416, 1947; Creep, R. O., and Deane, H. W.: Ann. New York Acad. Sc, 50:596, 1949; Applegarth, A.: Endocrinology, 44:197, 1949. 65. Dempsey, E. W.: Recent Prog. Hormone Research, 3:127, 1948. Organic Substances 107 acetone, sudanophilia ) of the lipid present at sites of keto- steroid production is really specific for ketosteroids, no other known type of substance displays the entire battery of reac- tions. Chemical considerations are decidedly against this as- sumption. In fresh tissues such reactions are entirely nega- tive; on mere standing or after treatment with mild oxidants they become increasingly positive— an indication that what- ever gives the reaction must be the result of an oxidative process. Ketosteroids would be expected to react directly, even in fresh tissues. The distribution pattern of carbonyl compounds as seen with the relatively aldehyde-specific Schiff's reagent is invariably either identical with, or ex- tremely similar to, that obtained by the use of *1<:etone rea- gents," both in the endocrine organs and elsewhere^^ ( brain, atherosclerotic plaques, necrotic tumors, etc.). The minor discrepancies occasionally observed are theoretically readily explainable by differences in the lipid solubility of the rea- gents and possibly by differences in reactivity of the several aldehydic substances present (steric factors?). Similar dif- ferences in reactivity have been reported in the cases of such relatively homogeneous groups of substances as desoxyribose- nucleic acids^^ and mucopolysaccharides,^^ even though the reactive groups within each class are the same. In a series of recent papers^^ Seligman and his group re- ported their studies on the carbonyl groups of formalin-fixed nervous tissue, adrenals, and testes, demonstrated by the 2-hydroxy-3-naphthoic acid hydrazide-Blue B Salt method. They came to the conclusion that the reacting carbonyl groups, "unmasked" by formalin, are of a ketonic rather than 66. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 51:133, 1942. 67. Lessler, M. A., and Kopac, M. J.: Anat. Rec, 108:578, 1950. 68. Monne, L., and Slautterback, D. B.: Exper. Cell Research, 1:477, 1950. 69. Ashbel, R., and Seligman, A. M.: Endocrinology, 44:565, 1949; Seligman, A. M., and Ashbel, R.: Cancer, 4:579, 1951; Seligman, A. M., and Ashbel, R.: Endocrinology, 49:110, 1951; Ashbel, R., Cohen, R. B., and Seligman, A. M.: Endocrinology, 49:265, 1951; Rabinovici, N.: Endocri- nology, 49:579, 1951. 108 Microscopic Histochemistry of an aldehydic character. The main points of their argu- ments are as follows: 1. It can be shown that the reaction is not due to plasmal, since treatment of unfixed tissue by mercuric chloride causes only a weak reaction. 2. Phenylhydrazones of aldehydes react with diazonium salts to form intensely colored formazans, while ketonic phenylhydrazones, in the absence of replaceable hydrogen, cannot couple with diazonium salts. This formazan reaction is negative with nerve tissue, adrenal lipid, etc. 3. Aniline and sulfanilic acid, according to Oster and Mu- linos^^ and Boscott and Mandl,"^^ selectively block aldehydes. The carbonyl groups in question are not blocked under the conditions specified. 4. The Angeli-Rimini test*^^ for aldehydes is negative with the tissues. The problem of hpid aldehydes versus ketones has been subjected to an experimental study by Gomori.'^^ His findings are as follows: 1. It can be shown in model experiments that fluorescence is a property of auto-oxidation products rather than of ketosteroids. 2. Positive carbonyl reactions can be obtained after for- malin-free fixatives, such as a saturated solution of picric acid in 50 per cent alcohol. Therefore, "unmasking" by formalin is a superfluous assumption. 3. Seligman's criteria for the nonaldehydic nature of lipid carbonyl groups are not valid. Formazans do not form from all aldehydes; for instance, periodate-treated glycogen is en- tirely negative. The results of blocking experiments permit 70. Oster, K. A., and Mulinos, M. G.: J. Pharmacol. & Exper. Therap., 80:132, 1944. 71. Boscott, R. J., and Mandl, A. M.: J. Endocrinol., 6:132, 1949. 72. Angeli, A., and Angelico, F.: Gazz. chim. ital., 34-1:50, 1904; Rimini, E.: Atti reale Accad. Lincei, ser. 5, cl. di sc. fis., mat. e nat., 17-11:360, 1908. 73. Gomori, G.: J. Lab. & CHn. Med., 39:649, 1952. Organic Substances 109 no conclusions whatsoever as to the aldehydic or ketonic na- ture of carbonyl groups. The Angeh-Rimini reaction is known to be negative with a number of aldehydes even in test-tube experiments;"^^ it is invariably negative histochemically with unquestionably aldehydic substances. 4. Estrone, testosterone, and desoxycorticosterone give negative reactions with Seligman and Ashbel's reagent under the conditions of the histochemical test. 5. The carbonyl reactions in tPie tissues are so intense that they must be due to a substance which constitutes a substan- tial percentage of the lipid material. On the basis of these findings, Gomori comes to the con- clusion that the "battery of reactions" is due not to keto- steroids but to auto-oxidation products of unsaturated fatty acid esters of cholesterol. In summary, it may be said that, for the time being, we do not possess a single reaction capable of identifying ketones, let alone ketosteroids, in the midst of a large bulk of hpid aldehydes. This view is shared by the vast majority of work- ers who have studied the problem critically (Gomori;^^ Aboim;^^ Albert and Leblond;^- Claesson and Hillarp;^^ Bos- cott, Mandl, Danielli, and Shoppee;^^ Boscott and Mandl;^^ and Sayers^^ ) . The invariable association of ketosteroids with aldehydes ( but not vice versa ) simply means that the steroid hormones are handled metabolically, at least up to a certain point, very much like cholesterol. In addition to aldehydes and ketones, a number of other compounds have been asserted to give a positive test with Schiff's reagent. According to Lison,'^^ ethylenic double bonds 74. Angeli, A., and Angelico, F.: Gazz. chim. ital., 33-11:245, 1903; Angeli, A., and Marchetti, G.: Atti reale Accad. Lincei, ser. 5, cl. di sc. fis., mat. e nat., 17-11:360, 1908. 75. Aboim, A. Nunes: Bull. Soc. port. sc. nat, 14:119, 1943. 76. Claesson, L., and Hillarp, N. A.: Acta anat, 3:109, 1947. 77. Boscott, R. J., Mandl, A. M., DanielH, J. F., and Shoppee, C. V^.: Nature, 162:572, 1948. 78. Sayers, G.: Physiol. Rev., 30:241, 1950. 79. Lison, L.: Bull, d'histol. appliq. a la physiol., 9:177, 1932. 110 Microscopic Histochemistry (such as in oleic and cinnamic acids) may react. However, this is denied by Verne^^ and Oster and Oster,^^ who find that a positive reaction is always due to contamination by products of oxidation. Gerard^^ and Verne^^ made the observation that the same lipid structures which stain positively with Schiff's reagent also show strong oxidative properties, e.g., they recolorize leuco-bases and stain intensely blue with the nadi-reagent, al- though the intensity of the aldehyde and oxidative reactions need not run parallel. This interesting observation may be explained by the invariable compresence of aldehydes and peroxides; furthermore, there is a possibility that part of the color obtained in the Schiff reaction is due to a pseudo-reac- tion: the oxidative recolorization of the reagent. However, differentiation should be easy; the true aldehyde color is ex- tremely resistant to acids, whereas regenerated fuchsin is readily decolorized by them. Histochemical methods for lipid aldehydes 1 ) For pre-formed aldehydes.— The best method is to use fresh, unfixed frozen sections. However, a few hours' fixation in formalin, followed by thorough washing in repeated changes of distilled water, is permissible, although the stain- ing of some structures (e.g., myelin sheaths) may be greatly weakened. ^^ Freshly boiled distilled water should be used to make up the formalin solution and to rinse the sections. Stain in Schiff's reagent ( straight or diluted with an equal volume of distilled water) for about 30-60 minutes. Rinse sections in a 1-3 per cent solution of Na bisulfite for a few minutes. Rinse in repeated changes of tap water. Counter- stain with hematoxylin. Mount either in glycerin-jelly or, after dehydration, in balsam. The fatty aldehyde-Schiff com- pound is lipid-insoluble. 80. Veme, J.: Ann. de physiol., 5:245, 1929. 81. Gerard, P.: Bull, d'histol. appliq. a la physiol., 12:274, 1935. 82. Veme, J.: Compt. rend. Soc. de biol, 133:75, 1940; Verne, J.: Bull, d'histol. appHq. a la physiol., 13:433, 1936. 83. Veme, J.: Bull, d'histol. appHq. a la physiol., 14:269, 1937. Organic Substances 111 2) For acetalphosphatides.^^—\Jse unfixed or formalin- fixed tissues. Block pre-formed aldehydes by treating the sections with a 2 per cent solution of either phenylhydrazine or hydroxylamine-HCl in an acetate buffer of pH 4.5-5.5 for 6-12 hours at room temperature. Rinse thoroughly in distilled water. Treat sections with a 2 per cent solution of HgCU for 15 minutes. Rinse in distilled water. Stain, etc., as under 1, above. Danielli^^ finds that hydrolysis in 0.1 N HCl for 15 minutes at room temperature is preferable to treatment with HgCk because the latter may have oxidative side effects. 3) For aldehydes secondary to the oxidation of unsatu- rated fatty acids. ^^—Tresit sections with HgCl2 or 0.1 N HCl; block aldehydes with one of the reagents mentioned under 2, above; wash and expose them to air for several days (for instance, in a flat dish containing a shallow layer of water ) . Stain, etc., as under 1, above. It is always advisable to extract a control section with sev- eral changes of alcohol-ether or acetone and to carry it through the entire procedure. The fat-free blank will help to rule out reactions given by nonlipid substances. C. PROTEINS, AMINO ACIDS, AND PRODUCTS OF PROTEIN METABOLISM The demonstration of the presence of protein substances as such is of relatively minor importance in histochemistry because proteins are ubiquitous components of all tissues. Occasionally, however, it may be necessary to ascertain the protein or nonprotein nature of certain structures, such as granulations. The identification of the individual amino acids is of much more interest but possible only ta a few special instances. Proteins The methods for the demonstration of proteins can be divided into two groups: precipitation and digestion tests. 1. The precipitation tests are based on the fact that many proteins retain their aflBnity for protein precipitants even 112 Microscopic Histochemistry though they have been precipitated previously by histologi- cal fixatives. a) The ferrocyanide reaction of Hartig-Zacharias} Method Treat the section for 10 minutes with an acidified solution of potassium ferrocyanide (1-5 per cent in dilute hydro- chloric acid; concentrations not important ) ; wash thoroughly and flood with a dilute solution of ferric chloride. Proteins stained blue. Ferric compounds (hemosiderin) will be blue before the application of ferric chloride. h) The tannin-ferric method of Salazar,^— Method \ Treat the section for 15-20 minutes with an acidified solu- tion of tannin ( 5-10 per cent tannin in 5-10 per cent acetic acid; concentrations not important); wash thoroughly and flood with a dilute solution of ferric chloride. Some protein structures, secretion granules, etc., stain gray-black. 2. The digestion tests, once widely used, yield very little information of a chemical nature, although they do differen- tiate between various protein substances. Pepsin digests col- lagen and reticulum; trypsin, elastic fibers. For details the interested reader is referred to the review of the topic by Spalteholz and others.^ This topic should be reinvestigated with the aid of purified proteolytic enzymes which have become available recently. Amino Acm Components of Proteins Of the reactions described for the histochemical identifica- tion of amino acids, only two are reagents for amino acids proper (ninhydrin and alloxan). The others are specific 1. Zacharias, E.: Bot. Ztschr., 41:208, 1883. 2. Salazar, A. L.: Compt. rend. Soc. de biol., 83:1655, 1920. 3. Verdauung, kiinstliche. In Enzyld. d. mikr. Technik (3d ed.; Berlin and Vienna: Urban & Schwarzenberg, 1927), 3:2220. Organic Substances 113 t> either for the aromatic nucleus or for phenoHc functions or for the guanidine grouping. The shades given by any of the reagents to be mentioned cannot be compared for intensity and sharpness with those of the better color reactions in histochemistry. Some of the reagents will react only with unfixed ( or incompletely fixed ) proteins and, even so, give shades so pale and indistinct as to make the test virtually worthless. With some of the other tests, the interpretation of the results is vague. With the ex- ception of the Millon and Sakaguchi reactions, which do give valuable information, the methods in this group will be men- tioned for the record rather than because of their usefulness. 1. Millon's reaction^ is one of the oldest tests for protein substances. Actually, the reagent is specific for phenols and, in the case of amino acids, for tyrosine. Of the several modifications published,^ Pollister's appears to be the most reliable. Method Any good fixation is suitable. Incubate sections at 30°- 37° C. in a solution containing 5 per cent mercuric acetate and 15 per cent trichloroacetic acid. After 5-10 minutes add about one-tenth volume of a 1 per cent solution of sodium nitrite and incubate sections for another 25 minutes. Rinse sections directly in 70 per cent alcohol, dehydrate, and mount. Tyrosine-containing proteins are stained in a shade of pink to brick-red. The shade is rather transparent, and it is advisable, for better visibility, to use sections not thinner than 10 jx. 2. The Sakaguchi test^ is specific for derivatives of guani- dine in which at least one hydrogen in each of the amino 4. Millon, E.: Compt. rend. Acad, sc, 28:40, 1849. 5. Bensley, R. R., and Gersh, L: Anat. Rec, 57:217, 1933; Serra, J. A., and Queiroz Lopes, A.: Port, acta biol., 1:51, 1945; Pollister, A. W.: Rev. d'hematol., 5:527, 1950. 6. Sakaguchi, S.: J. Biochem. (Japan), 5:25, 1925. 114 Microscopic Histochemistry groups is unsubstituted. Arginine is the only compound found in tissues which will give a positive reaction. Of the several modifications published/ Baker's is rela- tively the simplest and most reliable. Method Use celloidin sections or paraflBn sections protected with collodion. The thickness of the sections should be at least 10 ;x. Mix rapidly 2 ml. of 1 per cent NaOH, 2 drops of a 1 per cent solution of a-napthol in 70 per cent alcohol, 4 drops of a 1 per cent solution of Na hypochlorite ( a dilution of Clorox or of some similar household bleaching agent with 7-10 parts of water ) . Pour mixture on the slide immediately and leave it on for 15 minutes. Drain fluid off the slide, blot it, and immerse it in a mixture of 3 parts of pyridine and 1 part of chloroform. Mount in the same medium. Arginine-containing proteins stain in an orange-red shade. The full intensity of the color is permanent for a few hours only. Clearing in xylene and mounting in balsam is permissible, but fading of the color will be even faster. 3. The xanthoprotein reaction^ is specific for aromatic rings, which are nitrated by cold fuming nitric acid to yield yellow dyes. 4. Diazonium salts will couple with phenols and hetero- cyclic rings^ (tyrosine, histidine, proline, etc.) to yield yel- lowish or brownish azo dyes of low color intensity. 5. Syrupy phosphoric acid, especially with a trace of some 7. Serra, J. A.: Port, acta biol., 1:1, 1944; Serra, J. A.: Naturwiss., 32:46, 1944; Thomas, L. E.: J. Cell. & Comp. Physiol., 28:145, 1946; Baker, J. R.: Quart. J. Micr. So., 88:115, 1947; Thomas, L. E.: Stain Technol., 25:143, 1950; Warren, T. N., and McManus, J. F. A.: J. Nat. Cancer Inst., 12:223, 1951. 8. Raspail, F. V.: Nouveau systeme de chimie organique (3d ed.; Brussels: 1840), 1:161. 9. Pauly, H.: Ztschr. f. physiol. Chem., 42:508, 1904; Brunswik, H.: Ztschr. f. physiol. Chem., 127:268, 1923; Berg, W.: Arch. f. d. ges. Physiol., 199:656, 1923. Organic Substances 115 aldehyde (vanillin, furfural, etc.) added, gives with trypto- phane at 60°— 65° C. a fairly intense but unstable purple-red color (Romieu's reaction^^). 6. Ehrlich's p-dimethylaminobenzaldehyde reagent^^* ^^ condenses with a number of phenols, amines, pyrrole and indole derivatives^^ to give reddish and bluish dyes of un- known constitution. 7. a-amino acids give, when boiled with a 0.2-0.5 per cent solution of ninhydrin ( triketohydrindene hydrate), a blue coloration.^* This method works only with tissues not too thor- oughly fixed and unembedded. The color of the reaction has a great tendency to diffuse. 8. Alloxan in aqueous or alcoholic solutions gives a red color with a-amino acids. ^^ The shade is so pale as to make the reaction worthless for histochemical purposes. 9. Voss^^ recommends the use of o-diacetylbenzene (10 per cent in 70 per cent alcohol ) as a reagent for amino acids and proteins ( bluish-red color ) . Reactions 6-9, being specific for NH2 groups, will be nega- tive after formalin fixation.^^' ^'^ An entirely new avenue for the histochemical investigation of protein substances is proposed by Danielli.^^ He recom- mends the application of a number of important condensa- tion reactions of organic chemistry so modified as to yield 10. Romieu, M.: Compt. rend. Acad, sc, 180:875, 1925; Blanchetiere, A.: Compt. rend. Acad, sc, 180:2072, 1925; Blanchetiere, A., and Romieu, M.: Compt. rend. Soc. de bioL, 107:1127, 1931. 11. Ehrlich, P.: Med. V^chnschr., p. 151, 1901. 12. Lison, L.: Histochimie animale (Paris, 1936), p. 160. 13. Rohde, E.: Ztschr. f. physiol. Chem., 44:161, 1905; Fleig, M. C: Bull. soc. chim. France, 4th ser., 3:1038, 1908. 14. Ruhemann, S.: J. Chem. Soc, Tr., 97:2025, 1910; Berg, W.; Klin. Wchnschr., 2:1757, 1923. 15. Krasser, F.: Monatschr. f. Chem., 7:673, 1886; Hurtley, V^. H., and Wootton, V^. O.: J. Chem. Soc, Tr., 99:288, 1911; Giroud, A.: Protoplasma, 7:72, 1929. 16. Voss, H.: Ztschr. f. mikr.-anat. Forsch., 49:51, 1940. 17. Duliere, W. L.: Biochem. J., 30:770, 1936. 18. DanielH, J. F.: Cold Spring Harbor Symp. Quant. Biol., 14:32, 1950. 116 Microscopic Histochemistry colored end-products. Only two examples will be quoted to give an idea of the nature of the new tests: 1. Amino groups of the protein are condensed with amino- benzaldehyde; the aromatic amino group of the resulting SchiflF's base can be diazotized and coupled with napthol. 2. Phenolic groups of the protein are condensed with dini- trochlorobenzene;^^ the nitro groups are reduced to amino groups and the latter coupled with a diazonium salt. In both cases the final product will be an intensely colored azo dye. Theoretically, this approach is sound and can be expected to give excellent results. Antigens The microscopic localization of antigens in tissue sections is one of the newest and most promising fields of histochem- istry. The histochemical localization of antigens depends on the fact that it is possible to introduce various chemical groups into proteins without greatly changing their immunological properties.^^^^ If an antibody is coupled, e.g., with tetrazo- tized benzidine-R-salt conjugate, a red azoprotein is obtained which will be selectively adsorbed to the original antigen. This method has been used for in vitro aggulutination stud- ies.^^ Unfortunately, the shade of the azoprotein cannot be made dark enough to produce a satisfactory color contrast in combination with the appropriate antigen in tissue sections,. In 1941 Coons and associates developed a much more sen- sitive method by coupling antibodies with anthryl isocy- 19. Bost, R. W., and Nicholson, F.: J. Am. Chem. Soc, 57:2368, 1935. 20. Landsteiner, K., and Lampl, H.: Biochem. Ztschr., 86:343, 1918. 21. Heidelberger, M., and Kendall, F. E.: Science, 72:252, 1930. 22. Heidelberger, M., Kendall, F. E., and Soo Hoo, C. M.: J. Exper. Med., 58:137, 1933. 23. Reiner, L.: Science, 72:483, 1930; Heidelberger, M., and Kendall, F. E.: J. Exper. Med., 59:579, 1934. 24. Marrack, J.: Nature, 133:292, 1934. Organic Substances 117 anate^^ and later with fluorescein isocyanate.^^ In this way intensely fluorescent carbamido-proteins can be produced, which, if adsorbed on morphological structures, are easily visible under the microscope. The method is rather complicated. Fluorescein isocyanate itself is not available on the market but must be synthesized by a laborious procedure; the conjugated antibody must be carefully purified to eliminate its nonspecific components. However, the method has already produced most valuable results in the localization of several antigens^^ ( among others, ACTH)2Mn the tissues. Urea Urea is precipitated by a solution of xanthydrol in acetic acid as dixanthydryl urea, which forms beautiful rosettes of small needle-shaped crystals. The crystals are insoluble in water and in most organic solvents. This reaction is the only really specific test for urea.^^ Un- fortunately, its usefulness in histochemistry is limited because ( 1 ) the solvent severely damages histological structure and (2) the reaction is relatively slow, permitting considerable dilution and displacement of urea before precipitation occurs. For this reason the sensitivity of the reaction is low, and the localization is only approximate. The failure of the test to 25. Coons, A. H., Creech, H. J., and Jones, R. N.: Proc. Soc. Exper. Biol. & Med., 47:200, 1941. 26. Coons, A. H., Creech, H. J., Jones, R. N., and Berliner, E.: J. Im- munol., 45:159, 1942. 27. Coons, A. H., and Kaplan, M. H.: J. Exper. Med., 91:1, 1950; Kaplan, M. H., and Coons, A. H.: J. Exper. Med., 91:15, 1950; Coons, A. H., Snyder, J. C, Cheever, F. S., and Murray, E. S.: J. Exper. Med., 91:31, 1950; Hill, A. G. S., Deane, H. W., and Coons, A. H.: J. Exper. Med., 92:35, 1950; Coons, A. H., Kaplan, M. H., and Deane, H. W.: J. Nat. Cancer Inst., 10:1344, 1950. 28. Marshall, J. M.: J. Exper. Med., 94:21, 1951. 29. Policard, A.: Compt. rend. Soc. de biol., 78:32, 1915. 118 Microscopic Histochemistry demonstrate urea at relatively low concentrations led FeyeP** to believe that urea is not filtered through the glomeruli. Probably the best method is that of Oliver.^^ Method Fix small pieces of tissue in a mixture of 6 g. of xanthydrol, 35 ml. of alcohol, and 65 ml. of acetic acid for 6-12 hours. Heat the mixture gently and filter it before use. Dehydrate in alcohols, embed in paraffin. Counterstaining of the sections is permissible. The crystals are best recognized by their bire- fringence. Uric Acm Uric acid occurs in the tissues in the form of acid sodium urate, a substance of a very typical crystalline structure, spar- ingly soluble in water, insoluble in alcohol. Although some- what soluble in dilute alkalis, it is practically insoluble in ammonia. Treatment with ammonia converts it into the cor- responding ammonium salt, the crystalline structure of which is different from that of sodium urate. Heavy-metal salts of uric acid are quite insoluble and can be converted into colored compounds ;^^ however, such reac- tions lack specificity ( phosphates and carbonates behave the same way). This applies also to most of the silver methods recommended for its demonstration.^^ However, uric acid is one of the few argentaffin substances (see p. 58) occurring in the animal body, and this property can be utilized for its identification. This has been done, although under nonopti- mal conditions, by Gersh.^* The following method gives very sharp and selective pictures. 30. Feyel, P.: Compt. rend. Soc. de biol., 114:1155, 1933. 31. OUver, J.: J. Exper. Med., 33:177, 1921. 32. Saint-Hilaire, C: Ztschr. f. physiol. Chem., 26:102, 1898. 33. Courmont, J., and Andre, C: J. de phys. et de path, gen., 7:255, 1905; Tomita, W.: Tr. Jap. Path. Soc., 17:190, 1927. 