View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

HISTOCHEMICAL DEMONSTRATION OF FREE AMINO GROUPS IN CUTANEOUS * BRIAN POTTER, M.D.

An investigation into the distribution of aminobeen employed for this purpose histochemically groups in physiologic and pathologic skin was(3), and chloramine-T (4). These reagents give suggested by recent biochemical and histochemi-closely similar results under histochemical condi- cal developments concerning the interestingtions (2). The following discussion will be limited reagent ninhydrin. This chemical was firstto the action of ninhydrin, the reactions of which compound have been most exhaustively investi- synthesized in 1910 by Ruhemann who predicted gated. that it would prove of value in physiological In the reactions by which ninhydrin produces chemistry. Used for years in paper chromatog-aldchydcs, a violet color is formed independently. raphy for the demonstration of amino acids,Ninhydrin gives this violet colored product with and oligopeptides, the use of ninhydrinthe following physiologic compounds: ; has lately been extended to the demonstration ofalpha, beta, gamma, delta and epsilion amino those and phospholipids that containacids; oligopeptides (5), polypeptides and proteins amino groups. The color produced directly by(6); primary and secondary aliphatie amines (7); ninhydrin from amino compounds is too faint to2-amino sugars and 2-aminohexonic acids (5); and be of value in histochcmistry. However it hasphospholipids containing ethanolamine or serine (9). Ninhydrin does react with various non- recently been recalled that, besides a direct violetphysiologic amines; organic bases and other color, an aldchyde is produced by ninhydrin incompounds (10); but the color produced, if any, is the same reaction. Quite large amounts of alde-not violet unless free amino groups are available. hyde are formed in tissue sections in this way This is the basis of spot tests with ninhydrin, and are readily demonstrable by Schiff's reagent.popular in paper chromatography. The same color This provides an elegant method for the histo-reaction has been used in histochemistry as a test chemical demonstration of free amino groups. for proteins but the color produced, besides being faint, is due to a small molecule and consequently cHEMICAL BASIS OF THE REACTiON liable to diffuse. The violet color is formed by combination Organic compounds containing free amino groups can be degraded to by the actionbetween ninhydrin, reduced ninhydrin and the liberated ammonia (11). Besides the , the of a variety of inorganic and organic chemicalscolor and ammonia, carbon dioxide is liberated (1). The prototype reaction is that given by alpha amino acids and is known as the Strecker degrada-when an alpha carboxyl group is present, as in amino acids (12). tion. Inorganic oxidizing agents include alkaline sodium hypochloritc and calcium hypochlorite, The pentoscs produced from 2-amino sugars but these do not produce stable tissue aldehydesexist in furanose ring form and do not give a under histochemical conditions (2). Organiccolor with Schiff's reagent. Therefore carbohy- agents active in this oxidative degradation com-drates cannot give a positive reaction with the prise chiefly carbonyl compounds including theninhydrin-Schiff method. ninhydrin, alloxan and isatin, which have During fixation, washing and staining histo- logic sections are exposed to water. Amines, *Presentedat the Twentieth Annual Meetingamino acids, small molecule polypeptides and of The Society for Investigative Dermatology,amino sugars are soluble and consequently will Inc., Atlantic City, N. J., June 6, 1959. diffuse and dissolve. From the Section of Dermatology, Department of Medicine, University of Chicago, Chicago, In the course of infiltration with paraffin and Illinois. (Chief of Service: Stephen Rothman,again in deparaffinization, tissues are exposed to M.D.) cold organic solvents. This will remove most of This study was supported in part by a grant from the U. S. Public Health Service, National In-the phospholipids. stitute of Allergy and Infectious Diseases, Grant Consequently in sections from paraffin blocks No. E-1444(C3) and in part by a grant from the A. B. Kuppenheimer Research Endowment Fundonly proteins will be available to react with this of the University of Chicago. technic, except for any traces of phospholipid 245 1) Proteins and Related Compounds C RCHCOOH —RCHO + NH3 + CO2 + violet color (1)

NH2 aldehyde (one less car- bon atom) RCH2NH2 —RCHO + NH3 + violet color (1) aldehyde (same number of carbon atoms)

HOOCCHNHCOCH2NH2 —4aldehyde +N113 + violet color (5, 13)

I (postu- R lated)