34. Gersh, I.: Anat. Rec, 58:349, 1934. Organic Substances 119 Method Fix tissues in 95-100 per cent alcohol When mounting the sections on sHdes, float them on water for only a few seconds. Carry slides through xylene and alcohols. From last alco- hol, transfer them directly into methenamine-silver solution (see p. 60) buffered to pH zi= 9 and prewarmed to 37° C. Keep them in the incubator for about 30 minutes or until urate shows up in a black shade. Rinse slides, remove unre- acted silver with a dilute solution of Na thiosulfate. Gold toning is optional. Counterstain as desired. Under low and medium powers the crystalhne structure of the deposits appears to be preserved to perfection; under high power the slender needles are resolved into rows of fine black granules. The only source of error is the presence of calcifications; they will be disturbing only if present in large masses. Other- wise, silver phosphate and carbonate are relatively easily soluble in methenamine and will be washed out of the tissue during incubation. In the case of massive deposits, some sil- ver phosphate may remain undissolved. However, since it is not reduced to metallic silver ( provided that the slide is not exposed to strong light ) , the rinse in thiosulfate will remove it. In case of doubt, treat section for 1 or 2 minutes with a 0.2-0.5 per cent solution of nitric or hydrochloric acid in ab- solute alcohol before transferring it to the silver solution ( the acid must be washed off first with absolute alcohol). This treatment will completely eliminate calcifications. Urate de- posits will remain undissolved, but their crystal structure wiU be somewhat distorted. On the other hand, m'ate depos- its are removed readily by a dilute solution of lithium car- bonate, which will leave calcifications intact. Confusion with melanin and premelanin, which are also argentaffin and will blacken under the same conditions, is not likely. 120 Microscopic Histochemistry D. PROSTHETIC GROUPS Phenolic Substances, Especially Polyphenols The list of phenolic substances foiuid in animal tissues in- cludes tyrosine, adrenalin, certain propigments,^ the entero- chromaffin substance, the phenolic ketosteroids, and a few unidentified substances in lower species (oysters, cephalo- pods,^ toads,^ etc.^). Tyrosine was dealt with in the section on amino acids (p. 113). Phenolic ketosteroids have never been investigated histochemically except in a very uncritical way by Seeger.^ This section will be devoted especially to adrenalin and the enterochromaflBn substance. Before going into the specific histochemical properties of these substances, a few words must be said about color reactions for phenols. There are a number of characteristic color reactions for phenols, some of which can be utilized for their histochemical identification. They will be enumerated in the order of their importance. 1) The azo-coupling reaction.— At an alkaline reaction, phenols will couple with diazonium salts to form intensely colored, water-insoluble azo dyes.^ The shade of the dye de- pends on both the phenolic and the diazoic components; as a rule, a relatively low molecular weight of the components expresses itself in yellowish or orange shades; with increas- ing complexity of the molecule the shade will shift from orange to red, to purple, and finally to blue and black. How- ever, besides molecular weight, structure is also an important determinant of shade; azo dyes produced from a- and /3- 1. Lison, L.: Compt. rend. Soc. de biol., 106:41, 1931. 2. Lison, L.: Histochimie animale, pp. 158-59 (Paris, 1936). 3. Shipley, P. G., and Wislocki, G. B.: Contrib. Embryol., 3:73, 1915; Lison, L.: Compt. rend. Soc. de biol., 111:657, 1932. 4. Lison, L.: Compt. rend. Soc. de biol., 112:1237, 1933. 5. Seeger, P. G.: Ztschr. f. mikr.-anat. Forsch., 46:153, 1939. 6. Saunders, K. H.: The aromatic diazo-compomids and their technical applications (London: Edward Arnold & Co., 1936); Pratt, L. S.: The chemistry and physics of organic pigments (New York: John Wiley & Sons, 1947). Organic Substances 121 naphthol, respectively, differ from one another sharply in color. The reaction is highly specific; under conditions of moderate alkalinity all phenols except those substituted in the para and both ortho positions will couple; the only other substances found in the tissues which will give a similar reac- tion are heterocyclic compounds (histidine; probably also proline ) . According to Pauly J tryptophane does not couple; according to Danielli,^ it does. In the writer's experiments no color was produced with tryptophane with five different dia- zonium salts. In general, the color of azo dyes resulting from amino acids is yellowish or pale orange-brownish. 2) The indophenol reaction.— This consists in the forma- tion of bluish indophenol dyes when phenols unsubstituted in the para position are oxidized in the presence of aromatic amines. The reaction can be performed in a variety of ways. One of the simplest is Gibbs's method,^ in which the amine and the oxidant are combined into a single substance (2,6- dichloro- or dibromo-quinonechloroimide ) ; it also possesses the advantage of producing somewhat darker shades than the other variants. 3) The ferric chloride reaction.— Yhenoh give character- istic color reactions (green, blue, purple, depending on the nature of the phenol ) with a dilute solution of ferric chloride. The shades produced in tissue sections are much too pale to be useful. In addition to giving the reactions mentioned, diphenols are strong reducing agents. They will reduce an ammoniacal silver nitrate solution to metallic silver ("argentaffin reac- tion" ).^^' ^^ By dichromates and ipdates they are oxidized to quinones and other brownish, more or less insoluble, prod- ucts of poorly known constitution which ultimately precipi- 7. Pauly, H.: Ztschr. f. physiol. Chem., 42:508, 1904. 8. Danielli, J. F.: Cold Spring Harbor Symp. Quant. Biol., 14:32, 1950. 9. Gibbs, H. D.: J. Biol. Chem., 72:649, 1927. 10. Cordier, R.: Bull, d'histol. appliq. a la physiol., 4:161, 1927. 11. Hamperl, H.: Virchows Arch. f. path. Anat., 286:811, 1932. 122 Microscopic Histochemistry tate ("chromaffin reaction"); ^2' ^^ in the case of dichromates the precipitate will contain brownish chromium dioxide.^* In the case of a- and p-diphenols, which are powerful reducers, these reactions are prompt; the reducing power of m-di- phenols is much weaker, and the reactions take place over a period of hours rather than minutes. The corresponding aminophenols and diamines behave in an analogous way, but, since they are not known to occur in tissues, unless ad- ministered parenterally ( arsphenamine ) ,^^ their histochem- ical significance is very limited. On account of the effects of fixation, the situation in tissues is somewhat more compHcated than in the test tube. Non- protein phenols appear to be soluble in alcohol; at least, they are not preserved by alcohohc fixatives. Formalin, on the other hand, undergoes, condensation reactions (of the bake- lite type) with phenols, resulting in the formation of highly insoluble resin-like substances^^ of varying molecular sizes. The identifying reactions of the condensation products may differ sharply from those of the parent phenols. This is easy to understand, since the hydroxymethyl groups entering the benzene ring have a tendency to occupy the coupling ( ortho and para) positions; folding of the large polymer molecules formed later in the course of the reaction may cause instances of unpredictable steric hindrance or the approximation of reactive groups. The changes in the chemical properties of phenols by form- aldehyde can be studied conveniently by the use of Cou- jard sildes.^^ The phenols in question are dissolved in serum or dilute gelatin, and marks are made on slides with the solu- 12. Veme, J.: Bull. Soc. chim. biol., 5:227, 1923. 13. Gerard, P., Cordier, R., and Lison, L.: Bull, d'histol. appliq. a la phvsiol., 7:133, 1930. 14. Ogata, T., and Ogata, A.: Beitr. z. path. Anat. u. z. allg. Path., 71:377, 1922-23. 15. Jancso, N. von: Ztschr. f. d. ges. exper. Med., 61:63, 1928; Jancso, N. von: Arch. f. exper. Zellforsch., 6:444, 1928. 16. Coujard, R.: Bull, d'histol. appliq. a la physioL, 20:161, 1943. 17. Gomori, G.: Arch. Path., 45:48, 1948. Organic Substances 123 tions. The slides are fixed in formaldehyde vapor and subse- quently treated with various reagents. It will be observed that phenol, tyrosine, catechol, and hydroquinone lose their azo-coupling reactions after this treatment (or produce azo dyes so pale as to be indistinguishable from the shade of the control marks ) ; the indophenol reaction of phenol and cate- chol is completely abolished. The chromaflBn and argentaflBn reactions of catechol and hydroquinone are greatly weak- ened. If formalin fixation is followed by treatment with 5 per cent potassium dichromate for 24 hours, the argentafiin reac- tion of these diphenols is abolished. On the other hand, resorcinol and phloroglucinol fully retain their azo-coupling ability and will produce azo dyes of brilliant shades; their in- dophenol reaction also remains unimpaired. They will show an intense chromafiin and argentafiin reaction; the latter is resistant to treatment with dichromate. These observations are of great importance, as they show that in the case of phenolic substances the results of test-tube experiments cannot be used for their identification in fixed tissues. Adrenalin The adrenal medulla reduces alkaline silver solutions,^^ stains brown with dichromates,^^ gives azo dyes (of an in- conspicuous ochre-yellow shade, suggestive of catechol de- rivatives) with diazonium compounds, shows a definite in- dophenol reaction and a greenish staining with ferric chloride solutions. There can be little doubt that all these reactions are due to the presence of adrenalin. However, they can be obtained with fresh material only; once the tissue is fixed either in formalin or in alcohol, they become negative. With alcohol, loss of reactivity is probably due to extraction of adrenalin from the tissue; with formalin, chemical changes induced in the molecule must be held responsible for the negative reactions. For the histochemical demonstration of adrenalin, the tis- 18. Henle, J.: Ztschr. f. ration. Med., 24:142, 1865. 124 Microscopic Histochemistry sue should be fixed in Regaud's mixture ( 10 per cent formalin containing 5 per cent potassium dichromate ) . Mercury-con- taining fixatives, such as Zenker's fluid, give much poorer results. In the finished sections adrenalin-containing cells appear in a more or less dark-brown shade. The argentafiin reaction (Ogata," Baginski'^) is not rec- ommended; it does not give a sharp localization. The ENTEROCtmoMAFFiN (EC) Substance The granules of the EC cells show a number of interesting staining reactions, some of which are capable of a chemical interpretation. All the reactions to be mentioned here can be ehcited even after prolonged formalin fixation. 1. The granules are chromaflfin.^*^ 2. They are argentafiin,^^ even after fixation in dichromate- containing mixtures. 3. At an alkaline reaction tliey give intensely colored azo dyes with diazonium compounds.^^* ^^ 4. They give a fairly intense indophenol reaction,^* espe- cially with Gibbs's reagent. 5. They reduce ferriferricyanide to Prussian blue. In addition to these reactions, they stain intensely with some lake dyes (hematoxylin, gallocyanin, celestin blue, etc. ) in the absence of metal salts.^^ The chemical explanation of this staining property is not clear. They are also stained by various silver-impregnation techniques, such as Bodian's"^ (this is not an argentafiin reaction; see p. 58 ) . On the basis of their chemical reactions, which are the same as those of adrenalin in the test tube (Verne^^), Cor- 19. Baginski, S.: Bull, d'histol. appliq. a la physioL, 5:129, 1928. 20. Heidenhain, R.: Arch. f. mikr. Anat., 6:368, 1870. 21. Masson, P.: Compt. rend. Acad, sc, 158:59, 1914. 22. Cordier, R., and Lison, L.: Bull, d'histol. appHq. a la physiol., 7:140, 1930. 23. Lison, L.: Arch, de biol., 41:343, 1931. 24. Jonnard, R.: J. de phys. et de path, gen., 32:731, 1107, 1934. 25. Clara, M.: Ztschr. f. Zellforsch. u. mikr. Anat., 22:318, 1934-35. 26. Dawson, A. B.: Anat. Rec, 91:53, 1945. Ormnic Substances 125 fc5 diex^^ drew the conclusion that EC granules must contain an o- or a p-diphenol or aminophenol or diamine. Cordier and Lison,^^ in a later paper, found that all possibilities except that of an o-diphenol, with a short-chained substituent in one of the p-positions (there are two such positions on account of the two phenohc hydroxyls), can be ruled out. This view has received general acceptance, in spite of the fact that it does not explain either the discrepancies between the behav- ior of adrenalin and the EC substance or the slowness with which EC granules reduce alkaline silver solutions. Lison^ thinks that adrenalin is simply not fixed by formalin, while the EC substance ( or its protein matrix ) is. It would be dif- ficult to ascertain the correctness or incorrectness of this theory, the products of fixation being invisible. On the other hand, formalin-iodate or formalin-dichromate mixtures do fix both adrenalin and EC substance in the form of sharply locahzed brownish granules; yet adrenahn loses its typical reactions, while the EC granules retain them. Also, deriva- tives of catechol (adrenalin, 3,4-dihydroxyphenylalanine ) and of hydroquinone ( homogentisic acid) reduce silver solu- tions in a matter of seconds, while the EC granules require hours. In Coujard slides the reactions of resorcinol are invariably identical with those of the EC granules,^^ including the shade of the azo dyes produced and even the staining by gallocyanin and celestin blue. On the basis of these observa- tions, it appears safe to assume that the EC substance is a derivative of resorcinol and not of catechol. The EC cells fluoresce intensely in ultraviolet light.^"^ From his comparative studies of their fluorescence spectrum, Jacob- son^* came to the conclusion that these cells contain some derivative of pteridine. The evidence in favor of this assertion is convincing. However, his other theory, namely, that the typical reactions of the EC cells are due to a pteridine com- 27. Eros, G.: Zentralbl. f. allg. Path. u. path. Anat., 54:385, 1932. 28. Jacobson, "W.: J. Path. & Bact, 49:1, 1939. 126 Microscopic Histochemistry pound, cannot be accepted. Pteridine derivatives do not give a single one of the reactions of the EC cells in the Coujard experiment. For the histochemical staining of the EC cells the argentaf- fin reaction, azo-coupling, and the indophenol reaction are recommended. With all techniques, prompt fixation of the tissue is important because the EC substance undergoes fairly rapid decomposition. The reactions are quite intense with normal EC cells; in carcinoid tumors the intensity of staining is variable and may even be negative, depending on the degree of biochemical differentiation of the neoplastic cells. 1 ) The argentaffin reaction.— i Method Fix tissues preferably in Bouin's fluid or in formalin. Other formalin-containing mixtures are also usable, but the con- trast between the granules, and the background is impaired by dichromates and mercury salts. Treat sections for 10-60 minutes with Lugol's solution; de- colorize with thiosulfate. This treatment helps to suppress the staining of the background. Wash slides thoroughly in distilled water. Incubate them for 12-24 hours at 37° C. either in methenamine-silver (p. 60), buffered at pH 8-8.5 with a borate buffer, or in Fontana's"^ solution (p. 60). In- spect the slides under the microscope at intervals; as soon as the EC cells appear black, remove them from the solution. Sections stained with methenamine-silver can be toned with gold chloride; gold toning is not advisable after Fontana s solution because it may cause considerable fading of the granules. Rinse sections in a dilute solution of Na thiosulfate, wash them under the tap, counterstain as desired, dehydrate, and mount. The specificity of the method is satisfactory but, like that of all silver techniques, relative. On continued incubation 29. Fontana, A.: Dermat. Ztschr., 46:291, 1925-26. Organic Substances 127 (over 24 hours) a number of additional structures will be stained; an early and intense blackening of eosinophilic and neutrophilic granules is especially conspicious.^^' ^^ After 48 hours or more, practically the entire slide may turn solid black. If a darkish background develops on account of over- staining, differentiate slide as described under the Ag tech- nique for glycogen and mucin (p. 64). According to Burtner and Lillie,^^ the argentaflBn reaction can be greatly accelerated by performing it at 60° C. 2) The azo-coupling reaction.— Method Dissolve about 50 mg. of either Red B Salt or Black B Salt or Garnet GBC Salt (chemical constitutions, p. 171 ) in about 10 ml. of cold water; add a few drops of a saturated solution of borax and pour the more or less turbid solution on the slide. Leave it on for 30-60 seconds; wash slide under the tap, counterstain lightly with hematoxylin, dehydrate, and mount. EC granules stain deep orange (Red B Salt), rusty red-brown (Black B Salt), or red (Garnet GBC Salt). The background is light yellow. In order to obtain darker shades, Lison^^ recommends the use of tetrazotized diamines, in the hope that only one diazo group will couple with the tissue, and the second one can be coupled with naphthol. In this way intense purple or bluish shades could be produced. The method does work in prac- tice, but the background becomes stained so heavily that the net gain in contrast is negligible. The two-step modification of Clara^^ (using unilaterally diazotized diamines and diaz- otizing the other side after coupling with the tissue has taken place ) is not better than Lison's original method. 30. Cordier, R.: Arch, de biol., 36:427, 1926. 31. Burtner, H. J., and Lillie, R. D.: Stain Technol., 24:225, 1949. 32. Clara, M., and Canal, F.: Ztschr. f. Zellforsch. u. mikr. Anat., 15:801, 1932; Clara, M.: Ztschr. f. wissensch. Mikr., 51:316, 1934. 128 Microscopic Histochemistry 3 ) The indophenol reaction.— This method gives distinctly less brilliant pictures than the previous ones but is quite specific. Method Dissolve about 50 mg. of 2,6-dibromoquinonechloroimide (Eastman No. 2304) or the corresponding dichloro-com- pound (Eastman No. 2483) in 5-10 ml. of alcohol, add 40-50 ml. of water and a few drops of a saturated solution of borax. Immerse slides for 10-15 minutes. The solution will turn a dark gray-brown. Remove slides, wash, and counterstain them with a red nuclear dye. Dehydrate and mount. EC granules stain in moderately intense shades of gray-blue. SULFHYDRYL (ThIOl) GrOUPS Although sulfhydryl groups occurring in animal tissues are invariably carried by amino acids, their special biochemical significance warrants a discussion apart from the rest of the amino acids. Of the sulfur-containing amino acids, only glutathion is uncombined and freely diffusible. The others (cysteine, methionine, etc.) are incorporated into protein molecules, share the solubility properties of proteins, and are precipi- tated by fixatives. Sulfhydryl compounds are fairly strong reducing agents and are transformed by oxidation into un- reactive disulfides. Unless the tissues are examined fresh, most of their sulfhydryl groups will have undergone oxi- dation. Since glutathion is diffusible, its exact histochemical local- ization is impossible, even if the tissue is fixed in a liquid which will precipitate it quantitatively (neutral formalin containing 1-2 per cent of cadmium acetate or lactate ).^^ However, it is possible to localize sulfhydryl groups in pro- teins. All the methods, to be mentioned show the presence of 33. Binet, L., and Weller, G.: Bull. Soc. chim. bioL, 16:1284, 1934; Joyet-Lavergne, P.: Compt. rend. Soc. de biol., 128:59, 1938. Organic Substances 129 thiol groups only. Disulfides do not react unless they are first reduced to thiols. The most effective reducers are sodium cyanide or sodium sulfite (5 per cent solutions; treat tissue for about 10 minutes ) .^* It appears that trichloroacetic acid (2-20 per cent) or saturated ammonium sulfate also "reveal" or "unmask" disulfides; their mode of action is not clear. The shades, produced by the reagents are too pale to be observed in thin sections; thick frozen sections or teased preparations of fresh tissues are preferred. Fixation in for- malin-saline or alcohol is permissible. 1) The nitroprusside reaction (Bw^^a).^^— Alkalized nitro- prusside (about 5 per cent nitroprusside with 1-2 per cent ammonia added ) produces a fleeting purple coloration with -SH groups. If the tissue is dipped for a few seconds in a 5 per cent solution of zinc acetate and then transferred to the reagent, the color will be stable enough to permit dehy- dration and mounting of the tissue.^^ This reaction appears to be specific. The only other sub- stance giving a similar reaction is. creatinine; it is washed out of tissues by a short rinse. 2) The ferriferricyanide method (Chevremont and Fred- eric).^''—li tissues are treated with a freshly prepared solution containing about 0.2 per cent each of potassium ferricyanide and ferric ammonium citrate, with a few drops of dilute acetic acid added, -SH groups will reduce the ferricyanide and produce a precipitate of Prussian blue. The specificity of this method is limited; almost any reducing substance will be stained blue more or less promptly. 34. Joyet-Lavergne, P.: Bull, d'histol. appliq. a la physiol., 5:331, 1928; Joyet-Lavergne, P.: Compt. rend. Soc. de biol., 98:658, 1928; Serra, J. A.: Stain Technol., 21:5, 1946. 35. Buffa, E.: J. de physiol. et path, gen., 6:645, 1904. 36. Giroud, A., and BulHard, H.: Bull. Soc. chim. biol., 14:278, 1932; Giroud, A., and Bulliard, H.: Protoplasma, 19:381, 1933; Giroud, A., and Bulliard, H.: Bull, d'histol. appliq. a la physiol., 11:169, 1934; Giroud, A., and Bulliard, H.: Arch, d'anat. micr., 31:271, 1935. 37. Chevremont, M., and Frederic, J.: Arch, de biol., 54:589, 1943. 130 Microscopic Histochemistry 3 ) Bennett's method. ^^— This method is based on the well- known reaction between sulfhydryl groups and chloromer- curiphenol. The reagent is p-chloromercuriphenylazo-^-naph- thol, which is a poorly soluble red dyestuff. It will stain pro- teins, containing -SH groups in a very specific way but in a rather pale reddish shade. It would be interesting to synthesize chloromercuri-a- naphthol, which should also combine with sulfhydryl groups. In a second step, it could be coupled with a suitable diazo- nium salt (e.g., Blue B Salt) to yield a very dark purple- black azo dye. E. VARIOUS UNCLASSIFIED ORGANIC SUBSTANCES Flavoproteins Chevremont and Comhaire^ have devised a method for the histochemical demonstration of riboflavin, based on the fact that riboflavin, if first reduced to leucoflavin, will reoxi- dize in air to red rhodoflavin. Since frozen sections of for- malin-fixed material are used, only protein-bound riboflavin will be demonstrated. Method Move frozen sections around in 1-2 per cent HCl, contain- ing enough zinc dust to keep it bubbling, for about 30 min- utes. Rinse sections in distilled water and expose them to air for a few hours in a shallow layer of water. Mount in glyc- erin-gelatin. Flavoproteins are stained red. This is an untested method. Sodium hydrosulfite ( Na2S204 ) would be a simpler reducer than nascent hydro- gen and just as effective. There is a possibility that flavoproteins could be localized by their green fluorescence. The distinctive feature is the 38. Bennett, H. S.: Anat. Rec., 100:640, 1948; Bennett, H. S., and Yphantis, D. A.: J. Am. Chem. Soc, 70:3522, 1948; Mescon, H., and Flesch, P.: J. Nat. Cancer Inst., 10:1370, 1950. 1. Chevremont, M., and Comhaire, S.: Arch. f. exper. Zellforsch., 22:658, 1939. Organic Substances 131 immediate disappearance of fluoresence on the addition of hydrosulfite to the mounting medium. Lewisite Lewisite, a war gas, is a mixture of chlorovinylarsins. In an alkahne medium it decomposes to yield acetylene; the latter will precipitate cuprous ions in the form of red copper carbide. This reaction is utilized for the histochemical dem- onstration of lewisite. Method^ Prepare three stock solutions: A, an 8 per cent solution of cupric sulfate in distilled water; B, a 20 per cent solution of sodium sulfite in distilled water; C, a solution of 50 g. of sodium thiosulfate and 20 g. of sodium hydroxide in 80 ml. of distilled water to which 1 ml. of piperidine has been added. The latter enhances the red shade of the precipitate. For use, mix 10 ml. each of solutions A and B; when clear, add 10 ml. of solution C. The mixture is unstable and can be used for only about 1 hour. Dip frozen sections of fresh tissue in the reagent for a few seconds; wash them, dehydrate in alcohol, and mount in bal- sam. Counterstaining is not recommended. This is an un- tested method. "Mustard" (Dichlorodiethylsulfide) Gold chloride reacts with mustard to yield a bright yellow complex, which can be reduced to black metallic Au. Method^ Immerse frozen sections of fresh tissues in a 1 per cent solution of gold chloride for 2-3 minutes. Mount on a slide, flood with 5 per cent sodium hydroxide for 30 seconds. Wash, dehydrate, and mount in balsam. Shades of black and 2. Ferguson, R. L., and Silver, S. D.: Am. J. Clin. Path., 17:37, 1947. 