(14) HOOC'CHNHCOCHNHCOCHvNH2 aldehydes + CO2 + violet color (5)

I I (postu- R (Cl2)4 lated)

NH2 2) Phospholipids Containing Amino Groups CH2—O—CO—R'

CH—O—CH=CH—R" 0

CH2—O—P—O—CH2—CH2—NH2 —aldehyde + violet color (9) (postu- OH lated) Plasmalogen (or cephalin) containing serine or ethanolamine 3) 2-Aminosugars COOH

HC—NH2 CHO lOCH lOCH I +violet color HCOH HCOH

HCOH HCOH

CH2OH CH2OH Hexaminic acid Pentose (8) /O\Cl2 OH /0\ OH /\ OH HO/\ /OH HO /\ NH2 OH Hexosamine Pentose (8) + violet color 246 FREE AMINOGROUPS IN PROTEINS 247 firmly bound to proteins as lipoproteins. In NH NH frozen sections kept from solvents, cephalins and plasmalogens containing ethanolamine or CH—CH2—S—Hg—S—CH2—CH serine would react. In a word, then, substances in tissue sections C=O C==O giving a positive reaction with the ninhydrin-between protein molecules (17). Schiff technic are either proteins or lipoproteins. The acidified potassium dichromate of Zenker These findings of organic chemistry have beenfluid does not apparently react with amino partially confirmed by histochemical demonstra-groups. tion. Control sections digested by pepsin do not give the ninhydrin-Schiff reaction (3) because METHOD AND MATERIALS the amino-containing products are diffusible; Biopsy and autopsy specimens of normal and neither do control sections previously deaminateddiseased skin and scalp were fixed in 10% formalin- by nitrous acid (2). The acaciasaline (i.e. 4% formaldehyde), absolute acetone and agar give negative results (2). Model slidesor Zenker's fluid (with acetic acid). The Zenker- coated with albumin take the stain but pepsinfixed material was washed in running tap water and insulin do not produce as intense a shade asovernight. The procedure was then as follows: 1. Dehydrate, embed, section and mount. expected (2). The explanation for this may lie in the fact that each submolecule of insulin of 2. Deparaffinise sections, clear, hydrate. 3. Immerse in 5% w/v aqueous phenylhydrazine molecular weight 12,000 has only six free aminohydrochloride for 1hour. groups (15): two s-amino groups of lysine and 4. Wash under distilled water faucet and in four open peptide chains. The amino acid com-95% . ponents of proteins generally are bound together 5. Incubate in 0.5% w/v alcoholic recrystallised by peptide bonds through their amino and car-ninhydrin at 55° C. for 20 hrs. boxyl groups and the rest of their molecule is 6. Wash in two changes of 95% alcohol and tap free (16). The only sources of amino groupswater for 15 minutes. appear to be the terminal ones and the lysine and 7.Treatwith Schiff's reagent for 15 minutes. hydroxylysine side chains of some proteins. 8. Transfer directly to sulfite solution; 2 changes of 5 minutes each. FIXATION 9. Wash in tap water, 15 minutes. 10. Dehydrate, clear and cover. Since it is necessary for amino groups to be The third step, blockade of pre-existing tissue available after fixation to react with ninhydrin,aldehydes, is designed to prevent a direct Schiff the effects on amino groups of the fixatives usedreaction. Although these false positive colorations must be considered. are weak even with Zenker fixation (the plasma! Acetone, formalin and Zenker were chosen asreaction), without such blockade the production examples of non-additive denaturing; additiveof aldehydes could not be attributed solely to the action of ninhydrin. non-coagulating; and additive coagulating fixa- To achieve the maximum oxidative production tives respectively. of aldehydes by ninhydrin, the latter must be Acetone affects amino and other hydrophilicfreshly recrystallised. This was done by dissolving groups of proteins causing loss of their capacityninhydrin in 2 times its weight of hot water and to hold water and by breaking links and unfoldingboiling with decolorising carbon to absorb im- protein chains makes more amino and otherpurities. Pure ninhydrin is crystallised on stand- groups available (17). ing the filtrate in the refrigerator, is recovered in Formalin (formaldehyde) reacts with aminoa Buchner funnel, washed with cold water and groups to form methylene bridges dried in an oven (18). The concentration, solvent, temperature and —NH—CH2----NH--- time mentioned in Step 5 gave optimum results; lower temperature, higher concentration, less between adjacent protein chains (17). prolonged exposure and solution of ninhydrin in Mercuric chloride at the low pH. of Zenkerwater instead of alcohol all produced smaller fluid may be taken up by ionized amino groupsamounts of aldehyde. as the ion (HgC14), but its chief reaction is with Higher temperatures resulted in the evaporation sulfhydryl groups to form mercaptide links of the solvent. Placing the sections in a fresh 248 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