3. Silver, S. D., and Ferguson, R. L.: Am. J. Clin. Path., 17:39, 1947, 132 Microscopic Histochemistry purple-black indicate the presence of mustard. This is an untested method. F. PIGMENTS Pigments will be defined as substances visible in unstained preparations by virtue of their own color. They form an ex- tremely heterogeneous group as to both their origin and their chemical nature. Formalin pigment, for example, is an arti- fact, not originally present in the tissues but produced by the action of formaldehyde on hemoglobin. Some of the pig- ments are entirely foreign to the organism, introduced acci- dentally ( "exogenous" pigments, such as carbon and various metals); others are the products of normal or pathological metabolic processes within the organism ("endogenous" pigments ) . The number of colored substances found in various animal and plant species is very large, and but a small fraction of them have been identified chemically. In this section only such pigments as are encountered in tissue sections of man and the more common laboratory animals will be dealt with. I. Formalin pigment^ occurs in tissues fixed in acid for- malin or formalin-containing mixtures; it is found mainly in and around larger collections of blood. It forms brown-black granules of a crystalline structure. Two methods of removal are recommended. 1 ) Barrett's method.^— Treat the section for 10 minutes to 2 hours with a saturated solution of picric acid in alcohol. 2) Kar dose wit scKs method.^— Treat the section for 1-2 hours with a 2-3 per cent solution of concentrated ammonia water in 70-80 per cent alcohol. The formation of formalin pigment can be prevented by using a formaldehyde solution buffered to pH 6-7 with M/15 phosphate. 4. Lillie, R. D., and Hershberger, L. R.: Bull. Intemat. A. M. Mus., 27:136, 1947; Hershberger, L. R., and Lillie, R. D.: Bull. Intemat. A. M. Mus., 27:162, 1947. 5. Barrett, A. M.: J. Path. & Bact, 56:135, 1944. 6. Kardasewitsch, B.: Ztschr. f. wissensch. Mikr., 42:322, 1925. Organic Substances 133 II. The exogenous pigments are mainly carbon and various metals or metal-protein compounds, possibly also sulfides. Carbon can be recognized by its opaque blackness and by its absolute resistance to all bleaching agents and solvents. Metallic pigments have been discussed in the section on "Metalhc elements" (p. 29). III. The endogenous pigments fall into several classes ac- cording to their origin and chemical nature. Some of the individual pigments are chemically well-defined entities; others are a conglomeration of a number of related sub- stances which probably represent successive stages in the evolution of an ultimate colored substance. The most im- portant endogenous pigments will be divided into three groups : 1. Hematogenous pigments include hemoglobin and its degradation products, some of which contain ferric iron and some of which are iron-free. 2. Phenolic pigments are formed by the oxidation and polymerization of catechol and hydroquinone (possibly also tyrosine ) derivatives. 3. Lipogenous pigments result from the oxidation and polymerization of unsaturated fatty acids. Other chemically well-defined pigments, such as porphy- rins and carotenoids, very seldom occur in the tissues of higher species in concentrations high enough to be visible under the microscope. Crustacean tissues, on the other hand, may be very rich in carotenoids. 1) Hematogenous pigments.— A) Iron pigments.— a) Hemoglobin is demonstrated by its peroxidase action (p. 162). It should be remarked that un- altered hemoglobin is best identified by the zinc-leuco method, but its immediate degradation products, such as those seen in renal tubules 1 or 2 days after a hemolytic reac- tion, may not stain at all. They can be demonstrated by the benzidine method. It can be shown in test-tube experiments that the recolorization of zinc-leuco dyes by peroxide is a true enzymatic reaction requiring hemoglobin itself, whereas 134 Microscopic Histochemistry the benzidine reaction can be catalyzed also by nonprotein heme compounds. b) Hemosiderin is a yellowish, brownish, or greenish- brown granular pigment which resists the bleaching action of hydrogen peroxide and of permanganate and is not argen- taffin. It does not dissolve in dilute acids, or alkahs. It can be identified by the Prussian blue test (p. 38). c) Malaria pigment is an amorphous brown-black granu- lar substance, soluble in dilute alcoholic alkalis or Barrett's solution (p. 132); it is bleached by 3 per cent hydrogen per- oxide within 2 hours. It has repeatedly been identified as hematin."^ Naturally, the pigment will not give a direct reac- tion for iron. After destroying the organic matter with chlo- rine, hydrogen peroxide, or alkali, a positive Prussian blue reaction can be obtained.^ It is not possible to differentiate malaria pigment from formalin pigment histochemically. In diagnostic cases, the formation of formalin pigment must be prevented by proper fixation. B) Iron-free pigments —Bile pigments appear under the microscope as yellowish-green, more or less coarse granules which are not bleached by hydrogen peroxide or by perman- ganate and are not argentaffin. They are slowly converted by oxidants (hydrogen peroxide, Lugol's solution,^ nitrous acid,^^ dichromates, etc. ) into green biliverdin. Hematoidin is chemically identical with bilirubin and shares its reactions. 2 ) Phenolic pigments.— A) Melanin is an oxidation-polymerization product of dioxyphenylalanine and possibly also of tyrosine.^^ It forms yellowish-brown or grayish to almost jet-black granules, in- 7. Sinton, J. A., and Ghosh, B. N.: Rec. Malaria Survey India, 4:205, 1934; Morrison, D. B., and Anderson, W. A. D.: Pub. Health Rep., 57:90, 1942; Devine, J., and Fulton, J. D.: Ann. Trop. Med., 36:167, 1942. 8. Kosa, M.: Virchows Arch. f. path. Anat., 258:186, 1925; Okamoto, K.: Taishitzu Gaku Zasshi, 13:55, 1944. 9. Stein, J.: Compt. rend. Soc. de biol., 120:1137, 1925. 10. Okamoto, K., Sengoku, M., and Hirotani, N.: Taishitzu Gaku Zasshi, 14:30, 1948. 11. Hasebroek, K.: Fermentforsch., 5:1, 1922; Sato, K., and Brecher, L.: Arch. f. mikr. Anat., 104:649, 1925. Organic Substances 135 soluble in dilute acids and alkalis and promptly bleached by acidified permanganate and slowly ( in several hours to days ) by hydrogen peroxide. It is argentaffin in that it will reduce alkaline silver solutions (Foot's or alkaline silver-methen- amine, p. 60) in about 3-12 hours. It does not give the Prus- sian blue reaction. B) The pigment of ochronosis derives from an unidenti- fied derivative of hydroquinone (possibly homogentisic acid). It imparts a homogeneous yellow-brown to blackish hue to certain tissues (cartilage, degenerated elastic fibers, etc. ) . It is not argentaffin. Its most typical tinctorial property is its very intense, almost black, staining with cresyl violet.^^ 3 ) Lipo genie pigments.— Iron-free, brownish pigments, stainable by fat dyes (to some extent even after paraffin-embedding ) , rather resistant to dilute acids and alkalis and to bleaching by oxidants, have been described under various names (chromolipoid,^^ hemo- fuscin,^* lipofuscin,^^ "Abnutzungspigment,"^^ pigment of wear and tear, luteolipin,^^ ceroid,^^ cytolipochrome,^^ etc. ) . Their other properties include a brownish fluorescence, stain- ing by basic aniline dyes, especially fuchsin,^^ increase in basophilia after oxidation by permanganate,^^ metachromasia with methyl green,-^- ^^ a positive Schiff reaction after peri- odate treatment,^^' ^^ peroxide-like effects, etc.^^' ^^ Whether 12. Friedrich, H., and Nikolowski, W.: Arch. f. Dermat. u. Syph., 192:273, 1951. 13. Ciaccio, C: Biochem. Ztschr., 69:313, 1915. 14. Recklinghausen, F. D. von: Berl. Win. Wchnschr., 26:925, 1889. 15. Borst, M.: Pathologische Histologie (Leipzig: F. C. V^. Vogel, 1922). 16. Lubarsch, O.: Zentralbl. f. allg. Path. u. path. Anat., 13:881, 1902. 17. Rossman, L: Carnegie Inst. Washington Publ, 541:97, 1942. 18. LilHe, R. D., Ashbum, L. L., Sebrell, W. H., Daft, F. S., and Lowry, J. v.: Pub. Health Rep., 57:502, 1942. 19. GiUman, J., and Gillman, T.: Am. J. Path., 40:239, 1945. 20. Mallory, F. B.: Pathological technique (Philadelphia: W. B. Saunders Co., 1938). 21. Endicott, K. M., and Lillie, R. D.: Am. J. Path., 20:149, 1944. 22. Gyorgy, P.: Am. J. Clin. Path., 14:67, 1944. 23. Popper, H., Gyorgy, P., and Goldblatt, H.: Arch. Path., 37:161, 1944. 24. Elftman, H., Kaunitz, H., and Slanetz, C. A.: Ann. New York Acad. Sc, 52:72, 1949. 136 Microscopic Histochemistry or not all or most of these characteristics apply to all forms of the pigment cannot be decided, because there are only a few thorough studies available; in most reports only a few of the properties are mentioned. In the writer's experience the type and duration of fixation markedly affect staining prop- erties. Dichromate or moderately prolonged fixation in for- malin tends to enhance basophilia, the Schiff reaction, acid- fastness, and affinity to fat stains after paraffin-embedding, whereas alcohol or short fixation in formalin has the opposite effect. The pigment granules in one section, even within one cell, may exhibit considerable variation in staining intensity. Minor species differences have been noted by Lee.^^ It appears that all the pigments in this group are of an essentially similar nature. They all derive from the oxidation and polymerization of unsaturated fatty acids and represent various stages of one process. Staining reactions similar to the ones mentioned will develop in droplets of unsaturated fat injected into the tissues^^ and even in fats allowed to stand in air.^^' ^^ Since the chemistry of the oxidation of un- saturated fats is very incompletely understood, it is impos- sible to specify just what varying degrees of such staining properties as basophilia or acid-fastness could mean. How- ever, since all transitions from one extreme to the other may be observed in a single slide, it is doubtful whether any clas- sification within the group is biologically meaningful. Ceroid, for example, is probably only a late product in the matura- tion of lipogenic pigments in general, although the process may have been accelerated and have actually progressed beyond the normal average. Gillman and Gillman's cytosiderin^^ appears to be a mix- ture of hemosiderin with lipogenous pigment. 25. Glavind, J., Granados, H., Hartmann, S., and Dam, H.: Experentia, 5:84, 1949. 26. Lee, C. S.: J. Nat. Cancer Inst., 11:339, 1950. 27. Hass, G. M.: Arch. Path., 27:15, 1939; Endicott, K. M.: Arch. Path., 37:49, 1944. 28. Smith, J. L.: J. Path. & Bact., 11:415, 1906. CHAPTER VIII ENZYMES The microscopic identification and localization of enzymes in tissue sections presents the youngest offshoot of histo- chemistry. True, a few reactions purporting to reveal en- zymes were described a long time ago; however, some of them were shown later to be definitely nonenzymatic, and the enzymatic nature of others is still debatable. At the present time only a very hmited number of en- zymes can be demonstrated histochemically. The vast major- ity of enzymes are too labile to resist the manipulations neces- sary in histological technique. Others are resistant enough, but, so far, no reactions suitable for histochemical application have been devised for their demonstration. Fortunately, a fair number of enzymes, mainly of the hy- drolytic series, resist a certain amount of physical and chemi- cal treatment and will tolerate histotechnical manipulations without undue loss of activity. Moreover, they catalyze reac- tions which yield insoluble precipitates, suitable for micro- scopic localization. Whether or not an enzyme is sufficiently resistant to be demonstrated histochemically can be determined either by ( 1 ) exposing Coujard slides, marked with serial dilutions of an active extract, to the physical and chemical agents used in the histological routine (fixatives, dehydrating and clear- ing agents, hot paraffin, etc.) of by (2) chemical assay of the activity of tissue samples in the fresh state and after treatment with the agents mentioned. Experience shows, that fair histochemical results can be obtained even if as much as 90 per cent of the enzyme is destroyed in the course of manipulations. Of course, in such cases the microscopic pic- ture cannot be expected to reveal the full extent of activity, 137 138 Microscopic Histochemistry because at sites of low enzyme concentration activity is likely to drop below the threshold of sensitivity of the method. PREPARATION OF TISSUES FOR ENZYMATIC REACTIONS The tissue need not be absolutely fresh, but refrigeration is advisable whenever processing cannot be prompt. In most cases satisfactory results are obtained with tissues preserved in the icebox for several (often up to 48) hours, although the pictures may not be quite so sharp as with fresh tissue. If fixation is not permissible, fresh-frozen sections must be used. In such cases it is imperative to keep the enzyme un- dissolved; otherwise correct locahzation is impossible unless the enzyme happens to be insoluble. The simplest way to achieve this is to avoid contact with water or with dilute saline solutions and to use strong solutions of ammonium sulfate {¥2 saturated or better) up to the point in the proce- dure at which the localizing precipitate has been obtained. It is very likely that a rapid freezing-cutting method, such as that of Adamstone and Taylor^ or the simpler original procedure of Schultz-Brauns,^ followed by a brief fixation in acetone, would give good results with many enzymes. The optimal method for most enzymes is probably the freezing-drying technique. Unfortunately, most workers do not possess the equipment and must depend on simpler methods, similar to those used in the average laboratory of histology or pathology. It is to such methods that this chap- ter will be devoted. Fixation— As a rule, the best fixative for all enzymes is chilled acetone, which, of all fixatives, causes the least inac- tivation. A bottle of acetone should be kept in the icebox for routine use. Cytological details are not so good as one would like them to be, but they are satisfactory for most purposes if the slices are thin enough (not over 3 mm. in thickness). It is advisable to chill (not freeze) the tissue itself before 1. Adamstone, F. B., and Taylor, A. B.: Stain TechnoL, 23:109, 1948. 2. Schultz-Brauns, C: Klin. Wchnschr., 10:113, 1931. Enzymes 139 throwing it into the cold fixative; many shrinkage artifacts can be avoided in this way. Acetone is especially recom- mended for acid phosphatase. For other enzymes, cold ethyl alcohol ( 90-100 per cent ) is equally satisfactory and actually preferable because it gives a better cytological fixation and the tissue is easier to handle. A compromise fixative consist- ing of equal volumes of acetone and alcohol will give excel- lent results in the overwhelming majority of cases. Methyl al- cohol is entirely unsuitable as a fixative, since it destroys most enzymes. Most hydrolytic enzymes are reasonably re- sistant to formalin^ and can be fixed in 10 per cent formalin^ (preferably adjusted to pH 6-6.5 with a small amount of phosphate buffer). Again, fixation in the icebox is recom- mended. Formalin-fixed tissues should be cut frozen; embed- ding usually gives very poor results, even if the formalin is washed out carefully. Exceptions will be mentioned under the individual enzymes. Dehydration and embedding.— The successful use of frozen sections has been reported repeatedly, especially after for- malin fixation. Alcohol- and, even more, acetone-fixed tissues cut poorly by the freezing method. In addition, the enzymes are not always irreversibly precipitated by these fixatives. In general, embedding is preferable whenever feasible. For dehydration, alcohol or acetone can be used. For best results, dehydration should be carried out at icebox tempera- ture. After complete dehydration the pieces can be em- bedded though benzene or chloroform into paraflBn. Regular celloidin-embedding, although it may yield good results in some cases, is not recommended because it requires too long an exposure to alcohol-ether; arid, in addition, the dilute al- cohol in which celloidin blocks are stored is quite harmful to enzymes. Semi-embedding in dilute celloidin has been men- tioned previously (p. 15). Acetone-fixed tissues have a tendency to crumble during 3. Seligman, A. M., Chauncey, H. H., and Nachlas, M. M.: Stain Technol., 26:19,1951. 140 Microscopic Histochemistry cutting and to disintegrate when floated on water. The reni- edy is infiltration of the blocks with thin celloidin before paraffin-embedding. The temperature of the paraffin oven should not exceed 56° C, and the tissues should not be exposed to this tem- perature for longer than IM hours. Embedding can be has- tened by the use of reduced pressure.* A piece of rubber tubing, connected through a safety bottle to a water pump and introduced into the oven through its ventilating opening, will serve as a simple suction apparatus. The melted paraffin is kept in a wide-mouthed bottle which has a tight-fitting stopper, with a piece of glass tubing in its single perforation. The end of the rubber tubing is attached to the glass. The bulk of benzene or chloroform is removed by the first change of paraffin in an open dish (about 15-20 minutes). The pieces are now transferred to the suction bottle, and the vacuum is turned on gradually. Bubbling, quite lively at the start, soon decreases and ceases altogether in about 30 min- utes. A third change of paraffin in an open dish (about 10- 15 minutes) completes infiltration, and the pieces are then ready to be embedded. The following schedule of fixation and embedding is sug^ gested for all enzymes unless otherwise specified in the text: 1. Chill tissue for 15-30 minutes in the icebox. Fix shoes not thicker than 3 mm. in chilled acetone for 24 hours. It is advisable to trim the pieces even thinner with a razor blade as soon as they have gained some consistency (2-3 hours). 2. Dehydrate in two or three changes of absolute acetone or alcohol, about 12 hours each. 3. Transfer pieces to a mixture of equal volumes of alcohol and ether for a few hours. 4. Transfer to a 2-3 per cent solution of celloidin (collo- dion, U.S.P., diluted with ^A to 1 volume of alcohol-ether mixture) for 12-24 hours. 5. Drain pieces rapidly and carry them through two changes of chloroform, 32-1 hour each. ^ 4. Gomori, G.: Am. J. CHn. Path., 16:347, 1946. Enzymes 141 6. Embed in paraffin under reduced pressure, as described above. Steps 1-4 should be performed at icebox temperature. In the case of alcohol or acetone-alcohol fixation, steps 3 and 4 can be omitted; in step 5 benzene may be used instead of chloroform. Sections are cut at 3-8 fx, floated on lukewarm water, mounted with albumin-glycerol, and dried. The dried sec- tions should be melted in the paraffin oven ( 5-10 minutes ) . The coating of paraffin they acquire by this treatment ap- pears to have a protective action against atmospheric influ- ences ( oxygen, moisture ) . Unmelted sections may lose most of thek activity in a matter of a few weeks or months, whereas melted ones, just like uncut paraffin blocks, remain virtually unchanged for many years. For use, the slides are carried through xylene and alcohols to water. It may be advisable to protect the tissue against loss of enzyme with a thin layer of collodion. This can be done by flooding the slide (after the second alcohol) with dilute ( about 0.5 per cent) collodion in alcohol-ether, shaking off the excess, and hardening the membrane for a minute or so in 80-95 per cent alcohol. Some substrates do not pass the membrane readily. In such cases, protection with collodion should be omitted, and even the collodion from the embed- ding procedure should be removed by a short rinse in alco- hol-ether or acetone. HISTOCHEMICAL REACTIONS FOR ENZYMES Sometimes the primary product of the reaction is an in- soluble dye (e.g., red formazan from triphenyltetrazolium chloride) or a soluble one (e.g., acid aniline dyes from the corresponding leuco-dyes), permitting immediate visualiza- tion of sites of activity. Soluble dyes, as a rule, are much less reliable as far as correct localization is concerned, because they will stain the true sites of liberation only if they have an affinity to some structure present locally; otherwise they may diffuse away and may even stain distant and inactive structures for which they do possess an affinity. 142 Microscopic Histochemistry In other cases, including all hydrolytic enzymes, the prod- uct of enzymatic activity is a colorless and soluble substance (ions, phenolic compounds) and must be precipitated in a second step by a suitable reagent, added to the incubating mixture. If correct localization and high sensitivity are to be achieved, precipitation must be very prompt (of the veloc- ity of ionic reactions ) in order to bind the reaction product as fast as it is liberated, and the precipitate must be exceed- ingly insoluble. The less perfectly these conditions are satis- fied, the less sensitive the method and the poorer the local- ization of activity will be. These important principles will require a somewhat more detailed consideration. No substance is totally insoluble. If an "insoluble" com- pound is made from its components in a solution, it will pre- cipitate only if and when its solubility (or, in the case of a salt, solubihty product) is exceeded. Constancy of results, maximum sensitivity and correct localization in a histochemi- cal experiment can be attained only if the velocity of precipi- tation is almost infinite and the precipitate is practically in- soluble. These two aspects do not necessarily run parallel; even in the case of highly insoluble precipitates a transient phase of supersaturation or colloidal state may occur. For- tunately, this is very rarely observed in practice. In the opti- mal case all the reaction product v^ll be precipitated imme- diately and at the exact site where it is formed. Otherwise, some of it will diflFuse away from the primary sites of forma- tion before it can be precipitated and raise the concentration in the ambient fluid. In unfavorable cases very little or even none will precipitate locally, in spite of high enzymatic activ- ity; practically all the reaction product will be washed away and contribute to the saturation of the medium, first only around centers of high activity and later everywhere. The difference between the rates of hydrolysis and of loss by dif- fusion must have a minimum absolute value, proportional to the solubility of the precipitate, for a local precipitation to Enzymes 143 occur. This fact explains the interesting observation of "all- or-none" eflFects^' ^ in enzymatic histochemistry: either the minimum value is exceeded, and a local precipitation will be obtained; or it is not, and the reaction will be negative. In the borderline region, a minimal change in conditions may cause very marked differences in the extent of the reaction. Even serial sections incubated together in the same Coplin jar may exhibit wide variations in the pattern of distribution, on account of random convection currents in the substrate mixture, causing a temporary imbalance of concentrations and of temperature. Such chance variations may easily be misinterpreted as significant. False negative reactions at the sites of enzymatic activity are only one type of error due to too soluble a precipitate. A different kind of error may be caused by the diffusion of the undeposited reaction product into the incubating me- dium. If and when the point of saturation is reached, the solute will settle out indiscriminately all over the slide. Such precipitates are usually coarsely crystalline, easy to recog- nize, and not likely to be mistaken for the true localization of the enzymatic reaction. However, there exists another type of artffact that is easily confused with a tme reaction. Certain structures may possess an affinity for small molecules such as are produced by enzymatic hydrolysis and may adsorb them selectively, even from incompletely saturated solutions. A biological example of this is the in vitro calcification of cartilage in near-saturated solutions of calcium phosphate. A similar phenomenon is observed under the conditions of the histochemical method for alkaline phosphatase. Some structures, such as bone matrix and cell nuclei, themselves enzymatically inactive, readily adsorb calcium phosphate produced by enzymatic hydrolysis elsewhere, at active parts of the section.^ As can be expected, such artffacts will occur preferentially in the immediate vicinity of highly active cen- 5. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 70:7, 1949; Gomori, G.: Ann. New York Acad. So., 50:968, 1950. 6. Gomori, G.: J. Lab. & CHn. Med., 35:802, 1950. 144 Microscopic Histochemistry ters"^-^^ where a transient high concentration of phosphate ions may prevail for relatively long periods. This type of arti- fact will be accentuated on prolonged incubation. The sec- ondary staining of nuclei, themselves inactive, is a typical example of a false localization,^^- ^^ that is, a positive reac- tion due to enzyme contained in the tissue proper but appear- ing at sites other than those of the true primary microscopic localization. In a recent paper Johansen and Linderstr0m-Lang^^ give a quantitative evaluation of diffusion artifacts in the calcium- cobalt technique for alkaline phosphatase. They come to the conclusions that ( 1 ) calcium phosphate has a great tendency to form supersaturated solutions, and (2) the time required for the phosphate concentration to reach the critical level may vary from approximately 0.1 to 2.5 seconds. This time is ample to allow gross diffusion into the neighboring areas, where conditions may be more favorable for precipitation (the presence of preformed crystal nuclei; special chemical or physical configuration of cell structures ) . Therefore, it is very likely that calcium phosphate will precipitate at such preferential sites rather than in centers of high enzymatic activity. While the possibility of diffusion artifacts cannot be de- nied, the conclusions of Johansen and Linderstr0m-Lang are based on purely theoretical assumptions, some of which are not borne out by experimental facts and which make the situation look much worse than it actually is. For instance, the authors do not take into account the results of experi- ments showing that diffusion artifacts occur mainly under 7. DanieUi, J. F.: J. Exper. Biol., 22:110, 1946. 8. Jacoby, F., and Martin, B. F.: Nature, 163:875, 1949. 9. Martin, B. F., and Jacoby, F.: J. Anat., 83:351, 1949. 10. Feigin, L, Wolf, A., and Kabat, E. A.: Am. J. Path., 26:647, 1950. 11. Kroon, D. B.: Acta endocrinol., 2:227, 1949; Doyle, W. L.r Am. J. Anat., 87:79, 1950; Cleland, K. W.: Proc. Linnean Soc. New South Wales, 75:54, 1950; NovikofiF, A. B.: Science, 113:320, 1951. 12. Gomori, G.: J. Lab. & Clin. Med., 37:526, 1951. 13. Johansen, G., and Linderstr0m-Lang, K.: Acta chem. Scandinav., 5:965, 1951. Enzymes 145 nonoptimal conditions of incubation, such as too low pH values or Ca ion concentration. As the pH and Ca ion con- centration are raised, the localization becomes progressively sharper, and the stained areas shrink concentrically. Actually, under optimal conditions (p. 184) and an incubation time not exceeding IM hours, the diffusion artifacts described by Danielli^ and by Jacoby and Martin^- ^ cannot be repro- duced.^^ This observation, together with the facts that (1) artificial gradual supersaturation of the substrate mixture or the diffuse impregnation of the section with enzyme produce pictures entirely different form those obtained by the regular method and (2) the localization of activity is the same with the calcium-cobalt and the azo dye methods, does not sup- port the theory of the importance of false secondary localiza- tions. Furthermore, the thesis that phosphatase is distributed evenly (or at random) among all cells or even within one cell body cannot be accepted. Comparison with quantitative Coujard slides clearly shows that in many (if not most) cases the enzyme is concentrated in very small, discontinuous spots, the activity of which may exceed the average value for the whole tissue by a factor of the order of 100. Activities as high as 700 Bodansky units per gram of active structure (as against the average value of 1 unit, assumed by the au- thors) have been observed.^^ On the basis of the facts men- tioned, it is safe to assume that the highest intensity of the histochemical reaction coincides with centers of enzymatic activity. Further experiments will have to decide whether the diffusion artifacts persisting in spite of optimal condi- tions are within or beyond the resolving power of high-power dry objectives. It must be mentioned here that diffusion of the enzyme itself has been blamed for false localizations.^^' ^^' ^^ Diffusion 14. Gomori, G.: Unpublished. 15. Gomori, G.: Exper. Cell Research, 1:33, 1950. 16. Danielli, J. F.: Nature, 165:762, 1950. 17. Hebert, S.: Arch, de biol, 61:235, 1950; Yokoyama, H. O., Stowell, R. E., and Tsuboi, K. K.: J. Nat. Cancer Inst., 10:1367, 1950; Yokoyama, H. O., Stowell, R. E., and Mathews, R. M.: Anat. Rec, 109:139, 1951. 146 Microscopic Histochemistry of the enzyme from sections of embedded material must be considered unlikely/^- ^^ especially if collodion protection is used. At least, solutions in which a larger number of slides, carrying sections of highly active material, are soaked for hours, and even days, never acquire any demonstrable activ- ity. However, enzymes may be lost by diffusion from unfixed or poorly fixed tissues (frozen-dried material ) unless the prop- er precautions are taken (p. 12). The enzyme dissolved in the incubating medium may slowly decompose the substrate, cause its gradual saturation, and produce artifacts of the type described in the preceding passage. Theoretically, the enzyme itself may even be adsorbed in a secondary way on various striictures; however, it has been shown experimen- tally that such an adsorption can occur only under very ab- normal conditions,^^ not in the least likely to be met with in practice. In the case of precipitates which are not extremely insol- uble, diffusion currents simply wash away part of the reac- tion product and prevent its precipitation. On the other hand, the relative sluggishness of these currents may cause a somewhat different type of artifact if the precipitate is very insoluble, mainly in the case of some azo dyes but possibly even in the case of phosphates. Under such conditions the reagent may become exhausted in the vicinity of centers of very high activity more quickly than it can be supplied by diffusion from near by. With this lack of precipitant, the product of hydrolysis will remain unbound and free to dif- fuse until it encounters a fresh supply of reagent some dis- tance away. This will result in too little staining at the sites of true activity and an irregular precipitate of azo dye (or, in the case of Ca phosphate, staining of nuclei ) around them. This type of artifact can be prevented by agitating the slide vigorously during incubation, and so hastening the inter- change of fluid around the tissue. 18. Doyle, W. L.: Quantitative aspects of the histochemistry of phospha- tases. In Symposium on cytology (East Lansing: Michigan State College Press, 1951). Enzymes 147 While it may prove impossible to prevent artifacts com- pletely and in every case, every effort must be made to re- duce their occurrence. It is imperative to insure the prompt- est possible precipitation of reaction products formed. This can be done mainly by increasing the insolubility of the pre- cipitate, e.g., by lov^ ionic strength, an appropriate pH, and a high concentration of the precipitating ion. In addition, incubation should not be prolonged beyond the optimum.^^ The details of these measures will be given with the individ- ual techniques. Presaturation of the incubating medium with the precipitate to be formed is also practiced, but this pro- cedure is not very reliable. If saturation is not complete, it is of little help; if it is, the shghtest evaporation of the medium or change in its temperature may lead to supersaturation, with a resulting indiscriminate precipitation. A high rate of enzymatic activity is also important because of the critical nature of the difference between rates of hydrolysis and of diffusion. Enzymatic activity can be enhanced by a high substrate concentration, the use of activators and incubation at the optimal pH. Inhibitors should not be employed ex- cept for specific purposes. It is obvious that slow reactions for the precipitation of the primary product of enzymatic activity cannot localize cor- rectly, no matter how insoluble the final precipitate may be. For instance, indoxyl liberated enzymatically can be visual- ized by oxidation into insoluble indigo. The picture obtained, however, can serve only as a rough indication of the true distribution of activity, since the process of oxidation is a relatively slow one. The precipitate obtained in the course of enzymatic reac- tions is often colorless and invisible under the microscope. In such cases it must be transformed into a colored substance. The choice of the appropriate chemical procedure to ac- complish this will vary with the nature of the primary pre- cipitate. If the steps are well planned, there is Httle danger of producing spurious secondary localizations, because the 148 Microscopic Histochemistry reactions take place without any significant solubilization of the precipitate. False positive reactions, i.e., reactions misleadingly similar to a genuine one but not due to enzyme contained in the tissue, may be encountered occasionally. Only their causes will be enumerated here; their prevention will be dealt with under the individual methods. 1. The simplest mistakes result from confusing pigments with a positive reaction. Hemosiderin, for example, is con- verted by ammonium sulfide, a reagent often used in en- zymatic histochemistry, into green-black ferrous sulfide. Al- though the shade of this compound is quite different from the gray-black of cobalt sulfide or from the brown-black of lead sulfide, obtained in the course of enzymatic reactions, errors from this source are not impossible. 2. Preformed calcifications in tissues are essentially simi- lar in composition to some precipitates formed by enzyme activity and will be converted into the same colored end- products. 3. Heavy metals adsorbed by protein may constitute a source of error. 4. Certain substrates may undergo some degree of spon- taneous hydrolysis or oxidation during incubation. This will result in false positive reactions by the secondary adsorption of reaction products. 5. Bacterial contamination of the substrate is a rare com- plication if proper precautions are taken ( use of chloroform or camphor in the incubating medium), but it may occur once in a while, especially when incubation time is greatly prolonged. The medium may become grossly turbid, partly from the growth of microorganisms and partly from decom- position products of the substrate. The latter may be ad- sorbed on certain structures, causing false reactions. On account of the possibility of false positives, it is neces- sary that, whenever a new method is proposed, the enzy- matic nature of the reaction be proved. One or more of the Enzymes 149 following control tests, run alongside the regular method, can be used to rule out nonenzymatic reactions: 1. Treat the sections prior to incubation with agents known to destroy enzymatic activity (boiling water, strong mineral acid, oxidants, heavy metals, etc.). It must be re- marked that the resistance of enzymes varies widely; some are very easily destroyed, others are surprisingly resistant (e.g., hemoglobin is resistant to high temperatures and acids; myosin to precipitation by trichloroacetic acid). In some cases distinction between true and "pseudo"-enzymes must be arbitrary. 2. Use specific inhibitors ( cyanide, azide, eserine, etc. ) in an effective concentration. 3. Leave out the substrate from the incubating mixture. Whatever reaction persists under the conditions of these control tests cannot be du,e to an enzymatic effect. Of course, the picture in the regular slide may reveal a combination of enzymatic and nonenzymatic reactions. In such a case only the difference between the regular and the control slide can be attributed to enzymatic activity. The specificity of enzymes is rarely an absolute one. As a rule, one enzyme will attack several chemically related sub- strates ( although the optimal conditions of activity may vary with the substrate). Conversely, one substrate may be at- tacked by a number of enzymes (although, again, optimal conditions may vaiy with the enzyme ) . Whether one or sev- eral enzymes, are involved in any given reaction (or in re- lated reactions ) is a problem familiar to research workers in both bio- and histochemistry. In biochemistry the successful separation of a crude extract into fractions with distinctly different enzymatic properties is considered to be the evi- dence for the presence of several enzymes. In histochemistry, different topographical patterns of the distribution of activ- ity, obtained under different conditions (varying the sub- strate, pH; the presence of activators or inhibitors) is as- sumed to be the indication for the involvement of more than 150 Microscopic Histochemistry one enzyme. However, these different patterns must be quite constant and obtained under optimal conditions of precipi- tation in order to avoid the possible misinterpretation of random "all-or-none" effects and of diffusion artifacts. The classification of hist o chemically demonstrable en- zymes.—AW enzymes for which histochemical techniques are known belong in one of the two groups: (1) oxidative, (2) hydrolytic enzymes. 1. OXIDATIVE ENZYMES The oxidative enzymes fall into three groups: (a) dehy- drogenases, (h) oxidases, and (c) peroxidases. a) Dehydrogenases The dehydrogenases catalyze the transfer of hydrogen to immediate acceptors other than oxygen and peroxides, al- though the ultimate acceptor may be oxygen. They are rather delicate enzymes which are largely destroyed by any sort of fixation and completely destroyed by embedding. They are rapidly inactivated even on standing. They require coenzymes, and some of them are also linked to the diapho- rase or cytochrome systems.. The principle of their demonstration is the observation of the change in color of suitable hydrogen acceptors when they are reduced by the enzyme. The three main types of compounds used are: (1) methylene blue (Semenoff),^^ re- duced to colorless leuco-methylene blue and thus indicating the sites of activity by bleaching; (2) various tetrazolium compounds, introduced into enzyme research by Kuhn and Jerchel,-^' ^^ and Lakon;^^' ^^ they are reduced to bright red, 19. Semenoff, W. E.: Ztschr. f. Zellforsch. u. mikr. Anat., 22:305, 1934-35. 20. Kuhn, R., and Jerchel, D.: Ber. d. deutsch. chem. Gesellsch., 74:941, 1941. 21. Kuhn, R., and Jerchel, D.: ibid., 74:949, 1941. 22. Lakon, G.: Ber. d. deutsch. bot. Gesellsch., 60:299, 1942. 23. Lakon, G.: ibid., 60:434, 1942. Enzymes 151 purplish, or blue formazans, insoluble in water and soluble in fats; and (3) tellurites (first introduced into bacteriologi- cal technique by Klett),^* reduced to insoluble black ele- mentary tellurium. The methylene blue technique is not recommended be- cause (a) the negative image it gives allows only a very poor locahzation and (b) the method is cumbersome and must be carried out under strictly anaerobic conditions to prevent the reoxidation of the leuco-base by atmospheric oxygen. The tetrazolium method is the most sensitive of the three and aflFords an excellent localization, except for an occasional secondary staining of fat droplets by formazan, a complication which should be borne in mind to avoid mis- interpretation of the pictures. It should also be mentioned that some of the commercial batches of tetrazoHum com- pounds may contain oxidizing substances (probably lead tetra-acetate ) as an impurity and that these interfere with the reaction. The formation of dye should be quite notice- able after about 5-10 minutes of incubation if good active material is used (e.g., rat kidney); if not, the reagent must be recrystallized by dissolving it in a small volume of ab- solute alcohol and precipitating it with 4-5 volumes of ether. The tellurite method is considerably less sensitive, but it gives nice, sharp pictures. Even without the use of any substrate, positive reactions will be obtained in most cases on account of the presence of various endogenous substrates in the tissues. Although such reactions of an undefinable substrate specificity are usually weak, the substrate being exhausted rapidly, one should try to avoid them by rinsing the sections before incubation. For all practical purposes, reactions obtained in rinsed sections will be specific for the dehydrogenases of the substrate supplied. Maintaining the right pH under incubation is important. The optimum for the enzymatic activity is pH 7.3-7.6; at 24. Klett, A.: Ztschr. f. Hyg., 33:137, 1900. 152 Microscopic Histochemistry much higher pH values all the reagents mentioned will un- dergo nonenzymatic reduction by a number of substances, such as sulfhydryl compounds, polyphenols ( adrenalin ) , etc. Method According to Seligman and Rutenburg^^ the tissue need not be absolutely fresh; refrigeration for 4 hours at 4° C. does not cause any noticeable loss of activity; in fact, even fixation in chilled acetone for 4 hours causes only 40 per cent inactivation of the enzyme. Use frozen sections 25-50 ^ thick. Thinner sections often fail to stain ( destruction of cellular organization in the super- ficial layers or loss of cof actors by diffusion). Zweifach, Black, and Shorr^^ recommend the use of razor-blade hand sections, about 0.5-1 mm. thick. After the reaction has taken place, the pieces can be fixed in formalin, frozen, and cut. Loose cellular material (sediments, scrapings) are sus- pended in the incubating medium. The composition of the substrate mixture is not critical, except for the pH ( 7.3-7.6 ) . The concentration of the buffer (phosphate) should be 0.05-0.1 M; that of the substrate (suc- cinate, lactate, etc.), 0.1-0.2 M. Triphenyltetrazolium chlo- ride and neotetrazolium chloride^^ (a dimer of the former) are used in 0.5-1 per cent solutions; the more insoluble ditet- razolium chloride^^ (a dimethoxy derivative of neotetrazo- lium ) and potassium tellurite ( K2Te03 ) in a 0.1 per cent solu- tion. Incubation time at 37° C. will range from 20 minutes to 3 hours. The dye formed from triphenyltetrazolium is scarlet- red; that from neotetrazolium, purple-red; that from ditetra- zolium, in thin layers purple, in thick layers deep blue. The shdes should be examined within a few hours because for- 25. Seligman, A. M., and Rutenberg, A. M.: Science, 113:317, 1951. 26. Zweifach, B. W., Black, M. M., and Shorr, E.: Proc. Soc. Exper. Biol. & Med., 74:848, 1950. 27. Available from Pannone Chemical Co., Farmington, Conn. 28. Available from Dajac Laboratories, 511 Lancaster Ave., Leominster, Mass. Enzymes 153 mazan dyes have a tendency to secondary organization into coarse crystalline precipitates. Elementary tellurium is black or brown-black. The sections can be counterstained with hematoxylin or carmine; they should be mounted in glycerol or glycerol-jelly. b) Oxidases Oxidases are a motley group of enzymes having in com- mon the property of catalyzing the oxidation of various sub- strates, mainly phenols and amines, in the presence of at- mospheric oxygen. Their classification is very unsatisfactory. Substrate specificity is usually only relative; the same en- z}Tne will attack a number of substrates (although some of them more readily than others); and, conversely, the same substrate may be attacked by a number of different enzymes. It is impossible to go into the complicated and controversial details of the problem. The interested reader can seek more information in special textbooks of enzymology. Only a few enzymes of histochemical interest will be dealt with here. A) Indophenol oxidase {nodi oxidase; cytochrome oxi- dase ) .— A mixture of solutions of a phenol or naphthol and an aro- matic diamine is slowly oxidized on exposure to air, with the formation of intensely colored (usually blue) indophenol dyes, most of which are insoluble in water but very soluble in oils and fats. The reaction is immediate in the presence of strong oxi- dants, such as dichromate or hypochlorite. Ehrlich in 1885^^ showed by the injection of a mixture of alpha-naphthol and dimethyl-p-phenylene diamine into animals that the for- mation of indophenol blue is catalyzed by living tissues, in the absence of strong oxidants. Rohmann and Spitzer^^ found that ground-up tissues or even alcohol-ether-dried organ 29. Ehrlich, P.: Das SauerstoflFbediirfnis des Organismus (Berlin: A. Hirschwald, 1885). 30. Rohmann, F., and Spitzer, W.: Ber. d. deutsch. chem. Gesellsch., 28:567, 1895. 154 Microscopic Histochemistry powders had a strong catalytic effect. In 1907 Winkler^^ demonstrated that the same mixture, at an alkahne reaction, will stain pus cells in alcohol-fixed smears. Shortly after- ward Schultze^^ adapted the stain for use on tissue sections and emphasized the diagnostic value of the method in dis- tinguishing myeloid from lymphoid cells. From then on, for many years, the "indophenol oxidase" problem has been in the focus of interest. An extremely large number of papers has been published on its chemical background, its applica- tion, and the interpretation of its results. The term "nadi oxidase" (from a contraction of the first syllables of the two reagents ) , which has gained a wide acceptance, was coined by Graeff.^3 In 1922 von Gierke^^ noticed that the distribution of the reaction is not the same in fresh tissues treated with an un- alkalized reagent mixture as in fixed tissues treated with an alkalized one. Whereas in the first case the reaction is ex- tremely widespread throughout all tissues and not particu- larly intense in myeloid elements, in the second case it is limited almost exclusively to the granules of myeloid leuco- cytes, which stain very intensely. Furthermore, the diffuse reaction ( G for Gewebe = tissue nadi reaction of Graeff ) ^^ does not take place if freshly boiled distilled water is used as a solvent and is readily prevented by moderate heat (55° C.), by formalin, acetone, alcohol, acid, and alkali, etc. The myeloid reaction ( M reaction of Graeff ) ,^^ on the other hand, takes place even in oxygen-free water, is relatively heat-stable (up to 80° G.), insensitive to formalin (tissues kept in formalin for 6 years still give excellent reactions), to 31. Winkler, F.: Folia haemat., 4:323, 1907. 32. Schultze, W. H.: Beitr. z. path. Anat. u. z. allg. Path., 45:127, 1909; Schultze, W. H.: Miinchen. med. Wchnschr., 56:167, 1909; Schultze, W. H.: ibid., 57:2171, 1910; Schultze, W. H.: Zentralbl. f. allg. Path. u. path. Anat., 28:8, 1917. 33. Graeff, S.: Beitr. z. path. Anat. u. z. allg. Path., 70:1, 1922. 34. Gierke, E. von: Miinchen. med. Wchnschr., 58:2315, 1911. 