ninhydrin solution for a second period of 20 hoursnormal skin. In one pathologic specimen, amor- did not result in the development of any strongerphous material in the lumen of a clear cell tumor color. of sweat gland was stained pale violet. In serial In Step 6 washing the sections in running tapsections treated with the full ninhydrin-Sehiff water for as long as three hours did not cause any diminution of the color obtained with Schiff'stechnic this material became purple as expected, but in hair shafts the cortex did not do so. reagent. The Schiff's reagent and sulfite solution were each made by method one of Kligman and Mescon RESULTS (19). The purpose of the sulfite rinse was to re- In normal skin the nuclei and cytoplasm of the move excess colorless Schiff's reagent that mightepidermis, hair follicles, sweat glands and duets become colored nonspecifically by oxidisation onand erector pili muscles were quite brightly exposure to air (20). stained purple (Fig. 1). So were the collagen CONTROLS bundles especially the thicker ones, so that the papillary layer was fainter than the rest of the Sections treated in a fashion identically as the others except for Step 5 were incubated in abso-eorium. Sebaceous gland cells, having scanty cytoplasm, did not develop much color. Erythro- lute alcohol alone instead of ninhydrin solution. After the remaining steps these were examinedeytes in the lumen of vessels were nm- and found to be colorless. The production of colorhydrin-Sehiff negative and appeared as empty by the method outlined is thus shown to be theyellowish envelopes. Leukoeyte nuclei were col- action of ninhydrin. ored purple. Sections from which Sehiff's reagent was with In the normal epidermis the eorneal layer, held were examined after incubation in ninhydrinsquamous cell nuclei and cytoplasm all were solution. These were colorless except for a faintstained purple with a uniform intensity. In the violet staining of hair shafts in some sections ofhair follicle the hair shaft cortex was not colored; I ia-!c r

N

Fm. 1. Normal human skin, Zenker fixed, ninhydrin-Sehiff stain, 40X. The portion of hair shaft appears dark because of its natural pigmentation not depth of staining. FREE AMINO GROUPS IN PROTEINS 249 -: _ - •1 .;c) ,-t C ; I ....aL' ¼.. .4 -.c, fl4;'t I. 1 L * ci FIG.2. Hair follicle, Zenker, ninhydrin-Schiff, 80X. The hair shaft medulla takes the stains, the cortex does not. the medulla was purple; the inner root sheathpositive reaction in the keratogenous zone has purple. The outer root sheath developed a uni-been described (21). form color in cytoplasm and nuclei of about the The smooth muscle of the erector pili muscles same intensity as the surrounding connectivewas colored less strongly than the striated bundles tissue (Fig. 2). In some sections the upper part ofof voluntary muscle in the skin, for instance the the lower third of the hair and inner root sheathoccipitalis or frontalis of the scalp, which de - were particularly brightly colored (Fig. 3). Thisvelopeda bright purple color (Fig. 4). 250 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY 4 - .1 -— :* - - - :-.--. Il- .-. •r - _a- .±&t '- ' -'-iti — ' — -• - Tht_ .- - - -_ -— p — - '--C - 4 '- -- — -' S.- r 'f-- a • —-- .— 1. - — - — -' em —

FIG. 3. Keratogenous zone of hair, Zenker, ninhydrin-Schiff, 175X.Thelower circular dark area represents pigment, the rest staining. The above description is that of the results had a distribution similar to that described for after Zenker fixation. In normal skin after acetone Zenker fixed skin. fixation the structnres colored most strongly by the ninhydrin-Schiff method were collagen PATHOLOGIC CHANGES bundles and the keratin of the corneal layer. Morbid lesions in which hyperkeratosis or Fixation by formalin resulted in generally keratin cysts or tunnels were a feature were not weaker colorations but positive reacting materialsstained differentially by the ninhydrin-Sehiff .9.