35. Graeff, S.: Zentralbl. f. aUg. Path. u. path. Anat., 32:337, 1922. Enzymes 155 acetone, alcohol, alkali (actually, the optimum pH of the reaction is 12.5-13), very little sensitive to dichromate and Lugol's solution and many other agents which will promptly destroy practically all the known enzymes. However, it is sen- sitive to acids. In 1933 Keilin^^ came forward with the theory that the enzyme responsible for the G nadi reaction is identical with cytochrome oxidase. According to this theory, dimethyl-p- phenylene diamine is oxidized by cytochrome. Reduced cytochrome is then reoxidized by cytochrome oxidase in the presence of atmospheric oxygen. Keilin's ideas have become generally accepted, and the designation "cytochrome oxi- dase" has superseded the terms "labile" or "G" indophenol oxi- dase. Since cytochrome oxidase is a rather sensitive enzyme, readily destroyed even by drying and intolerant to formalin, etc., it is obvious that it cannot be responsible for the effects seen in the stable or M nadi reaction. First of all, by definition the stable enzyme is not an oxi- dase, since it is effective in the absence of atmospheric oxy- gen. In fact, it is altogether questionable whether enzymatic action is involved in the stable reaction. It is difficult to con- ceive of an enzyme which is resistant to formalin indefinitely, to Lugol's solution, and to dichromate for over 30 minutes. The writer was unable to confirm the sensitivity of the reac- tion to HgCL, reported repeatedly. He could verify the find- ings of Heringa and Ruyter:^^ blood smears after 30 minutes' exposure to a 2 per cent solution of HgCb, followed by re- moval of the metal with Lugol's solution, were indistinguish- able from those given by untreated smears. The exceedingly high pH optimum (around 13) is also an unlikely attribute of an enzyme. Graeff^^ and Katsunuma^^ voiced early opin- 36. Keilin, D.: Ergebn. d. Enzymforsch., 2:239, 1933. 37. Heringa, G. C, and Ruyter, J. H. C: Acta brev. Neerland., 5:118, 1935. 38. Katsunuma, S.: Intrazellulare Oxydation und Indophenolblausynthese (Jena: G. Fischer, 1924). 156 Microscopic Histochemistry ions in favor of the catalytic but nonenzymatic nature of the reaction. It is known that the formation of indophenol blue from the nadi reagent can be catalyzed nonenzymatically. The posi- tive reaction given by unsaturated fats has been mentioned before (p. 110). Remesow^^ obtained positive reactions with irradiated cholesterol. In model experiments filter-paper strips moistened with a solution of linseed oil or oleic acid in chloroform were allowed to dry for 12-24 hours. An in- tense blue color was observed after 1 minute's immersion in the nadi reagent. Schiimmelfeder^^ reported negative results with oleic acid in similar experiments. It must be assumed that he did not allow enough time for the formation of per- oxides. The presence of peroxides in oleic acid was proved by Lison.*^ He shook a solution of benzidine and hemoglobin with oleic acid (in place of H2O2 as the peroxide) and ob- tained the typical shade of benzidine blue in the aqueous phase, just as if hydrogen peroxide had been employed. Nadi-positive lipid substances have been prepared by sev- eral workers. ^^ The lipid nature of the granules of the mye- loid elements has been demonstrated repeatedly, and the morphological appearance of the Sudan-stained substance is identical with that of the nadi-positive granules. ^^ Sehrt^* could prove that sudanophilia and staining by the nadi reac- tion invariably run parallel; if a lipid solvent abolishes one, it will abolish the other. On the basis of these facts, it would seem that the enzy- matic theory is a superfluous assumption; the formation of 39. Remesow, I.: Virchows Arch. f. path. Anat., 285:591, 1932. 40. Schiimmelfeder, N.: Virchows Arch, f, path. Anat., 317:707, 1950. 41. Lison, L.: Bull. Soc. chim. biol., 18:185, 1936. 42. Neumann, A.: Arch. f. exper. Zellforsch., 6:298, 1928; Gutstein, M.r Biochem. Ztschr., 207:177, 1929; Magat, J.: Compt. rend. Soc. de biol., 116:1367, 1934. 43. Dietrich, A.: Ergebn. d. allg. Path. u. path. Anat., 13:283, 1909; Marinesco, G.: Compt. rend. Soc. de biol., 82:258, 1919. 44. Sehrt, E.: Miinchen. med. Wchnschr., 74:139, 1927. Enzymes 157 indophenol blue under the conditions of the nadi M reaction can be explained satisfactorily by the presence of fat perox- ides in the myeloid granules. The simultaneous presence of aldehydes would greatly accelerate the reaction.**^ The in- hibition by cyanide, often quoted as an evidence of the en- zymatic nature of the reaction, actually does not prove any- thing. Cyanide is known to poison many inorganic catalysts; moreover, it may act simply as a powerful reducer. In the model experiments mentioned, cyanide completely inhibited the staining of the paper slips and very greatly slowed down the spontaneous oxidation of the nadi mixture in air. This interpretation of the reaction renders the arguments meaningless as to whether the distribution of the reaction does or does not correspond to the true localization of the enzyme.*^ One version of the enzymatic theory, however, cannot be refuted with certainty. Although the reaction will take place even in the absence of any enzyme, a uniquely hardy per- oxidase, capable of utilizing fat peroxides, might be present and contribute its catalytic eflFect. Methods 1) For cytochrome oxidase {labile or G nadi reaction).— Use small fragments of fresh tissue or razor-blade slices. Frozen sections of fresh unfixed tissue can also be used, al- though freezing may cause partial inactivation of the en- zyme. To about 25 ml. of a phosphate buflFer of pH 7.2-7.6 add 1-2 ml. of a 1 per cent solution of alpha-naphthol in 40 per cent alcohol and the same amount of a 1 per cent solu- tion of dimethylparaphenylenediamine ( p-aminodimethyl- anilin) hydrochloride. Incubate in a shallow dish for about 10 minutes or until the reaction is clearly visible. Rinse sec- 45. Feigl, F.: Qualitative analysis by spot tests (New York: Elsevier Pub- lishing Co., 1946), p. 345. 46. Dietrich, A.: Zentralbl. f. allg. Path. u. path. Anat., 19:3, 1908; Hollande, A. C: Compt. rend. Acad, sc, 178:1215, 1924; Prenant, M.: Bull, d'histol. apphq. a la physiol., 2:329, 1925. 158 Microscopic Histochemistry tions; if desired, counterstain with lithium carmine. Mount in glycerin-jelly. The preparations are not permanent. A posi- tive reaction consists of a blue color. 2 ) For fat peroxides. ( a ) Stable or M nadi reaction.— Fix in formalin-alcohol or formalin. Frozen sections are prefer- able, but excellent results are often obtained with paraffin sections. Smears should be fixed in 95 per cent alcohol or formalin-alcohol (1:9). The incubating mixture is similar to the previous one, except for the pH. Instead of phosphate buffer, use distilled water, and add 2-3 drops of strong am- monia water or N NaOH to 50 ml. of the mixture. A positive reaction consists of a blue color. The results of this method can be checked by the {h) Glavind-Granados-Hartmann-Dam technique.'^"' Prepare two solutions: A, dissolve 40 mg. hematin in 10 ml. of pyridine, add 20 ml. of glacial acetic acid; B, dissolve 25 mg. of leuco- 2,6-dichlorophenolindophenol in 3.5 ml. of absolute alcohol, add 5 ml. distilled water. This latter solution is unstable and must be used fresh. Mix 0.75 ml. of solution A with the total of solution B and apply the mixture to the sections (or smears ) for 3-5 minutes. Wash in water, mount in glycerol- jelly. Fatty peroxides stain red. B) Nonspecific phenol oxidase.— Dark brownish or purplish staining of the granules of mye- loid leucocytes can be obtained by incubating alcohol- or formalin-alcohol-fixed smears with slightly alkaline solutions of various phenols, such as tyrosin, adrenalin, hydroqui- none,^^ alpha-naphthol,^^"^^ and dioxyphenylalanin.^^ The conditions of this reaction have not been investigated in de- 47. Glavind, I., Granados, H., Hartmann, S., and Dam, H.: Experientia, 5:84, 1949. 48. Kreibich, C: V\^ien. klin. Wchnschr., 23:701, 1910; Kreibich, C: ibid., 23:1443, 1910. 49. Loele, W.: Miinchen. med. Wclmschr., 57:1394, 1910. 50. Loele, W.: Die Phenolreaktion ( Aldaminreaktion ) und ihre Bedeu- tung fiir die Biologic (Leipzig: W. Klinkhardt, 1920). 51. Loele, W.: Virchows Arch. f. path. Anat., 262:39, 1926. 52. Bloch, B., and Peck, S. M.: Folia haemat., 41:166, 1930. Enzymes 159 tail, and no opinion can be expressed as to whether or not it is of an enzymatic nature. It must be remarked here that myeloid granules show peculiar reactions when incubated with phenolic substances (especially naphthols) under the conditions mentioned. If the solutions used are stale and somewhat colored by spon- taneous oxidation, an intense purple-black coloration of the granules becomes evident within 10 minutes. With fresh solutions, darkening of the granules is, absent or minimal; however, the binding of naphthol by the granules can easily be demonstrated by suitable reactions, such as azo-coupling or the Gibbs reaction (formation of indophenol blue on ex- posure to dichloro- or dibromoquinoneimide ) . Moreover, the naphthol-treated granules exhibit unusual staining reactions. They will stain intensely with basic dyes^^- ^^ and with fat dyes, such as Sudan III (de Bruijn).^^ Even in paraffin sec- tions, myeloid granules will stain intensely with Sudan III or Sudan Black B after 10 minutes' exposure to an alkalized, saturated aqueous solution of naphthol. The nature of this "phenolophilia" or "naphtholophilia"^^ is not clear (quinone- ethylenic condensation?). C ) Dopa oxidase.— In 1916 Bloch and Ryhiner^^ found that the basal layers of the epidermis, leucocytes, and other tissue elements con- tain an enzyme which oxidizes Z-3,4-dihydroxyphenylalanin (dopa) to a blackish pigment, indistinguishable from natural melanin. In subsequent papers'^^ Bloch offered good evidence that dopa is attacked by two different enzymes: (1) a non- specific phenol oxidase which is relatively resistant to various 53. Graham, G. S.: J. M. Research, 35:231, 1916-17. 54. De Bruijn, P. P. H.: Acta Neerland. morph. norm, et path., 2:232, 1939. 55. Bloch, B., and Ryhiner, P.: Ztschr. f. d. ges. exper. Med., 5:179, 191&-17. 56. Bloch, B., and Loffler, W.: Deutsches Arch. f. klin. Med., 121:262, 1916-17; Bloch, B.: Ztschr. f. physiol. Chem., 98:226, 1917; Bloch, B.: Arch. f. Dermat. u. Syph., 124:129, 1917; Bloch B.: Am. J. M. Sc, 177:609, 1929; Bloch, B., and Schaaf, F.: Klin. Wchnschr., 11:10, 1932. 160 Microscopic Histochemistry chemical and physical agents and is present in a number of structures, especially in myeloid leucocytes, and (2) an ab- solutely specific dopa oxidase which does not act on any other substrate, is easily inactivated by chemical and phys- ical agents, and is present only in cells concerned with the elaboration of melanin (basal layer of epidermis, chromato- phores, cells of melanoma, etc.). Although Bloch's theory was criticized by several authors ( Heudorf er,^^ Przibram^^ ) , it appears that the virtual specificity of the enzyme of mel- anin-producing cells in vertebrate species is a well-estab- lished fact. The only compound, besides dopa, to be attacked by this enzyme is oxytyramin ( Mulzer and Schmalf uss ) .^^ In insects and other species, other melanizing enzyme systems may be present ( Hasebroek ) ,^^ and even vertebrates possess other enzymes which will convert tyrosin and related com- pounds to melanin. The pathway of melanin formation by both types of enzymes, is essentially the same and leads through derivatives of indole^^ (ring-closure of the side chain ) . To rule out nonspecific phenol oxidases, it is advis- able to incubate control sections with a different substrate (e.g., naphthol or tyrosin). Method {Laidlaw and Bhckhergs^^ modification) Use frozen sections of fresh material or of tissue fixed for only a few hours in 5 per cent formalin. Longer fixation may cause partial inactivation of the enzyme. Rinse sections very briefly in distilled water and transfer them into a 0.1 per cent solution of dihydroxyphenylalanin (the commercial sub- stance is a mixture of the stereo-isomers ) , buffered with a phosphate buffer to pH 7.3-7.5, in an open dish for 4-5 57. Heudorfer, K.: Miinchen. med. Wchnschr., 68:266, 1921; Heudorfer, K.: Arch. f. Dermat. u. Syph., 134:339, 1921. 58. Przibram, H.: Arch. f. Entwcklngsmech. d. Organ., 48:140, 1921. 59. Mulzer, P., and Schmalfuss, H.: Med. Klin, 27:1099, 1931. 60. Hasebroek, K.: Fermentforsch., 5:1, 1922. 61. Duliere, W. L., and Raper, H. S.: Biochem. J., 24:239, 1930; Mason, H. S.: J. Biol. Chem., 168:433, 1947. 62. Laidlaw, G. F., and Blackberg, S. N.: Am. J. Path., 8:491, 1932. Enzymes 161 hours. Temperature should be between 20° and 37° C. It is advisable to change the incubating solution once or twice to avoid the deposition of a melanin precipitate (by spontane- ous oxidation of the substrate). At pH 7.7 the reaction is much faster (about 1 hour), but the danger of precipitates is also increased. Rinse sections, counterstain as desired, de- hydrate, and mount. Sites of dopa oxidase activity appear dark brown-gray or brown-black. For greater contrast, mel- anin formed during the reaction can be blackened by silver (see under "Pigments"). D) Amine oxidase.— This enzyme, originally described by Hare,^^ oxidizes ali- phatic and aromatic amines, such as isoamylamine, tyramine, histamine, tryptamine, etc., first to the corresponding alde- hydes and then to acids. Ammonia and hydrogen peroxide are formed in the course of the reaction. The corresponding amino acids ( tyrosin, histidine, etc. ) are not attacked. Oster and Schlossmann^^ described a method for the histo- chemical localization of amine oxidase, based on the demon- stration of the aldehyde formed. The detailed technique is as follows: Fresh-frozen sections are used. Since tissues contain plas- malogen, which, on standing, may break down to plasmal, an aldehydic compound, it is necessary first to transform plasmal into a nonreacting form to avoid its being mistaken for newly formed aldehyde. This is accomplished by treating the sections with a solution of bisulfite; the bisulfite addition compound of plasmal does not react with the usual aldehyde reagents. After thorough washing, the sections are incubated with the substrate (tyramine) and treated with Schiff's rea- gent. The aldehyde formed is demonstrated by the localized development of an indigo-blue shade. The theoretical foundation of this method is not sound. 63. Hare, M. L. C: Biochem. J., 22:968, 1928. 64. Oster, K. A., and Schlossman, N. C: J. Cell. & Comp. Physiol., 20:373, 1942. 162 Microscopic Histochemistry First of all, neither the enzyme*^^ nor the aldehyde formed ( p-hydroxyphenylacetaldehyde ) is insoluble enough to re- main at its original site; both would diffuse out into the solu- tion. Second, the oxidation of tyramine by amine oxidase goes beyond the aldehyde stage^^- ^^ unless the aldehyde formed is promptly trapped (e.g., by semicarbazide ) . In addition to the theoretical objections, it can be shown experimentally that positive reactions are obtained even in the absence of substrate and after the treatment of the sec- tions with chemicals which almost certainly destroy the enzyme (trichloroacetic acid). *^^ The chemical nature of the aldehyde giving the peculiar blue shade with Schiff's reagent is not known, but it appears to be either a modified form of plasmal or some other lipid aldehyde produced from unsaturated fats by nonenzymatic oxidation. It is not likely to be an aromatic compound, such as would be formed by enzymatic action, because the shades given by simple aromatic aldehydes are much more reddish. c) PEROXroASES Enzymes of the peroxidase group catalyze the transfer of oxygen from hydrogen peroxide and other peroxides to a variety of acceptors. Chemically, most if not all of the perox- idases appear to be heme proteins. They are quite resistant to various chemical and physical agents, especially to acids and heat. It must be remembered that some substrates may be at- tacked by peroxides (not necessarily all peroxides) directly^^ and that some of the catalysts are not of pyotein nature ( iion porphyrins; in some instances even inorganic Fe salts in con- junction with ascorbic acid).^^ Therefore, a reaction obtained 65. Blaschko, H., Richter, D., and Schlossmann, H.: Biochem. J., 31:2187, 1937. 66. Bemheim, M. L. C: J. Biol. Chem., 93:299, 1931. 67. Gomori, G.: Ann. New York Acad. Sc, 50:968, 1950. 68. Dixon, M.: Biochem. J., 28:2061, 1934. 69. Bezssonoff, N., and Leroux, H.: Bull. Soc. chim. biol, 28:286, 294, 608, 1946. Enzymes 163 with peroxide and a suitable substrate is not, in itself, a cogent proof for the presence of true peroxidase. Even when it is known that a morphological structure does contain per- oxidase (e.g., leucocytes), it is always questionable whether the reaction obtained under the conditions of the histochemi- cal experiment is due to enzymatic action or to a nonprotein catalyst ( "pseudo-peroxidase" ) . For the histochemical demonstration of peroxidase, benzi- dine,"^^ naphthol,^^ and various leuco-dyes are utilized^-' ^^ in the presence of hydrogen peroxide. Benzidine is oxidized to a blue ( quinhy drone ) or brown ( quinoneimine ) dye; naph- thol to a purple-black one, the chemical nature of which is not clear; leuco-dyes are recolorized to their original shades. It is important to run controls without peroxide because positive reactions may be obtained even in its absence (see previous section, "Phenol Oxidase" ) . The two main histological localizations of peroxidase are hemoglobin and myeloid granules. However, the reactions given by hemoglobin differ from those obtained in myeloid granules in several important points: 1. Zinc leuco-dyes are readily recolorized by hemoglobin but not by myeloid granules. 2. In the benzidine reaction the optimal concentration of H2O2 is about 0.01 M in the case of myeloid granules and much higher, about 0.1 M, in the case of hemoglobin. 3. The activity of myeloid granules is rapidly destroyed by heating to 75°-80° C. or by extraction with a warm chloro- form-methyl alcohol mixture. The activity of hemoglobin is entirely resistant to these influences. In model experiments, stale (peroxidized) linseed oil im- bibed in filterpaper strips gives rather intense reactions with benzidine but none with leuco-dyes or with naphthol. An- 70. Adler, O., and Adler, R.: Ztsclir. f. physiol. Chem., 41:59, 1904; Lepehne, G.: Beitr. z. path. Anat. u. z. allg. Path., 65:163, 1919; Good- pasture, E. W.: J. Lab. & Clin. Med., 4:442, 1919. 71. Loele, W.: Foha haemat., 14:26, 1912. 72. Lison, L.: Compt. rend. Soc. de bioL, 106:1266, 1931. 73. Jacoby, F.: J. Physiol., 103:25P, 1944-45. 164 Microscopic Histochemistry other interesting observation is the fairly intense staining of both myeloid granules and cell nuclei with peroxide-free acidified solutions of benzidine, prepared with tap water ( eflPect of free chlorine? ) . These facts clearly indicate that the oxidation of various substrates by peroxides can be promoted by more than one mechanism; in fact, even the invariable involvement of per- oxides in what is called a "peroxidase reaction" is not certain. The clarification of this problem will require much more investigative work. In histochemistry the most important application of the peroxidase reaction is the demonstration of hemoglobin. After hemolytic reactions, hemoglobin retains its enzymatic properties ( in renal tubules ) for only 24-36 hours, although its other staining reactions remain unchanged for several days. Methods Smears are fixed with acetone, alcohol, or formalin-alcohol (1:10). For tissues the same fixatives or formalin-saline are recommended. It is important that the fixative should not hemolyze the red cells, thereby causing a diffusion of the reaction for hemoglobin. Frozen sections, are the best, al- though in most cases excellent results are obtained after celloidin- or paraffin-embedding. A) Lisons'^^ zinc-leuco method gives the best results for fresh hemoglobin. Prepare the following staining solution: Dissolve 1 g. of either acid fuchsin, acid violet, or patent blue in 100 ml. of a 2 per cent acetic acid solution; add 5-10 g. of zinc dust, and boil mixture until the dye is decolorized (patent blue will bleach only to a pale brown). Add 2 more ml. of concentrated acetic acid. Keep the mixture in the ice- box. In the case of recolorization, heat briefly until decolor- ized once more. For use, filter 10 ml. and add 1 ml. of com- mercial (3 per cent ) hydrogen peroxide to it. Pour mixture over the slide. Hemoglobin will stain very intensely almost Enzymes 165 immediately, in the shade of the original dye. Counterstain as desired. Mount in balsam. B ) The benzidine method will stain even somewhat stale, degraded hemoglobin ( nonenzymatic heme catalysis?). Of the numerous modifications of this method, the technique of Osgood^^ and Washburn^^ is recommended. Prepare a 0.2-0.3 per cent solution of benzidine in 95 per cent alcohol. To each 100 ml. of the solution add about 0.5 g. Na nitroprusside dissolved in a few ml. of water (nitro- prusside appears to enhance blue shades in preference to brown ones). This stock solution will keep for months if refrigerated. When staining smears, pour the solution over the slide for 3 minutes. The alcoholic solution will serve as a fixative. This step may be omitted in the case of tissue sections. Decant. Mix equal volumes of the solution and a 1:5 dilution (for hemoglobin) or a 1:50 dilution (for myeloid granules) of commercial (3 per cent) hydrogen peroxide; pour it over the slide and leave it on for about 5 minutes. Decant, wash slide briefly in water, counterstain with a red nuclear stain, dehydrate, and mount in balsam. Myeloid granules appear intensely dark blue; hemoglobin appears in shades of dark brown to blue. The addition of a few drops of an acetate buffer of about pH 4.5 to the benzidine-H202 mixture will make hemoglobin stain in a clearer blue shade. The naphthol technique of Loele"^^ is not recommended. APPENDIX Unna's "Reduktionsorte" and "Sauerstofforte'* The terms Reduktionsorte ("sites of reduction") and Sauerstofforte ("sites of oxygen") were coined by the German dermatologist Unna*^^ to denote structures which reduce oxidizing substances or show the presence of "free" oxygen, respectively. The first are demonstrated by 74. Osgood, E. E.: Atlas of hematology (San Francisco: Stacey, 1937). 75. Washburn, A. H.: J. Lab. & Clin. Med., 14:246, 1928. 76. Unna, P. G.: Arch. f. mikr. Anat, 78:1, 1911; Unna, P. G.: Med. KHn., 8:951, 1912; Unna, P. G.: Berl. klin. Wchnschr., 50:809, 1913; Unna, P. G.: Abderhalden's Handb. d. biol. Arbeitsmeth., V-22:62, 1928. 166 Microscopic Histochemistry potassium permanganate (brown coloration) or by ferriferricyanide (precipitation of Prussian blue), the latter by the recolorization of leu CO methylene blue. The application of the modern redox-indicator dyes to the tissues*^"^ should give results far more delicately graded than the methods of Unna. However, it must be admitted that the oxidizability of organic compounds by permanganate or by ferricyanide is theoretically mean- ingful: it indicates the presence, in the structures stained, of com- pounds with oxidation-reduction potentials lower than those of the oxidants mentioned. Owing to its very powerful oxidizing properties, the specificity of permanganate is practically nil, since it is reduced by the majority of tissue components. Ferricyanide, being a much milder oxidant, does give a fairly selective staining of the more strongly re- ducing groups (-SH, polyphenols). On the other hand, the results of staining with leuco methylene blue cannot by any means be construed to reveal anything about the distribution of "free" oxygen (whatever this term may mean) or of oxidizing systems in the tissues. Unna stained sections in a solution of leuco methylene blue containing a large excess of the reducing agent (formaldehyde sulf oxylate ) . While the tissue is in this solution, it shows no indication of staining. On rinsing the sections in water, the color of methylene blue becomes apparent, and the picture is very similar to, although not absolutely identical with, that obtained with a regular solution of methylene blue. Obviously, the affinity of leuco-dye to basophilic structures is practically the same as that of the parent-dye, allowing for minor differences caused by structural changes in the dye molecule. When the excess of the reducer is removed by rinsing and the tissue is ex- posed to atmospheric oxygen, the latter will restore the bound leuco- dye to its original color. There is absolutely no need to assume any oxidizing activity residing in the tissue proper. Unna's work has been received wtih much criticism"^^ but has been defended stubbornly by the author, who ultimately resorted to such fanciful hypotheses as the "oxypolar affinity,'"^^ meaning some sort of chemical attraction between compounds rich in oxygen and poor in oxygen. A slight modification of Unna's leuco methylene blue technique was revived by Roskin.^^ as a diagnostic aid for the recognition of malig- 77. De Robertis, E., and Moura Gongalves, J.: Endocrinology, 36:245, 1945. 78. Oelze, F. W.: Arch. f. mikr. Anat., 84:91, 1914; Schneider, H.: Ztschr. f. wissensch. Mikr., 31:51, 478, 1914; Rothman, S.: Jadassohn's Handb. d. Haut- u. Geschlechtskrankh., I, 2:330, 1929. 79. Unna, P. G.: Arch. f. mikr. Anat., 87:96, 1915. 80. Roskin, G.: Ztschr. f. Krebsforsch., 35:140, 1931; Roskin, G.: Ztschr. f. Zellforsch. u. mikr. Anat., 14:781, 1931-32; Roskin, G., and Semenoff, W.: Enzymes 167 nant changes. Roskin and co-workers and Voinov^^ assert that malig- nant cells (in unfixed smears) do not stain with the dye while non- maHgnant ones do. Ludford^^ and the present writer^^ were unable to verify Roskin's claims. 2. HYDROLYTIC ENZYMES The hydrolytic enzymes demonstrable histochemically be- long, with the exception of two individual enzymes— phos- phamidase and glucuronidase— in the group of esterases; that is, they hydrolyze esteric linkages. In modern usage the term "esterase" is restricted to those enzymes with a substrate specificity for carboxyhc esters. The other ester-split dug en- zymes derive their names from the acid components of the esters they hydrolyze preferentially, since it appears that their specificity is determined by the acid moiety (phos- phatase, sulfatase). Methods for Hydrolytic Enzymes in General Depending on the substrate, hydrolysis yields an acid ion and an alcohol or a phenol ( or, in the case of phosphamidase, an acid and an amide). Reactions have been devised for the demonstration of either the acid or the alcoholic moiety. The acids are demonstrated by their regular precipitation reactions with metal ions. Theoretically, many cations could be used to trap the acid ions; in practice, however, only a few are suited for histochemical apphcation, the majority being too toxic to enzymes or incompatible with the sub- strate at the optimal pH of the reaction. The ones most often used are calcium and lead; cobalt, iron, and copper are employed in one method each. . The principle is to add a cation, known from analytical Ztschr. f. Zellforsch. u. mikr. Anat., 19:150, 1933; Roskin, G., and Solow- jewa, W.: Acta cancrol., 1:464, 1934-35; Roskin, G.: Bull. biol. et med. exper., 3:375, 1937; Roskin, G.: Bull, d'histol. appliq. a la physiol., 15:20, 1938; Roskin, G., and Struve, M. E.: Stain Technol., 22:83, 1947. 81. Voinov, V. A.: Klinicheskaia meditsina, 18:51, 1940. 82. Ludford, R. J.: Arch. f. exper. Zellforsch., 17:339, 1935. 83. Gomori, G.: Unpublished. 168 Microscopic Histochemistry chemistry to form a highly insoluble precipitate with the acid ion in question, to the incubating mixture. As the acid is liberated, its ions will be trapped in statu nascendi by the cation, to yield a precipitate at the site of formation. The precipitate formed is usually colorless and not easily seen under the microscope. Therefore, it must be trans- formed into a colored, easily observable compound. In the case of the heavier metals the sections can be treated directly with a suitable reagent. Soluble sulfides, for example, will transform precipitates of Pb, Co, Fe, and Cu into blackish, exceedingly insoluble sulfides. Fe and Cu can also be trans- formed into the corresponding f errocyanides ( blue and red- brown, respectively). When the primary precipitate is a Ca salt, it can be demonstrated by the methods mentioned in the section on calcium ( p. 33 ) . The alcoholic (or phenolic) moiety can be demonstrated only if it is a thio-alcohol or a naphthol. Some thio-alcohols form highly insoluble precipitates with heavy metals; naph- thols can be visualized as azo dyes. The azo-coupling reaction is rapidly becoming the basis of several excellent histochemical techniques and therefore de- serves a somewhat detailed discussion. Its application to in- soluble tissue phenols has been described in the section on phenolic substances (p. 120). In the case of more soluble substances, such as are produced by enzymatic hydrolysis, a number of other important points have to be considered. Under suitable conditions, diazonium salts will couple with aromatic amines and hydroxy compounds (and, in addi- tion, with a number of heterocyclic compounds) to form brightly colored, very insoluble azo dyes. Hydroxides (phe- nols and naphthols ) couple optimally at an alkaline reaction (pH 8 and up), whereas amines couple at an acid reaction (pH 3-5). This is only a general rule; while it is true that the coupling ability of phenols and naphthols rapidly de- clines as the neutral point is approached, there are a few types of naphthols which couple quite readily, almost at the Enzymes 169 rate of an ionic reaction, even at pH 6.3-6.5. Amines (naph- thylamines ) couple much more slowly, and the sluggishness of the reaction precludes the use of azo-coupling for their histochemical demonstration (diffusion artifacts). Any naphthol will not couple indiscriminately with any diazonium compound. Each new combination must be tried out in the test tube before its application to histochemistry can be considered. If coupling does take place, the color of the azo dye formed depends on both the naphtholic and the diazoic components used. The shade produced by any specific combination is not necessarily uniform but depends to some extent on factors such as the concentration of the ingredients, pH, temperature, etc. Azo dyes of phenols are not sufficiently insoluble for histo- chemical use; for this reason, only naphtholic compounds are employed. Azo dyes formed from ^-naphthol have very brilhant shades but have a tendency to remain in a colloidal solution for a short time before precipitating. This will cause some loss of fine structural details and false localizations (staining of acidophilic structures) around centers of high activity.^ Even the addition of high concentrations of salt to the incubating mixture (making the dye more insoluble by the "salting-out" effect) does not help much. There is also a possibility that the coupling process itself is a rela- tively slow one in the case of ^-naphthol, especially at pH values lower than about 8.5. On the other hand, a-naphthol and especially certain derivatives of ^-naphthol, e.g., the naphthols of the AS series (chemically, substituted amides of 2-hydroxy-3-naphthoic acid ) , give exceedingly sharp and detailed pictures. The shades given by a-naphthol are some- what dull; those given by the AS naphthols, very brilliant. Until recently, the only naphtholic substrates readily avail- able on the market were the acetate of a-naphthol and the benzoate of /5-naphthol. At present, practically all the im- portant substrates are made commercially. The Na salts of 1. Gomori, G.: J. Lab. & Clin. Med., 33:802, 1950. 170 Microscopic Histochemistry the phosphates and sulfates are quite soluble in water; esters of organic acids are very poorly soluble and require the ad- dition of some acetone or propylene glycol to the incubating mixture to obtain clear solutions. Fortunately, the substrate concentrations required are very low ( 3-10 mg. of crystalline substrate to a Coplin jar, or 0.00015-0.0005 M). Diazonium salts, which formerly had to be synthesized freshly for each experiment, are now available commercially in a wide variety and in a stabilized form. As dry powders, they can be kept in the icebox, if protected from light and moisture, almost indefinitely. However, their solutions are quite unstable and should be used immediately. Some of them keep better than others, but, on standing, even the most stable will decompose within about 2 hours. The rate of de- composition increases steeply with temperature and pH. The decomposition products are dark and have a tendency to stain the background, especially acidophilic structures, in a murky brownish shade. If incubation must be extended be- yond 15-20 minutes, it is advisable to cool the substrate mix- ture with ice and to employ the lowest possible pH value compatible with good enzymatic activity and prompt azo- coupling. If longer incubation is carried out, the entire mix- ture should be renewed every 30 minutes or so. Table 1 lists, diazonium compounds which have been found to be useful because they give intense shades and are rela- tively stable. The azo dyes formed from tetrazotized dianisidine and a- naphthol are insoluble in alcohol or xylol (but soluble in a mixture of the two; blot slide carefully between the last al- cohol and xylene ) and can be mounted in balsam. All other azo dyes formed from the ingredients mentioned are more or less soluble in absolute alcohol and especially in xylene. They must be mounted in glycerol- jelly or some similar medium. As mentioned, azo-coupling either does not take place at all or is sluggish below pH 6. The pictures obtained, even in the best case, lack precision of detail, and centers of activity Enzymes 171 are surrounded by a colored halo. To circumvent this diffi- culty, Seligman and associates^ have suggested the use of esters of highly insoluble naphthols (such as 6-bromo-^- naphthol), in the absence of a diazonium salt, whenever in- cubation must be carried out at a low pH. The precipitated naphthol could then be azo-coupled in a second step, at an alkaline reaction. The idea sounds feasible but does not work well in practice. Naphthols insoluble enough to precipitate with a sharp localization will not couple unless enough or- ganic solvent ( acetone, alcohol ) or strong alkali is added to effect some degree of solution. However, under such con- ditions, gross blurring of the picture will result. For this rea- TABLE 1 Composition and Trade-Name of Compound* Tetrazotized o-dianisidine (Diazo Blue B Salt) Diazotized naphthylamine Diazotized 4-cliloro-2-aminoanisole (Diazo Red RCSalt) • ••• Diazotized 3-nitro-4-aminoanisole (Diazo Bordeaux GP Salt) • ■ Diazotized 5-nitro-2-aminoanisole (Diazo Red B Salt) ,•• Diazotized 3-nitro-4-aininoanisole (Diazo Red G Salt) • ■• Diazotized 4-nitro-2-aminoanisole (Diazo Scarlet R Salt) ••• Diazotized ortho-amino-azotoluene (Diazo Garnet GBC Salt) .-..•• Diazotized 4-benzoylamino-2,5-dimethoxy- aniline (Diazo Fast Blue RR Salt) To Be Used with Naphthol a AS a AS a a a a Shade of Azo Dye Purple-black Purple Deep blue Black Carmine red Carmine red Red-brown Purple-brown Red-brown Red-brown Red-brown Red Carmine red Black Purplish red Purplish blue * The proprietary names of these compounds may vary slightly with individual manufacturers. However, if ordered by the trade-names listed above, they will be readily identified. 2. Seligman, A. M., Nachlas, M. M., Manheimer, L. H., Friedman, O. M., and Wolf, G.: Ann. Surg., 130:333, 1949. 172 Microscopic Histochemistry son, the azo dye technique is not recommended for the dem- onstration of enzymes which require incubation at a pH lower than 6.3. Esters of a- and ^-naphthol are usually so rapidly hydro- lyzed that diffusion artifacts result unless the incubating mix- ture is agitated all the time (p. 146). The simplest way of insuring thorough agitation is to use a low-speed electric or air-pressure stirrer with a long, narrow ( about /2 inch wide ) strip of stainless steel as a paddle, immersed almost to the bottom of the Coplin jar. Only two slides can be stained at one time; one in the first and one in the fifth slot of the Coplin jar; the tissues should face the stirrer. With esters of naph- thol-AS which are hydrolyzed relatively slowly, stirring is not necessary. Phosphatases The enzymatic hydrolysis of phosphoric esters by animal tissues was first described by Grosser and Husler^ in 1912. Ever since this original publication, the phosphatases have attracted a tremendous amount of scientific interest, and their literature has grown so vast that its mere review would fill a sizable volume. The classification of the phosphatases is not entirely satis- factory and is based partly on substrate specificity and partly on pH optima. A large number of individual phosphatases have been described; in some instances, however, the criteria of specificity are not sufficiently clear-cut. It should be remarked that, by a somewhat loose usage of terms, the existence of specific enzymes may be implied with- out any proof. For instance, the hydrolysis of glycerophos- phate in the neutral range has been reported under the title of "neutral glycerophosphatase," although, in all likelihood, it is due to the combined activities of nonspecific acid and alkaline phosphatases, measured at a nonoptimal pH. Even if such a possibihty is admitted later in the text, it would be more correct to avoid the use of such ambiguous terms. There 3. Grosser, P., and Husler, J.: Biochem. Ztschr., 39:1, 1912. Enzymes 173 is some overlapping between individual enzymes and groups of enzymes; thus alkaline phosphomonoesterase will hydro- lyze nucleic acids quite readily, although most of the phos- phate is in diesteric linkage; some pyrophosphatases will at- tack adenosinetriphosphate. Table 2 is essentially the classification of Folley and Kay,* expanded somewhat to include more recent data. It contains most but not all of the dephosphorylating enzymes whose specificities are reasonably well established. For more de- tailed information, the reader is referred to textbooks of enzymology.^ Of the large variety of phosphatases, histochemical meth- ods are available for nonspecific alkaline phosphatase, for 5- nucleotidase, acid phosphatase, and phosphamidase. A few data have been published on pyrophosphatase^' ^ and meta- phosphatase.^' '^ Glick and Fischer^ reported the histochemi- cal localization of adenosinetriphosphatase in plant and ani- mal tissue; but the specificity of their results has been ques- tioned.^ Soulairac and Desclaux^^ obtained positive results for adenosinetriphosphatase in rat muscle. Landschiitz^^ found that adenosinetriphosphatase can be demonstrated in the cells of the Ehrlich mouse carcinoma. No histochemical methods have been devised for the re- mainder of the group. Some of the phosphatases are so sen- sitive that they will not tolerate fixation and/or embedding (hexosediphosphatase, adenosinetriphosphatase of muscle). 4. Folley, S. J., and Kay, H. D.: Ergebn. d. Enzymforsch., 5:159, 1936. 5. Albers, H.: Phosphatasen. In Nord and Weidenhagen's Handb. d. Enzymologie (Leipzig: Akad. Verlagsgesellschaft., 1940), 1:408; Roche, J.: Phosphatases. In Sumner and Myrback^s The enzymes (New York: Aca- demic Press, 1950), I, 1:473. 6. Nickerson, W. J., Krugelis, E. J., and Andresen, N.: Nature, 162:192, 1948. 7. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 70:7, 1949. 8. GHck, D., and Fischer, E. E.: Science, 102:429, 1945. 9. Moog, F., and Steinbach, H. B.: Science, 103:144, 1946. 10. Soulairac, A., and Desclaux, P.: Compt. rend. Soc. de biol,, 143:470, 1949. 11. Landschiitz, C.: Experientia, 6:232, 1950. TABLE 2 Classification of Phosphatases Name of Enzyme and Site of Occurrence pH Optimum Substrate I. Phosphomonoesterases : a) Acid (Taka,^ yeast,^ liver") b) Acid (Spleen,*^ prostate,^ citrus fruitO 3-4 4.5-5.5 Around 6 6.5 Around 9 Around 9 Around 8 4.5,6.8, and 8.5 3.5-8.9 7 7.6 8.4 5.5and7.8 5.1-7.6 5-6 7 5 . 6-6 and 9 Glycerophosphate Various monoesters c) Neutral 1. Yeast,^ red cells^ Various monoesters 2. Glucose-6-phosphatase* Qiver) Glucose-6-phosphate d) Alkaline 1. Nonspecific (various organs) 2. Hexosediphosphatase (liver) J 3. 5-nucleotidase'' (nervous tis- sue, muscle) Various monoesters, nucleic acid Hexosediphosphate 5-nucleotides II. TIT Phosphodiesterases : Bran,' snake venom,™ serum*^. . Pyrophosphatase" Diesters (diphenylphosphate) Salts and esters of pyrophos- TV MetaohosDhataseP phoric acid Salts of metaphosphoric acid V. Unclassified special phosphatases: a) Adenosinetriphosphatase 1) Muscle^ Adenosinetriphosphate 2) Snake venom'' Adenosinetriphosphate b) Phytase^ Phytin c) Lecithinase* Lecithin d) Phosphoprotein phosphatase*^ e) Polyphosphatase'*^ Casein Salts of polyphosphoric acids VI. Phosphamidase^ (various organs) Substituted amides of phos- phoric acid » Inouye, K.: J. Biochem. (Japan), 10:133, 395, 1928. b Albers, H., and Albers, E.: Ztschr. f. physiol. Chem., 235:47, 1935. <> Bamann, E., and Salzer, W.: Biochem, Ztschr., 286: 147, 1936. d Davies, D. R.: Biochem. J., 28:529, 1934. « Kutscher, W., and Wolbergs, H.: Ztschr. f. physiol. Chem., 236:237, 1935. ' Axelrod, B.: J. Biol. Chem., 167:57, 1947. K Schaffner, A., and Bauer, E.: Ztschr. f. physiol. Chem., 232:66, 1935. t Roche, J.: Biochem. J., 25: 1724, 1931; Roche, J., and Bullinger, E.: Enzymologia, 7: 278, 1939. i Swanson, M. A.: J. Biol. Chem., 184:647, 1950. i Gomori, G.: J. Biol. Chem., 184:647, 1950; Roche, J., and Bouchilloux, S.: Bull. Soc. chim. biol,, 32:739 1950. ' k Reis, J.': Bull. Soc. chim. biol., 16:385, 1934; Reis, J.: Enzymologia, 2:110, 1937-38; Gulland, J. M., and Jackson, E. M.: Biochem. J., 32:597, 1938. • Imanishi, Y.: Biochem. Ztschr., 237:406, 1932; Uzawa, S.: J. Biochem. (Japan), 15: 11. 1932. ™ Uzawa, S.: J. Biochem. (Japan), 15:19, 1932; Gulland, J. M., and Jackson, E. M.: Biochem. J., 32:590, 1938. ° Plumel, M.: Bull. Soc. chim. biol., 30:55, 1948; Fleury, P., Courtois, J., and Plumel, M.: Bull. Soc. chim. biol., 32:40, 1950. 0 Lohmann, K,: Biochem. Ztschr., 203:172, 1928; Kurata, K.: J. Biochem. (Japan), 14:25, 1931; Ochiai, E.: Biochem. Ztschr., 253:185, 1932; Takahashi, H.: J. Biochem. (Japan), 16:447, 1932; Bamann, E., and Gall, H.: Biochem. Ztschr., 293: 1, 1937; Da Cunha, D. P.: Compt. rend. Soc. de biol., 124:1023, 1937; Schaffner, A., and Krumey, F.: Ztschr. f. physiol. Chem., 255:145, 1938; Rezek, A.: Chem. Abstr., 42:7808, 1948; Binkley, F., and Olson, C. K.: J. Biol. Chem., 186:725, 1950. p Kitasato, T.: Biochem. Ztschr., 201:206, 1928. 1 Bailey, K.: Biochem. J., 36: 121, 1942; DuBois, K. P., and Potter, "V. R.: J. Biol. Chem., 150: 185 1943. r'Zeller, E. A.: Experientia, 4:194, 1948; Zeller. E. A.: Helvet. chim. acta, 33:281, 1950. » Suzuki, U., Yoshimura, K., and Takaishi, M.: Bull. Coll. Agr. Tokyo Imp. Univ., 7:503, 1906- Enzymes 175 In other cases their substrate and pH requirements are such that no suitable precipitation reactions have been developed for their demonstration. Enzymes with pH optima around the neutral point present especially great dijfficulties because, on the one hand, Ca salts cannot be used efiFectively below pH ±: 8.5 and, on the other hand, the Pb salts of the sub- strates are not suflBciently soluble at a pH higher than ± 5. Alkaline Phosphatases Phosphatases with a pH optimum around 9 occur in most organs of almost all species examined. The largest group, that of the nonspecific enzyme ( s ) , will hydrolyze any monoester of phosphoric acid and, in addition, nucleic acids. High polymer, native desoxyribose nucleic acid is not attacked^^"^^ and must be depolymerized first if intended for use as a sub- strate. The rates of hydrolysis of various substrates vary over a wide range, and the optimal pH also depends on the nature of the substrate. As a rule, aromatic esters are hydrolyzed optimally at a higher pH (9.7-10) than aliphatic ones (8.1- 9).^^ All enzymes of the group are activated by Mg. They are inhibited by cyanide (except hexosediphosphatase,^^ which is activated but cannot be demonstrated histochemically ) . Cysteine is also a strong inhibitor. ^"^ Whether or not nonspecific alkaline phosphatase is a 12. KmgeHs, E. J.: Biol. Bull., 90:220, 1946. 13. Krugelis, E. J.: Genetics, 31:221, 1946. 14. Ross, M. H., and Ely, J. O.: J. Cell. & Comp. Physiol., 34:71, 1949. 15. King, E. J., and Delory, G. E.: Biochem. J., 33:1185, 1939. 16. Gomori, G.: J. Biol. Chem., 184:647, 1950. 17. Albers, H.: Ber. d. deutsch. chfem. Gesellsch., 68:1443, 1935; Hoff- mann-Ostenhof, O., Moser, H., and Putz, E.: Experientia, 4:352, 1948. 8; Plimmer, R. H. A.: Biochem. J., 7:43, 1913; Rapoport, S., Leva, E., and Guest, G. M.: J. Biol. Chem., 139:621, 1941; Courtois, J.: Bull. Soc. chim. biol., 30:37, 1948. *Zeller, E. A.: Enzymes as essential components of bacterial and animal toxins. In Sumner and Myrback's The enzymes I, 2:986 (New York: Academic Press, 1950). " Harris, D. L.: J. Biol. Chem., 165:541, 1946; Feinstein, R. N., and Volk, M. E.: J. Biol. Chem,, 177:339, 1949. V Neuberg, C, and Fischer, H. A.: Enzymologia, 2: 191, 241, 360, 1937-38; Frankenthal, L., Rob- erts, J. S., and Neuberg, C: Exper. Med. & Surg., 1:386, 1944. w Ichihara, M.: J. Biochem. (Japan), 18:87, 1933; Bredereck, H., and Geyer, E.: Ztschr. f. physiol. Chem., 254: 223, 1938; Lora Tamayo, M., and Martin Municio, A.: An. real Soc. espan. de fis. y quim., ser. B.47:149, 1951. 176 Microscopic Histochemistry single enzyme or a group of enzymes with closely similar, although not identical, properties has been a matter of con- siderable argument. There is a large body of chemical data available in support of the theory of plurality.^ Alkaline phos- phatases of different sources may show marked difiFerences in substrate preference and in their susceptibility to various activators and inhibitors. However, even if all the enzymes reported to exist do actually represent truly individual en- tities, the majority of them appear to be destroyed in the course of fixation and/or embedding. There are a consider- able number of reports asserting that the localization of ac- tivity depends on the substrate used and that the differences observed are due to the presence of several enzymes;^^^** ^'^ however, since the results have been obtained mostly under highly unfavorable conditions (prolonged incubation at low pH levels ) , the evidence cannot be accepted at face value. The slight variations due to different substrates may very well be considered the expression of differences in the rates of hydrolysis and in optimal pH values, the enzyme being identical in all cases ( see p. 172 ) . The writer found that un- der optimal or near-optimal conditions the pictures obtained with a large variety of substrates were indistinguishable from one another. Moog,^^ Ross and Ely,^* Zorzoli and Stowell,^^ and other s^^ have come to a similar conclusion. The only exception is 5-nucleotide : this substance is hydrolyzed by a specific enzyme. It should be emphasized specifically that 18. Deane, H. W., and Dempsey, E. W.: Anat. Rec, 94:456, 1946; Dempsey, E. V^., and Deane, H. W.: J. Cell. & Comp. Physiol., 27:159, 1946; Dempsey, E. W., and Singer, M.: Endocrinology, 38:270, 1946; Deane, H. W.: Am. J. Anat, 80:321, 1947; Dempsey, E. W., and Wislocld, G. B.: Am. J. Anat., 80:1, 1947; Siillmann, H.: Ztschr. f. Vitamin-, Hormon- u. Fermentforsch., 1:374, 1947-48; Lagerstedt, S., and Stenram, U.: Acta anat., 10:348, 1950; Newman, W., Feigin, I., Wolf, A., and Kabat, E. A.: Am. J. Path., 26:257, 1950. 19. Moog, F.: Biol. Bull., 86:51, 1944. 20. Zorzoli, A., and Stowell, R. E.: Anat. Rec, 97:495, 1947. 21. Bevelander, G., and Johnson, P. L.: Anat. Rec, 104:125, 1949; Hebert, S.: Arch, de biol., 61:235, 1950. Enzymes 177 results obtained with nucleic acids and with diesters of phos- phoric and pyrophosphoric acid as substrates (in the case of the last two substrates, by the azo dye technique, p. 184) are indistinguishable from those seen with the use of glyc- erophosphate,^^ for example. There is no histochemical indi- cation of the presence of diesterases or pyrophosphatases in paraffin-embedded animal tissues. According to Emmel,^^ renal phosphatase and intestinal phosphatase in the mouse can be differentiated from each other by their markedly different sensitivities to cyanide and acid. His findings have been verified,^^ but, since they do not apply to species other than the mouse,^^ they must not be interpreted as an indication of a general difference between renal and intestinal phosphatase. In view of the conflicting findings and opinions, the prob- lem of the presence of several substrate-specific alkaline phosphatases in embedded tissues requires critical re-ex- amination. Friedenwald and CrowelP^ and Maengwyn-Davies and Friedenwald^^ found, by the use of various substrates and activators, that fresh unfixed tissues contain substrate-spe- cific phosphatases which cannot be demonstrated in fixed tissues. This appears to be an important discovery which deserves thorough investigation. Technical details.— Freshness of the tissue is not a very important factor. Good results may be obtained with refrigerated tissue fixed as late 22. Gomori, G.: Unpublished. 