...—,•;•.—. -- - -'*'

FIG.tc 4. Occipitalis (striated) muscle, Zenker, ninhydrin-Schiff, 106X. The lightly staining oval struc- tures represent nerve.

1

FIG. 5. Ichthyosis hystrix, formalin, ninhydrin-Schiff, 11OX. The keratin appears densely stained where it is more compact. 251 252 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

.5 tt. P% Fin.6.Lipoid proteinosis, formalin, ninhydrin-Schiff, 67X. The homogenous material reacts posi- tively but less strongly than collagen. reaction. It was noted in ichthyosis hystrix thatacellular zone beneath the epidermis was stained the adherent keratin was brightly colored purplewith approximately the same intensity as other where it was most compact. This correspondedstructures in the corium. with areas that were somewhat basophilic rather In sections of senile elastosis the amorphous than acidophilic when stained with hematoxylinmaterial in question also was not stained differ- and eosin (Fig. 5). entially but with the same intensity as adjacent Parakeratosis could not be distinguished fromcollagen. orthokeratosis by the ninhydrin-Schiff method. In sections from a rheumatoid nodule the foci Formalin fixed sections of a number of patho-of necrobiosis contained hyalinized areas colored logic specimens containing homogeneous materialstrongly purple by this technic. were selected and treated by the ninhydrin-Schiff technic. These comprised examples of colloid HISTOLOGIC USE milium, lipoid proteinosis and balanoposthitis For histopathologic purposes, as contrasted xerotica obliterans. with histochemical, sections were lightly counter- In colloid milium the papillomatous lobules ofstained with hematoxylin for identification of homogeneous substance were stained equallynuclei and orientation. strongly as collagen in the corium of the same With ninhydrin-Schiff the cytoplasm of granu- lesion; that is to say, they developed a lightlar cell myoblastoma was stained brilliantly purple color. Homogeneous material in the widepurple to a degree corresponding to that of superficial band of the coriuxn and around thevoluntary muscle (Figs. 7 and 8). sweat glands in lipoid proteinosis achieved a The cytoplasm of leiomyomas and of those lighter color than the more sclerotic of collagengranulomas that were examined, on the other bundles (Fig. 6). hand, took only a pale stain. A reticulohistio- In balanoposthitis xerotica obliterans thecytic granuloma was studied but the giant cells FREE AMINO GROUPS IN PROTEINS 253 — t? -

FIG. 7. Granular cell myoblastoma, formalin, ninhydrin-Schiff and hematoxylin, 80X. Nuclei are counterstained for differentiation. proved difficult to identify without hematoxylin;free amino groups and (b) that the substance when counterstained the cytoplasm of these stained represents protein or lipoprotein. cells had a basophilic tint. This technic is therefore a more specific histo- chemical test for proteins than dinitrofluoro- benzene or azomethine methods, which would A purple color with Schiff's reagent after thereact with aminopolysaccharides as well. action of ninhydrin signifies (a) the presence of The ninhydrin-Schiff reaction is affected by 254 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

FIG. 8. Portion of same section as Fig. 7, 625X. The granular cytoplasm is strongly ninhydrin-Schiff positive.