23. Emmel, V. M.: Anat. Rec, 95^59, 1946; Emmel, V. M,: ibid., 96:423, 1946; Emmel, V. M.: ibid., 103:445, 1949; Emmel, V. M.: J. Nat. Cancer Inst., 10:1365, 1950; Emmel, V. M.: Proc. Soc. Exper. Biol. & Med., 75:114, 1950; Emmel, V. M.: Anat. Rec., 106:270, 1950. 24. Gomori, G.: Ann. New York Acad. Sc, 50:968, 1950. 25. Friedenwald, J. S., and Crowell, J. E.: Bull. Johns Hopkins Hosp., 84:658, 1949. 26. Maengwyn-Davies, G. D., and Friedenwald, J. S.: J. Nat. Cancer Inst, 10:1379, 1950. 178 Microscopic Histochemistry as 48 hours after removal, although there may be some blur- ring of the picture, owing to diffusion of the enzyme. Freezing-drying appears to preserve the enzyme much better than chemical fixation and embedding; the di£Ference is especially conspicuous in cases of low activity. TABLE 3 Fixation Formalin 80 per cent alcohol Absolute alcohol . . Acetone Percentage of^Enzymatic Activity Preserved Alkaline Phosphatase r 0-24^ 30« 65f /80« 75f \70g ± 100^ Acid Phosphatase ±30=* lO" 70« ±80b 20e 90b llo 178 22° a Emmel, V. M.: Anat. Rec, 95: 159, 1946. b Cappelin, M.: Bull, d'histol. appliq. a la physiol., 24: 155, 1947; Cappelin, M.: Monit. zool. ital., (suppl.) 56:256, 1948. ° Rabinovitch, M., Junqueira, L. C, and Fajer, A.: Stain Tech- no!., 24: 147, 1949. d Seligman, A. M., Chauncey, H. H., and Nachlas, M. M.: Stain Technol.,26:19, 1951. e Cappelin, M.: Bull, d'histol. appliq. a la physiol., 24: 155, 1947. « Doyle, W. L.: Proc. Soc. Exper. Biol. & Med., 69:43, 1948. B Stafford, R. O., and Atkinson, W. B.: Science, 107:279, 1948. h Cleland, K. W.: Proc. Linnean Soc. New South Wales, 75:35, 1950. All fixatives cause considerable inactivation of both alka- line and acid phosphatases. The effect of the procedures of fixation and embedding has been studied chemically by sev- eral workers, and the quantitative data are show in Table 3. Cappelin^'^ finds that fixation in chloroform (if chloroform can be called a fixative ) preserves 98 per cent of the activity of both acid and alkaline phosphatases. The discrepancy between the results of various workers is even more marked in reality than it would appear from 27. Cappelin, M.r Bull, d'histol. appliq. a la physiol., 24:155, 1947; Cappelin, M.: Monit. zool. ital. (suppl.), 56:256, 1948. Enzymes 179 Table 3, since many of the figures quoted are actually aver- ages, computed from widely divergent data. Some of the dif- ferences may be due to purely technical variations (tem- perature; length of exposure to the fixative; size of pieces used, and, consequently, rate of penetration by the fixative) and also to the different types of tissue used. The latter point is important, since the relative amounts of sensitive and re- sistant enzymes may vary considerably in different tissues. To mention only one example: v^hile a large percentage of the alkaline phosphatase activity of liver tissue is due to un- stable hexosediphosphatase, intestinal mucosa contains very little, if any, of this enzyme. If the effect of various fixatives is tested in Coujard sUdes, purified intestinal phosphatase being used as the enzyme, al- cohol and acetone are found to cause relatively little loss of activity (distinctly less than 50 per cent in 72 hours at 5° C); neutralized formalin destroys over 75 per cent of the enzyme in less than 24 hours at room temperatue but has very little effect at icebox temperature. Decalcification of tissues by the regular procedures v^ill destroy all phosphatases. Lorch^^ and Creep, Fischer, and Morse^^ report that small pieces of bone can be decalcified, after alcohol fixation, in a citrate buffer of pH 4.4-5 (expo- sure, several days). After decalcification, the tissue is reac- tivated around pH 9 (barbital buffer), dehydrated, and em- bedded. To v^hat extent the enzyme is preserved has not been determined, but enough activity remains to permit localization by the usual methods. Frozen sections show^ a higher activity than embedded tissues (DanielU);^^ hov^ever, there is a danger of loss of en- zyme by diffusion, since fixation in acetone or alcohol will not render the enzyme completely and irreversibly insoluble. Embedding in paraflfin causes a further loss in activity, esti- 28. Lorch, I. J.: Nature, 158:269, 1946. 29. Creep, R. O., Fischer, C. J., and Morse, A.: Science, 105:666, 1947; Creep, R. O., Fischer, C. J., and Morse, A.: J. Am. Dent. A., 36:427, 1948. 30. DanieUi, J. F.: J. Exper. Biol., 22:110, 1946. 180 Microscopic Histochemistry mated at 40-45,^^ 65,^^ and 75 per cent,^^ respectively. On the other hand, CappeHn^^ and Cleland^^ find that embed- ding causes very httle inactivation. In Coujard experiments, embedding did not inactivate the shdes noticeably except after formalin fixation ( ±50 per cent). No data are available on the effects of celloidin-embedding. ParaflBn blocks, in the writer's experience, retain their ac- tivity unchanged for over 10 years. Lison^* finds that they may deteriorate in 2 months. Cut ribbons can be kept in the icebox for many months; melted-on sections can be stored at room temperature for years without the slightest loss in activity. Danielli, Doyle,^^ and Lison, on the other hand, observed definite inactivation of the enzyme in stored paraffin sections. 1 ) The calcium phosphate method,— The first histochemi- cal procedure was published by Takamatsu^^ in 1938, in a journal not readily accessible to Western readers. In 19-39 the procedure was republished by Takamatsu^^ and simultane- ously described independently by Gomori.^^ It was based on the principle that, if sections are incubated with glycero- phosphate at an alkaline reaction in the presence of Ca ions, the phosphate ions liberated will be precipitated in statu nas- cendi ( at the site of formation ) as insoluble Ca phosphate. The latter is then transformed, in a second step, into metallic silver or black cobalt sulfide. The composition of the incubating mixture may be varied within fairly wide limits without much difference in the final results. The pH of the solution should be between 9 and 9.8 (in the lower ranges in the case of aliphatic substrates, in the higher ranges in the case of aromatic ones). Below pH 31. Doyle, W. L.: Proc. Soc. Exper. Biol. & Med., 69:43, 1948. 32. Stafford, R. O., and Atkinson, W. B.: Science, 107:279, 1948. 33. Cleland, K. W.: Proc. Linnean Soc. New South Wales, 75:35, 1950. 34. Lison, L.: Bull, d'histol. appliq. a la physiol., 25:23, 1948. 35. Takamatsu, H.: Manshu Igaku Zasshi, 31:34, 1938. 36. Takamatsu, H.: Tr. Soc. Path. Jap., 29:492, 1939. 37. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 42:23, 1939. Enzymes 181 9 the intensity of the reaction rapidly declines; only sites of highest activities will be stained after short periods, of incu- bation (up to 2 hours), and on greatly prolonged incubation diffusion artifacts may become very disturbing. Almost any buffer with a suitable pK can be used (NH4CI-NH4OH; bar- bital; 2-amino-2-methyl-l,3-propanediol). Borax is better avoided because it inhibits the hydrolysis of glycerophos- phate and of certain other substrates and because of its in- compatibility with higher concentrations of Ca. The concen- tration of the buffer should be 0.05-0.1 M. The cheapest and easiest available substrate is glycerophosphate, any commer- cial brand of which can be used; the recommended concen- tration, 0.01-0.03 M. Instead of glycerophosphate, any phos- phoric monoester, compatible with Ca at pH 9, can be used. In the case of unstable substrates, such as adenosinetriphos- phate or acyl phosphates, supersaturation artifacts are un- avoidable but can be recognized a such because they will be present even in inactivated sections. The importance of the concentration of Ca ions, has be- come appreciated only recently. It has been known for some time that nuclei will stain rather intensely around centers of high activity, and opinions have been expressed that this phenomenon is an artifact due to diffusion of Ca phosphate or of the enzyme itself (pp. 144-45). The experiments of Novikoff^^ and Gomori^^ have shown decisively that the alka- line phosphatase reaction of cell nuclei is due to the secondary adsorption of Ca phosphate only; the nuclei do not contain any enzyme either originally or by secondary adsorption ( ex- cept possibly in cases of poor fixation). The latter point is also proved by the invariable lack of any nuclear reaction with the azo dye method (Lorch,^^ Novikoff, and Gomori). In the Ca-Co method, the staining of nuclei is due to the relatively low concentration of Ca ions in the incubating 38. Novikoff, A. B.: Science, 113:320, 1951. 39. Gomori, G.: J. Lab. & Clin. Med., 37:526, 1951. 40. Lorch, J.: Quart. J. Micr. Sc, 88:159, 1947. 182 Microscopic Histochemistry mixture (0.01-0.02 M, Gomori; 0.03 M, Takamatsu). If this concentration is increased to about 0.1 M, nuclear reaction and other diffusion artifacts are completely eliminated unless incubation is unduly prolonged. Higher concentrations of Ca inhibit phosphatase activity. Ruyter^^ proposes the use of Mg in the presence of an am- monia buffer instead of Ca. The precipitate obtained will be magnesium ammonium phosphate. This method is not rec- ommended, because the precipitate is rather coarsely crystal- line and does not permit fine localization. However, the addi- tion of a small amount of Mg (around 0.005 M) to the in- cubating medium is useful because of its activating effect. Alkaline phosphatase is not too sensitive to minor varia- tions in temperature; incubation at any temperature between 30° and 45° C. will do. The length of incubation may be varied between wide limits, depending on the intensity of the reaction desired. However, it should be borne in mind that prolonged incuba- tion, even under optimal conditions, favors diffusion artffacts ( DanieUi,*' Gomori^' ) . It would be difficult to estabhsh a def- inite time limit beyond which it is not safe to go. In the writer's experience, after 4 or 5 hours false localizations be- come quite noticeable, and after 12-16 hours they may be widespread and intense. However, this complication need not arise except in special cases, since prolonging incubation beyond 2-3 hours is seldom, if ever, indicated. Although the precipitate is clearly visible in polarized light,'^-^ much sharper pictures are obtained by any of the color reactions mentioned under "Calcium." The best pic- tures are obtained with the use of silver and cobalt salts. Dorfman and Epshtein*^ prefer the Turnbull blue method 41. Ruyter, J. H. C, and Neumann, H.: Biochim. et biophys. acta, 3:125, 1949. 42. Danielli, J. F.: Nature, 165:762, 1950. 43. Belanger, L. F.: Proc. Soc. Exper. Biol. & Med., 77:266, 1951. 44. Dorfman, V. A., and Epshtein, S. M.: Doklady Akad. Nauk S.S.S.R., 72:977, 1950. Enzymes 183 (ferrous sulfate followed by f erricyanide ) . Boume^^ suggests staining with alizarinesulfonate. A very interesting method of visualization has been de- scribed by Dalgaard.^^ If glycerophosphate containing radio- active phosphorus is used, the precipitate can be demon- strated by its radioautogarph. The secondary introduction of radioactive lead into the precipitate is also possible.*^ The sensitivity of the method, as determined by the Cou- jard method, is about 20-30 ^LtM -units (0.5-1 Bodansky unit) per gram of active structure, the length of incubation being 1 hour. Proportionahty holds for the period of between 2-3 minutes and 12 hours. Sources of error.^^— a) False negative reactions (i.e., negative reactions in spite of the chemically verified presence of enzyme ) will be obtained whenever the concentration of the enzyme is below the threshold of sensitivity of the method. Inactivation of the enzyme by technical error is an unlikely complication, pro- vided that the method suggested is adhered to with reason- able accuracy. However, if for special reasons experimental conditions have to be changed considerably (use of low pH values, inhibitors, high ionic strength ) , negative results must be interpreted with caution. b) False positive reactions were briefly mentioned on page 148. Pigments can be seen in unincubated sections; some of them can be identified by specific reactions (hemo- siderin, ceroid, etc.). Preformed calcifications which will give reactions indistinguishable from true enzymatic ones can be removed before incubation by treating the sections for about 10 minutes with a citrate buflFer of pH 4.5-5. The adsorption of cobalt salt on tissue protein is not likely to cause any trouble. Artifacts due to the spontaneous or bac- 45. Bourne, G.: Quart. J. Exper. Physiol., 32:1, 1943. 46. Dalgaard, J. B.: Nature, 162:811, 1948. 47. Barka, T., Szalay, S., Posalaky, Z., and Kertesz, L: Kiserletes orvos- tud., p. 1, 1951. 48. Gomori, G.: J. Lab. & Clin. Med., 35:802, 1950. 184 Microscopic Histochemistry terial decomposition of substrates were discussed earlier. Ruyter^^ finds that positive reactions may be obtained after prolonged incubation in alkaline solutions of Ca salts not containing any substrate. The writer was unable to duplicate this observation. c) False localizations have been dealt with in detail (pp. 142 ff.). Method recommended Fixation and embedding (see pp. 138 ff. ) . It is advisable, al- though not strictly necessary, to protect slides with collodion. Composition of the incubating mixture: 3 per cent (±0.1 M) solution of Na glyc- erophosphate 5-10 ml. 2 per cent (±0.2 M) solution of calcium chloride 20-25 ml. 10 per cent (±0.5 M) solution of magnesi- um chloride (or sulfate) About 10 drops Na barbital (powder) 1 knifepointful (0.5-1 g.) Distilled water to make 50 ml. Instead of barbital, other buffers (p. 221) may be used. Should the mixture be turbid ( presence of free phosphate in the substrate), it must be filtered before use. Incubate sections around 37° C. for 1-4 hours. Wash them under the tap for about 1 minute. Immerse in a 1-2 per cent solution of any soluble cobalt salt (acetate, chloride, sulfate, nitrate) for 5 minutes. Wash under the tap for 1-2 minutes. Immerse in a dilute solution of colorless or light-yellow ammonium sulfide ( a few drops to a Coplin jarful of distilled water) for 5 minutes. Wash under the tap. Counterstain as desired. Dehydrate and mount. Sites of alkaline phosphatase activity appear in black. 2 ) The azo dye method.— The original form of this method was devised by Menten, Junge, and Green^^ in 1944; modi- 49. Menten, M. L., Junge, J., and Green, M. H.: J. Biol. Chem., 153:471, 1944; Menten, M. L., Junge, J., and Green, M. H.: Proc. Soc. Exper. Biol. & Med., 57:82, 1944. Enzymes 185 fications were published by Manheimer and Seligman,^^ by Loveless and Danielli,^^ and by Gomori.^^ The first two methods use /3-naphthyl phosphate as a substrate; Loveless and Danielli, a complicated azo dye phosphate; Gomori, a- naphthyl phosphate. The method of Gomori is the simplest; it employs only commerically available chemicals and does not require cooling by ice. The pictures obtained with it are remarkably sharp and much more detailed than those given by methods using ^-naphthyl phosphate. Method Fixation and embedding as in the Ca-Co method. Protection with collodion is better avoided because the membrane may be stained rather intensely by decomposi- tion products. Composition of the incubating mixture.— Dissolve about 10 mg. of Na a-naphthyl phosphate ( available from Dajac Lab- oratories, 511 Lancaster Ave., Leominster, Mass.) in a few ml. of distilled water. Add a few ml. of a 4-5 per cent solu- tion of borax (Na2B4O7*10H2O), about 40 ml. of cool dis- tilled water (not warmer than 20° C.), and a few drops of a 10 per cent solution of magensium chloride or sulfate. Stir into the mixture 20-50 mg. of any of the following diazonium salts: Blue B, Red RC, Bordeaux GP, Red G, or Fast Blue RR. Incubate the slides for 10 to 30 minutes or until the desired intensity of staining is obtained. The slides may be removed from the incubating mixture and inspected under the micro- scope repeatedly. Mechanical stirring of the solution during incubation is highly advisable ( p. 172 ) . Wash slides; counters tain with hematoxylin or with alum carmine; diflFerentiate with acid alcohol (alcohol concentra- tion, 70-80 per cent ) ; wash again and either mount in glyc- erol-jelly or dehydrate and mount in balsam (p. 170). Sites 50. Manheimer, L. H., and Seligman, A. M.: J. Nat. Cancer Inst., 9:181, 1948. 51. Loveless, A., and Danielli, J. F.: Quart. J. Micr. Sc., 90:57, 1949. 186 Microscopic Histochemistry of activity appear in shades listed in Table 1 (p. 171); back- ground, yellowish. The only possible source of error is the staining of entero- chromaffin cells in a reddish shade, but only after formalin fixation. The azo dye method (one or the red color variants), fol- lowed by Kossa's technique for Ca, is very suitable for the simultaneous demonstration of preformed calcifications and of sites of phosphatase activity. It should supersede the older cobalt-suLfide-lead-acridine-red method.^^ 5- Nucleotidase The existence of the enzyme 5-nucleotidase has, been re- ported by Reis^^ and by GuUand and Jackson,^* on the basis of chemical studies. The substrates of the enzyme are 5- nucleotides (muscle adenylic acid, inosinic acid, and, possi- bly, adenosinetriphosphoric acid). The pH optimum is around 7.8, but the enzyme is quite active even at pH 9. The histochemical method^^ for 5-nucleotidase is very sim- ilar to the method for alkaline phosphatase; in fact, it may be identical with it except for the substrate. However, if sharp differentiation from alkahne phosphatase is desired, it is better to perform incubation at pH ± 8.3. At this pH, 5- nucleotidase is fully active, while the activity of alkaline phosphatase is only about one-third of the maximum. A slight disadvantage of this low pH is a tendency toward diffusion artifacts. It can be offset almost completely by a sufficiently high concentration of Ca ions. The substrate used is muscle adenylic acid^^ (regular 52. Gomori, G.: Am. J. Path., 19:197, 1943. 53. Reis, J.: Bull. Soc. chim. biol., 16:385, 1934; Reis, J.: Enzymologia, 2:110, 183, 1937-38; Reis, J.: ibid., 5:251, 1938-39; Reis, J.: ibid., 2:183, 1937-38; Reis, J.: Bull. Soc. chim. biol., 22:30, 1940; Reis, J.: Biochem. J., 46:xxi, 1950; Reis, J.: ibid., 48:548, 1951. 54. Gulland, J. M., and Jackson, E. M.: Biochem. J., 32:597, 1938. 55. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 72:449, 1949. 56. Available from the Schwarz Laboratories, 202 E. 44th St., New York 17; from Nutritional Biochemicals, Cleveland 28, and from the Sigma Chemi- cal Co., 4648 Easton Ave., St. Louis 13. Enzymes 187 adenylic acid is a 3-nucleotide ) . For histochemical purposes the cheaper grades appear to be just as good as the more expensive ones. Method Dissolve about 20 mg. of muscle adenylic acid in 20 ml. of a 0.1-0.2 barbital or tris ( hydroxymethyl ) -aminomethane buffer of pH 8.3. Add 20 ml. of a 2 per cent CaCl2 solution and a few drops of a 10 per cent MgCL solution. Incubate sections for 2-5 hours at 37° C. Convert Ca phosphate pre- cipitate into cobalt sulfide, as in the method for alkaline phosphatase. The reaction obtained with 5-nucleotide is more wide- spread than that given by glycerophosphate or 3-nucleotides and includes, in addition to the sites of activity of nonspecific alkaline phosphatase, certain tracts in the central nervous system and the smooth muscle of blood vessels and of the urinary bladder. The findings appear to be compatible with the assumption that, while nonspecific alkaline phosphatase will attack both glycerophosphate and the two nucleotides, 5-nucleotidase cannot hydrolyze substrates other than 5-nucleotide. Lecithinase ( Phospholipase ) Lecithin possesses 4 esteric linkages, and specific enzymes hydrolyzing each one of the linkages have been demon- strated. A H2C— i— O— OCRi B HC— j— O— OCR2 C D -f- H2C— i— OPOO— i— CH2CH2N (CH3)3 The enzymes attacking bonds A and B belong among the lipases, while those attacking bonds C and D are phospha- tases or, more exactly, phosphodiesterases. For detailed in- 188 Microscopic Histochemistry formation and bibliography the reader is referred to the ex- cellent review by Zeller.^^ When paraffin sections of animal tissues are incubated with a "solution" of lecithin ( or cephalin ) in the presence of Ca ions around pH 7, no reaction whatsoever is obsei'ved, even on prolonged incubation (up to 12 hours ).^^ At this pH, the Ca salts of fatty acids, if any were liberated, would precipitate promptly (see the Tween technique, p. 203), while Ca phosphate would not, unless the formation of phos- phate ions were very rapid. At pH 9 a picture essentially iden- tical with the pattern of distribution of alkaline phosphatase is obtained.^^ It should be mentioned here that Dempsey and Deane,^^ and Dempsey and Wislocki^^ did obtain posi- tive reactions with lecithin around the neutral range but only after greatly prolonged incubation (24-72 hours). The arti- facts produced under such conditions have been discussed (p. 176). The findings seem to indicate that the enzymes responsible for the hydrolysis of bonds A and B cannot be demonstrated histochemically. This is surprising in view of the fact that such enzymes are known to occur in many animal tissues and that they are quite resistant to heat. The enzyme active at pH 9 and demonstrated histochemically acts either as leci- thinase C or as lecithinase D. In the first case, the primary precipitate would be the Ca salt of phosphorylcholine; in the second case, Ca phosphatidate. Whether the primary pre- cipitate is further hydrolyzed to Ca phosphate cannot be decided on the basis of data available. The interesting point is that, just as in the case of the sim- 57. Zeller, E. A.: Enzymes as essential components of bacterial and animal toxins. In Sumner and Myrback's The enzymes (New York: Academic Press, 1951). 58. Gomori, G.: Unpublished. 59. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 70:7, 1949. 60. Dempsey, E. W., and Deane, H. W.: J. Cell. & Gomp. Physiol., 27:159, 1946. 61. Dempsey, E. W., and Wislocki, G. B.: Am. J. Anat., 80:1, 1947. Enzymes 189 pier phosphodiesters (p. 177), the distribution of activity is identical with that of alkaline phosphatase. It appears very likely that the nonspecific enzyme is responsible for the histo- chemical reactions obtained with the use of phosphodiester substrates. Acid Phosphatase Phosphatases with very low pH optima have been de- scribed in fungi and in higher plants. A phosphatase with a pH optimum of ±: 5 was found in the spleen and liver by Davies^^ and later in the prostate by Kutscher.^^ Acid phos- phatases, as a rule, are not activated by Mg and almost in- variably are greatly inhibited by fluoride. Enzymes of animal origin may show considerable differences in respect to their behavior toward inhibitors, as shown in Table 4. TABLE 4 The Inhibition of Acid Phosphatases by Various Substances* Plasma Red Cells Adrenal Bile Prostate Alcohol 3* 3-4 f' '^ 0ci 2b 3d 4« 0_le, f,K,h 4b, i Only below pH 4.6; 0 at pH 5d. e 0b lb 4b 2b 3b, 0 Bile acid Cu++ F- Formalin Tartrate 3d 0e 4« 0b, i 4d,e, j *0 indicates no inhibition; 4, maximum inhibition; 1-3, intermediate degrees; — indicates no data available. a Gutman, E. B., and Gutman, A. B.: J. Clin. Investigation, 17:473, 1938. b Abul-Fadl, M. A. M., and King, E. J.: J. Clin. Path., 1:80, 1948. 0 Herbert, F. K.: Biochem. J., 38:23, 1944. d Abul-Fadl, M. A. M., and King, E. J.: Biochem. J., 42:28, 1948. e Abul-Fadl, M. A. M., and King, E. J.: Biochem. J., 45:51, 1949. f Gutman, E. B., and Gutman, A. B.: Proc. Soc. Exper. Biol. & Med., 47:513, 1941. 8 Behrendt, H.: Proc. Soc. Exper. Biol. & Med., 54:268, 1943. h King, E. J., Wood, E. J., and Delory, G. E.: Biochem. J., 39:24, 1945. > Abul-Fadl, M. A. M., and King, E. J.: J. Path. & Bact., 60: 149, 1948. i Seligman, A. M., and Manheimer, L. H.: J. Nat. Cancer Inst., 9:427, 1949; Seligman, A. M., Nachlas, M. M., Manheimer, L. H., Friedman, O. M., and Wolf, G.: Ann. Surg., 130:333, 1949. 62. Davies, D. R.: Biochem. J., 28:529, 1934. 63. Kutscher, W., and Wolbergs, H.: Ztschr. f. physiol. Chem., 236:237, 1935. 190 Microscopic Histochemistry In addition to the drfferences tabulated, it appears that the ratio Hydrolysis of phenylphosphate Hydrolysis of glycerophosphate is very much higher for the enzyme of red cells than for that of the plasma.^* The original histochemical method^^ for acid phosphatase utilizes the hydrolysis, of glycerophosphate at pH 5 in the presence of Pb++ ions. Fixation in cold acetone and em- bedding in paraffin are recommended. The results of this technique are not nearly so satisfactory or consistent as, those of the method for alkaline phospha- tase. A number of papers complaining of its capriciousness and unreliabihty have been published. ^^^^ When the causes of failures are analyzed, they are found to fall into six groups. 1. Inactivation of the enzyme in the course of fixation, embedding, storage, and incubation. There can be no doubt that only a relatively small fraction of the enzyme survives fixation and embedding. Inactivation may be fairly uniform or peculiarly patchy; the boundary between active and inac- tive areas is often quite sharp. Wolf, Kabat, and Newman^^ find that a layer at a certain distance from the surface of the block shows the best preservation of the enzyme. The writer is under the impression that embedding and not fixation (provided that cold acetone is used) is the main offender; 64. Behrendt, H.: Proc. Soc. Exper. Biol. & Med., 54:268, 1943; Gutman, E. B., and Gutman, A. B.: Proc. Soc. Exper. Biol. & Med., 47:513, 1941; Gutman, E. B., and Gutman, A. B.: J. Clin. Investigation, 17:473, 1938; Herbert, F. K.: Biochem. J., 38:23, 1944; King, E. J., W^ood, E. J., and Delory, G. E.: Biochem. J., 39:24, 1945. 65. Gomori, G.: Arch. Path., 32:189, 1941. 66. Moog, F.: Proc. Nat. Acad. Sc, 29:176, 1943. 67. Hard, W. L., and Lassek, A. M.: J. Neurophysiol., 9:121, 1946. 68. Montagna, W., Nobaclc, C. R., and Zak, F. G.: Am. J. Anat., 83:409, 1948; Tissieres, A.: Acta anat., 5:224, 1948. 69. W^olf, A., Kabat, E. A., and Newman, W.: Am. J. Path., 19:423, 1943. Enzymes 191 especially a high temperature of the paraflBn oven and long exposure of the tissue to hot paraflBn seem to be harmful. According to several authors (Hard and Lassek,^^ Doyle J^ Goetsch and Reynolds"^^), the enzyme is gradually inacti- vated on storage of paraflBn blocks, ribbons, or mounted sections. Doyle asserts that lead ions inhibit the enzyme very mark- edly (about 85 per cent). Using the regular histochemical substrate mixture, the writer found about 35 per cent inhibi- tion by lead in test-tube experiments. 2. The use of unfixed or partially fixed, unembedded tis- sue. Frozen sections may show a relatively high activity, but the aflBnity of undenatured proteins to lead salts is likely to produce false positive reactions. Lead adsorbed by paraflBn- embedded tissues can be washed out readily by dilute acetic acid (exception under 4); however, frozen sections of un- fixed or acetone-fixed tissues will hold lead so stubbornly that even prolonged washing in strong acetic acid cannot remove it completely. Formalin-fixed protein has much less aflBnity to lead. 3. Failure to use an acid rinse after incubation. Even de- natured proteins may adsorb some lead from the substrate mixture but will release it easily when rinsed in dilute acetic acid. The acid rinse is an important step; it removes protein- bound lead but leaves enzymatically produced lead phos- phate untouched. If it is omitted (as it apparently was by Newman, Kabat, and Wolf^^ and by Takeuchi and Ta> noue^^), the result will be a nonenzymatic staining of the background, mainly of nuclei. . 4. Certain structures such as axons retain a high aflBnity to lead even after paraflfin-embedding; in fact, axons can be 70. Doyle, W. L.: Proc. Soc. Exper. Biol. & Med., 69:43, 1948. 71. Goetsch, J. B., and Reynolds, P. M.: Stain Technol., 26:145, 1951. 72. Newman, W., Kabat, E. A., and Wolf, A.: Am. J. Path., 26:489, 1950. 73. Takeuchi, T., and Tanoue, M.: Kumamoto M. J., 4:41, 1951. 192 Microscopic Histochemistry stained by lead impregnation ( Mallory'^^ ) . Nonenzymatic staining of axons by lead can easily be mistaken for an en- zymatic reaction. "^^ Control sections with either inactivated ( by Lugol's solution or by boiling water ) tissue or with active tissue incubated in the presence of 0.005 M fluoride must be used to detect nonenzymatic impregnation; it may be ex- tremely difficult or impossible to determine whether the pic- ture obtained is not a composite of both enzymatic and nonenzymatic staining. 5. Nonoptimal composition of the incubating mixture. The effect of the composition of the substrate mixture on the constancy of the results has been studied by Gomori, who found that the ratio between the concentrations of buf- fer and substrate is an important factor. An unduly high concentration of buffer will greatly reduce the sensitivity of the method, thereby giving rise to "all or none" phenomena. The original formula is definitely nonoptimal, and there can be no doubt that some of the failures obtained with it are due to this fact. 6. Undertermined factors. This group includes the queer- est types of failure, unexplainable by any of the causes enu- merated. To mention only a few examples: sections cut from a single block stain uniformly one day but utterly refuse to stain the next day, although the incubating mixture is pre- pared in exactly the same way both times; out of a ribbon of several serial sections, mounted on one slide, one or two stain exceedingly poorly while the majority show an excellent reaction. Fixation of thin blocks in cold acetone, rapid embedding at a temperature not exceeding 56° C, and the use of re- cently cut sections and of the correct substrate mixture will produce good results in a vast majority of instances; however, occasional unexplainable failures cannot be eliminated com- pletely. 74. Mallory, F. B.: Am. J. Path., 12:569, 1936. 75. Bartelmez, G. W., and Bensley, S. H.: Science, 106:639, 1947; Heinzen, B.: Anat. Rec, 98:193, 1947; Lassek, A. M.: Stain Technol., 22:133, 1947. Enzymes 193 1) The lead method."'^— Composition of the substrate mix- ture: In 500 ml. of a 0.05 M acetate buffer of pH 5 dissolve 0.6 g. of lead nitrate (about 0.003 M) and add 50 ml. of a 3 per cent (about 0.1 M) solution of Na glycerophosphate. The mixture will become turbid; the degree of turbidity de- pends on the percentage of y8-isomer contained in the brand of glycerophosphate used. A mixture of about equal parts of the two isomers will cause much less turbidity than the more commonly sold mixtures containing around 75 per cent of y8-salt. Keep the solution in the incubator at 37° C. for 24 hours, filter it. Add a small amount (about 5 per cent) of distilled water to the filtrate to prevent precipitation on evaporation. The mixture is ready for use and will keep in the icebox for months. If it becomes turbid, it should be discarded. Carry sections through xylene and alcohols to water. Col- lodion coating of the sections is advisable; it appears to pre- vent loss of enzyme by diffusion (Doyle^^), especially at sites of low activity, although it may cause some patchiness of the reaction ( Goetsch and Reynolds"^^ ) . Rinse slides thor- oughly in distilled water. Incubate in the substrate mixture around 37° C. for 1-24 hours. Human prostate usually requires 1-^2 hours of incuba- tion, other tissues 6-S hours or more. Sites of activity will become a chalky white from the deposition of lead phos- phate. Rinse slide first in distilled water, then for a minute or so in 1-2 per cent acetic acid and once more in distilled water. Immerse slide in a dilute solution of ammonium sul- fide (a few drops to a Coplin jarful of distilled water) for about 2 minutes. Wash under the tap. Counterstain as de- sired. Dehydrate in alcohols; clear in gasoline or tetrachloro- ethylene, and mount in clarite or some similar resin dissolved in the same solvents. Toluene or xylene should not be used; they will cause some fading of the stain. Sites of activity are indicated by the dark brown -black precipitate of PbS. 76. Gomori, G.: Stain Techno!., 25:81, 1950; Wang, K. J.: Chinese J. Physiol., 17:317, 1950. 194 Microscopic Histochemistry For special purposes the method can be modified by changing the pH of the solution or by using other substrates. Above pH 6 the activity of alkaline phosphatase rapidly in- creases, and one may obtain combination pictures of the dis- tributions of acid and alkaline phosphatase. The choice of substrates is rather limited because the lead salts of most phosphoric esters are very insoluble at pH 5 or higher. Res- orcinol phosphate and adenosinetriphosphate are suitable substrates; they give pictures which differ more or less mark- edly from those obtained with glycerophosphate."^^ This may be an indication of the existence of more than one acid phos- phatase in the tissues, but the point will require further in- vestigation. Abolins^^ finds that the pattern of distribution of activity in the anterior pituitary varies with the pH of the substrate solution, and he attributes the differences to the presence of several enzymes. Sources of error are the same as in the method for alkaline phosphatase, with the addition of impregnation artifacts of nerve tissue. The sensitivity of the method ( Coujard's method ) is about the same as that of the technique for alkaline phosphatase. 2) The azo dye method. —Seiigman and Manheimer"^^ rec- ommend a-naphthyl phosphate as a substrate, in the pres- ence of diazotized aminoanthraquinone. Sites of activity are shown in a reddish-brown shade. This method is not recom- mended because of unavoidable gross diffusion artifacts. Phosphamedase An enzyme hydrolyzing phosphamides has been described in animal and plant tissues. "^^ In animals its natural substrates are probably phosphocreatine, phosphoarginine, and pos- sibly some other less-well-known compounds possessing an N-P bond. 77. Abolins, L.: Nature, 164:455, 1949. 78. Seligman, A. M., and Manheimer, L. H.: J. Nat. Cancer Inst., 9:427, 1949. 79. Waldschmidt-Leitz, E., and Kohler, F.: Biochem. Ztschr., 258:360, 1933; Ichihara, M.: J. Biochem. (Japan), 18:87, 1933; Bredereck, H., and Geyer, E.: Ztschr. f. physiol. Chem., 254:223, 1938. Enzymes 195 For the histochemical demonstration of phosphamidase the substrate used is phophoric acid p-chloroaniHde (p- chloroaniHdophosphonic acid). This compound is relatively stable and easier to prepare than the other phosphamides. There is some doubt concerning its correct formula; Rorig^^ believes that the compound synthesized by the method of Otto^^ is actually a derivative of diamidophosphoric acid. It appears that different batches prepared by the same method are not necessarily identical in composition. The data on the optimal conditions of activity (pH, acti- vators, etc.) of this enzyme are rather vague. The histo- chemical method to be described has been worked out by the method of trial and error only. Typical pictures are ob- tained only in the pH range between 5 and 6; in the alkaline range, chloroaniHdophosphonic acid as a substrate gives dis- tribution patterns indistinguishable from the regular alkaline phosphatase reaction. Takamatsu and Sho^^ have used phos- phocreatine as a substrate at an alkaline pH; the description of the results obtained is not clear enough to permit the drawing of conclusions. Method^^ Fixation in acetone or alcohol. Sections should not be coated with collodion. Dissolve 1 g. p-chloroanilidophosphonic acid^^ in an ice- cold mixture of 1 ml. of concentrated ammonia and 40 ml. of distilled water with vigorous shaking. The substance usu- ally does not dissolve completely. Filter. Add 2 drops of a 0.5 per cent phenolphthalein solution. Titrate back with M acetic acid to a barely perceptible pink shade (about 3 ml. of acid will be required). Fill up to 50 ml. with distilled water. Keep in the icebox. For use, add 0.8-1 ml. of a 5 per 80. Rorig, K.: J. Am. Chem. Soc, 71:3561, 1949. 81. Otto, P.: Ber. d. deutsch. chem. Gesellsch., 28:616, 1895. 82. Takamatsu, H., and Sho, E.: Tr. Soc. Path. Jap., 32:90, 1942. 83. Gomori, G.: Proc. Soc. Exper. Biol. & Med., 69:407, 1948. 84. Obtainable from Dajac Laboratories, 511 Lancaster Ave., Leominster^ Mass. 196 Microscopic Histochemistry cent solution of lead nitrate to 50 ml. of a 0.05 M maleate buffer of pH 5.5-5.6; shake until the initial white precipitate dissolves. Add a few drops of a 0.2 M solution of MnCl2 and 1.5-2 ml. of the phosphonate stock solution. If the mixture becomes turbid, place it in the paraffin oven and keep it there for 10-30 minutes or until the turbidity settles. Filter it into a Coplin jar. Incubate sections for 12-24 hours, at 37° C. Since the substrate is not entirely stable at pH 5.5 but de- composes slowly into free phosphate and p-chloroaniline, the Coplin jar must be supported in an inclined position, with the tissues facing downward, to avoid, as far as possible, the indiscriminate precipitation of lead phosphate all over the tissues. This maneuver will cause the heavy precipitate to settle on the back surface of the slides. Remove slides from the incubating mixture, wipe precipi- tate from back surface. Rinse them thoroughly in distilled water. Remove superficial precipitate by moving the slides around in a 0.1 M citrate buffer of pH 4.5-5. As soon as the glass appears completely clear around the tissue, rinse slides under the tap. This is the most delicate step in the entire procedure. Insufficient differentiation will leave a coarse black precipitate in the finished section, while overdifferen- tiation may remove part or all of the lead phosphate depos- ited by enzymatic action. Treat slides with ammonium sulfide, etc., as in the method for acid phosphatase. In its present form the method is not entirely dependable. The difficulties are essentially the same as with the acid phosphatase technique (lack of uniformity in staining, some- times even in the case of consecutive serial sections; unex- plained failures to get a positive reaction). Moderate amounts of the enzyme are found in practically all animal tissues; very high concentrations can be demon- strated in the vast majority of maUgnant epithehal neo- plasms. As a rule, the intensity of the reaction roughly par- allels the degree of histological malignancy. Polyps of the Enzymes 197 colon form an interesting exception; they all react intensely, even though they may be completely benign histologically. Phosphorylase Enzymes catalyzing the reaction Polysaccharide + inorganic phosphate ^ glucose- 1-phosphate (where the polysaccharide may be either starch of glycogen) have been described by Cori and Cori,^^ Kiessling,®^ and Hanes.^' If the reaction is started from the right side, a trace of polysaccharide is needed as a primer; the equilibrium of the reaction depends mainly on the pH. At pH ± 5.7, the equilibrium is shifted to the left; at pH 7-7.6, to the right. Successful attempts at the demonstration of phosphorylase have been made by Yin and Sun^^ and by Cobb.^^ Yin and Sun used sections of starch-free, water-soaked sections of soybean; Cobb used paraffin sections of frozen-dried carti- lage from which glycogen had been removed by saliva. In both cases the tissues were incubated with glucose- 1-phos- phate at pH ±6 for 30 minutes to 6 hours; starch and gly- cogen, respectively, were demonstrated by specific stains. This method deserves further investigation, especially in re- spect to the optimal conditions, of the reaction. Zymohexase (Aldolase) Allen and Bourne^^ have pubhshed a method for the dem- onstration of zymohexase. It is based on the fact that the 85. Cori, C. F., and Cori, G. T.: Proc. Soc. Exper. Biol. & Med., 34:702, 1936; Cori, C. F. and Cori, G. T.: ibid., 36:119, 1937; Cori, C. F.: Endo- crinology, 26:285, 1940; Cori, G. T., ^nd Cori, C. F.: J. Biol. Chem., 135:733, 1940. 86. Kiessling, W.: Biochem. Ztschr., 298:421, 1938. 87. Hanes, C. S.: Proc. Roy. Soc. London, B, 128:421, 1940; Hanes, C. S.: ibid., 129:174, 1940. 88. Yin, H. C, and Sun, C. N.: Science, 105:650, 1947. 89. Cobb, J. D.: The morphological distribution of glycogen and glyco- proteins in the cells and extracellular materials of growing bones (thesis. University of IlHnois, 1949). 90. Allen, R. J. L., and Bourne, G. H.: J. Exper. Biol., 20:61, 1943. 198 Microscopic Histochemistry enzyme splits hexosediphosphate into two molecules of triose phosphate. The latter hydrolyzes spontaneously at an alka- line reaction, and the phosphate liberated is visualized much as in the regular method for alkaline phosphatase. lodoace- tate and fluoride are added to the incubating mixture to pre- vent the dismutation of triose phosphate and the hydrolysis of hexosediphosphate by alkaline phosphatase, respectively. Frozen sections (whether fixed or unfixed, not clear from the text) must be used, because paraffin sections give no re- action. This method is open to criticism on several counts. Zymo- hexase itself is quite soluble and cannot be expected to re- main in situ under the conditions of the method. But the main objection is this: even if it were granted that hexosedi- phosphate will be attacked by only zymohexase under the conditions specified (and this is certainly not true, since reg- ular alkaline phosphatase hydrolyzes hexosediphosphate quite readily and is not inhibited by fluoride), the second step, namely, that of the spontaneous dephosphorylation of triose phosphate, is far too slow to permit localization, since triose phosphate is highly diffusible. There can be no doubt that its decomposition will take place almost quantitatively in the ambient fluid and not in the tissue. The illustrations given in the original paper do not corre- spond to the distribution of regular alkaline phosphatase. However, it is impossible to tell just what is demonstrated by the method. It is quite likely that some undetermined un- stable phosphatase, intolerant to embedding, is responsible for the reaction. SULFATASE Extracts of plant and animal tissues hydrolyze a variety of sulfuric esters, such as chondroitinsulfuric acid,^^ sulfates of 91. Neuberg, C, and Hoffmann, E.: Biochem. Ztschr., 234:345, 1931. Enzymes 199 indoxyl,^- phenols,^^ estrogens,^^ sugars,^^ and thiogluco- sides.^^ There seem to be four different enzymes involved, v^ith rather sharp substrate specificities. For further details the reader is referred to Fromageot's excellent review ar- ticle.^^ Sulfatases hydolyzing sulfates of phenols^^ and of thioglu- cosides^^ have been demonstrated in the tissues of higher animal species. The enzymes are quite resistant to dehydra- tion by acetone.^^ The concentration of enzyme is very low (maximum, 50-60 ^M of substrate hydrolyzed per gram of tissue and per hour;^^ in most cases activity is very much lower, usually around 1 per cent of the phosphatase activ- ity) loo Theoretically, either the sulfate or the phenolic component could be demonstrated by histochemical methods. The sul- fate method was tried with nitrophenyl sulfate as a sub- strate,^^^ without any success, although this compound is hydrolyzed at a much higher rate than other esters. Ohara and Kurata^^^ have reported positive results by the lead sul- fate technique, using phenyl sulfate or 8-hydroxyquinoline sulfate as substrates. Another method by the same authors demonstrates the sulfate ion by precipitating it with benzi- dine; the latter is then demonstrated with y8-naphthoquinone sulfonate. These methods seem to work with acetone-fixed, 92. Derrien, M.: Bull. Soc. chim. France, ser. IV, 9:110, 1911. 93. Neuberg, C, and Kurono, K.: Biochem. Ztschr., 140:295, 1923. 94. Cohen, H., and Bates, R. W.: Endocrinology, 44:317, 1949. 95. Soda, T., and Hattori, C: Proc. Imp. Acad. Tokyo, 7:269, 1931. 96. Neuberg, C, and Wagner, J.: Ztschr. f. d. ges. exper. Med., 56:334, 1927. 97. Fromageot, C: Ergebn. d. Enzymforsch., 7:50, 1938. 98. Neuberg, C, and Simon, E.: Biochem. Ztschr., 156:365, 1925. 99. Huggins, C, and Smith, D. R.: J. Biol. Chem., 170:391, 1947. 100. Rosenfeld, L.: Biochem. Ztschr., 157:434, 1925; Hommerberg, C: Ztschr. f. biol. Chem., 200:69, 1931. 101. Gomori, G.: Unpublished. 102. Ohara, M., and Kurata, Y.: Igaku to Seibutsugaku, 16:213, 1950. 200 Microscopic Histochemistry unembedded tissue only. Seligman and co-workers^^^ have used the sulfate of 6-bromo- and 6-benzoyl-^-naphthol on frozen sections of tissues fixed briefly in formalin. A long in- cubation (24 hours) is required. Human tissues do not at- tack the benzoyl compound. No detailed description of the method was given. Esterases The term "esterase" will be used in the sense of enzymes hydrolyzing esters of carboxylic acids. Esterases obtained from various sources and by various procedures may exhibit markedly different enzymatic prop- erties, especially with respect to substrate specificity and to sensitivity to activators and inhibitors. Although the classifi- cation of esterases is still a partly controversial issue, it is generally agreed that they can be divided into two large groups: namely, aliesterases,^^^ hydrolyzing esters of N-free alcohols, and cholinesterases, hydrolyzing esters, of choline. The aliesterases are subidivided into lipases, which preferen- tially split fats and oils, and esterases, the substrates of which are simpler esters of monohydric alcohols. This classification is not complete; it includes only the more widespread and thoroughly investigated enzymes which are also of histo- chemical interest. Table 5 shows the most important bio- chemical differences between esterases. It should be made clear that all the differences mentioned are relative rather than absolute and that there are many ex- amples of overlapping between enzyme types. While some lipases possess all the features enumerated under "Lipase" and some esterases all the features of "Esterase," some en- zymes occupy an intermediate position, in that in some re- spects they behave like a lipase and in others like an esterase. 103. Seligman, A. M., Nachlas, M. M., Manheimer, L. H., Friedman, O. M., and Wolf, G.: Ann. Surg., 130:333, 1949; Seligman, A. M., and Nachlas, M. M.: Cancer Research, 10:240, 1950. 104. Richter, D., and Croft, P. C: Biochem. J., 36:746, 1942. TABLE 5 Biochemical Differences between Esterases i. aliesterases (substrates: esters of n-free alcohols) 1. Substrate preferences: a) Chain length of fatty acid*' ^ h) Branching of chain of fatty acid*. . c) Ahphatic or aromatic nature of fatty acid* d) Nature of alcohol moiety e) Rates of hydrolysis of nitrophenol esters of C2-C5 fatty acids° /) Optical isomers'^' ® 2. Activators and inhibitors: a) Quinine^' e- i^- » 6) Arsanilic acid^' S' ' c) FluorideJ' ^' ' d) Bile acids^' ^' ^ A. Lipase Long (> 12) Straight chain Aliphatic Glycerol B. Esterase Short (< 12) Iso chain Aromatic Monohydric alcohols 2<3<4<5 2<3>4>5 The two types of enzymes often favor opposite optical isomers in an unpre- dictable way. Inhibition No effect Slight inhibition Activation No effect Inhibition Marked inhibition Inhibition II. CHOLINESTERASES (SUBSTRATES: ESTERS OF CHOLINE) 1. Substrate preferences'' . 2. Optimal substrate concentration" . 3. Selective inhibitorsP A. So-called True or Specific Cho- linesterase Acetylcholine, mecholyl Low (±10-3 5 M) Nitrogen mus- tard B. So-called Pseudo- or Nonspecific Cholines- terase™ Other choline esters High (± 10-2 M) Diisopropylfluoro- phosphate, per- caine^ a Terroine, E. F.: Ann. sci. nat., zool., X^ ser., 4: 1, 1942. ^ Scli0nheyder, F., and Volqvartz, K.: Enzymologia, 11: 178, 1944. •= Huggins, C, and Moulton, S. H.: J. Exper. Med., 88: 169, 1948. d Willstatter, R., and Memmen, F.: Ztschr. f. physiol. Chem., 138:216, 1924. e Ammon, R.: Fermentforsch., 11:459, 1929-30; Rona, P., and Ammon, R.: Ergebn. d. Enzym- forsch.,2:50, 1933. f Rona, P., and Pavlovic, R.: Biochem. Ztschr., 130:225, 1922. 8 Rona, P., and Pavlovic, R.: Biochem. Ztschr., 134:108, 1923. b Rona, P., and Takata, M.: Biochem. Ztschr., 134:118, 1923. i Rona, P., and Haas, H. E.: Biochem. Ztschr., 141:222, 1923, i Kastle, J. H., and Loevenhart, A. S.: Am. Chem. J., 24:491, 1900; Loevenhart, A, S., and Peirce, G,: J. Biol. Chem., 2:397, 1906-7. k Nachlas, M. M., and Seligman, A. M.: J. Biol. Chem., 181:343, 1949. 1 Seligman, A. M., Nachlas, M. M., and Mollomo, M. C: Am. J. Physiol., 159:337, 1949. ™ Click, D.: Science, 102:100, 1945. " Nachmansohn, D., and Rothenberg, M. A.: Science, 100:454, 1944. o Mendel, B., Mundell, D. B., and Rudney, H.: Biochem. J., 37:473, 1943; Augustinsson, K. B.: Nature, 162:194, 1948. p Mendel. B., and Hawkins, R. D.: Biochem. J., 41 :22, 1947; Adams, D. H., and Thompson, H. S.: Biochem. J., 42:170, 1948; Adams, D. H.: Biochim. et biophys. acta, 3:1, 1949.