the mode of action of fixatives. Fibrous proteinsgroups physically available, formalin combining such as soft keratin and collagen reacted moredirectly with them and Zenker reacting with free strongly after acetone, and cellular constituentsamino groups only secondarily. more strongly after Zenker, than either reacted The only structures in normal skin that were after formalin. This is explicable in the light ofnot stained at all by ninhydrin-Schiff were knowledge of the mechanism of fixation by theseerythrocytes and the hard keratin of hair cortex. chemicals, acetone making more free aminoTherefore free amino groups cannot be available FREE AMINO GROUPS IN PROTEINS 255 to alcoholic ninhydrin in these structures, eitherby voluntary muscle and the keratogenous zone because of their absence or because of nonpene-of hair follicles. tration of the reagent. It has been reported that 5. Erythroeytes and hard keratin of hair cor- erythrocytes (after Zenker fixation) and hairtex do not react and therefore do not contain cortex also failed to stain in an entirely differentavailable amino groups. (azomethine) histochemical method for the 6. The amorphous substances present in col- demonstration of amino groups (22). bid milium, lipoid proteinosis and senile elastosis A direct violet color was formed in hair shaftsreact and consequently contain protein. The by ninhydrin alone and yet aldehydes were nothomogenized zone of balanoposthitis xerotica produced in the cortex. It is not certain whetherobliterans and hyalinized areas of necrobiosis in this was due to the observed staining of therheumatoid nodule contain protein. medulla or to absorption of the diffusible dye by 7. The cytoplasm of granular cell myoblastoma the hair. reacts equally as strongly as that of striated Materials such as voluntary muscle and themuscle and more strongly than leiomyomas and keratogenous zone of the hair are known to con-granulomas. This may be taken as evidence of sist of protein but react more strongly positivemyoblastie origin of this tumor, and is an example than other structures in the same sections alsoof histopathologie, in addition to histoehemieal, known to contain protein. These are interpreteduse of this technic. as having a greater concentration of amino groups; myoglobin is in fact known from bio- Acknowledgement is made to The Director, chemical studies to be rich in amino groups (22).Armed Forces Institute of Pathology for the loan Provided that pre-existing aldehydes haveof sections of colboid milium and reticulohistio- been blockaded, any color developed with Schiff'scytie granuloma and to Mrs. A. Spencer Lucas reagent after ninhydrin must be accepted as afor technical assistance. positive reaction. Positive staining material infiltrated into or stored in pathologic lesions REFERENCES such as eolloid milium and lipoid proteinosis is 1. SCHONBEEG A. AND MOTJBACHER R.: The streeker degradation of a-amino acids. Chem. therefore readily identified as containing protein. Reviews, 50: 261—277, 1952. Similarly the homogenized zone of balano- 2. BUR5TONE, M. S.: An Evaluation of Histo- posthitis xerotica obliterans is interpreted directly chemical Methods for Protein Groups. J. Histoehem. and Cytoehem., 3: 32—49, as being protein. The amorphous and fibrous January, 1955. material that appears in senile elastosis is also 3. YA5UMA, A. AND IcHIKAwA, T.: Ninhydrin- Schiff and Alloxan-Sehiff staining. J. Lab. identified as containing protein by comparing and Clin. Med., 41: 296—299, 1953. ninhydrin-Schiff treated tissue with serial sec- 4. Csiu, C. H. U., F0GEL50N, M. H. AND SwIN- tions stained by routine methods. YARD, C. A.: A histoehemical test for alpha amino acids. J.Histoehem.& Cytochem., The comparison of the strongly positive cyto- 1:391—2, 1953. plasm of granular cell myoblastoma with that of 5. DENT,C.E.: A study ofthe behaviour of some myoglobin and contrast with the weaker reac- sixtyamino acidsand other ninhydnn reactingsubstances onphenol-'eollidine' tions given by leiomyoma and granuloma cyto- filterpaper chromatograms, with notes as plasm is in favor of the myoblastie hypothesis of to the occurrence of some of them in bio- the origin of this interesting tumor. logicalfluids. Biochem. J., 43: 169—180, 1948. 6. RUTHEMAN, S.: Triketohydrinedene Hydrate. SUMMARY Part III. Its Relation to alloxan. J. Chem. Soc. Trans., Vol. 99: 793—800, 1911. 1. Ninhydrin deaminates compounds contain- 7. FEIGL, F.: Spot Tests in Organic Analysis. ing free amino groups and produces aldehyde 6th English Edition Amsterdam. Elsevier groups in place of the latter. Publishing Co., 1956. 8. KENT, P. W. AND WHITEHOU5E, M. W.: Bio- 2. Aldehydes give a purple color with Sehiff's chemistry of the Amino Sugars. New York, Academic Press Inc., 1955. reagent. 9. LEA, C. H. AND RHODES, D. N.: Ninhydrin 3. In washed and lipid-extracted tissue sec- reaction of unhydrolysed phospholipids. tions amino groups demonstrated in this way are Biochim. et Biophys. Acta, 17: 416—23, 1955. 10. RUHEMANN, S.: Cyclic Di- and Tri-ketones. present in proteins and lipoproteins only. J. Chem. Soc. Trans. 97: Pt. II, p. 1438—49, 4. In normal skin the brightest color is given 1910. 256 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

— TriketohydrindeneHydrate, Ibid p. 17. BARER, J. H.: Principles of Biological Micro- 2025—31. technique. New York, John Wiley & Sons 11. MACFADYRN, D. A.: On the mechanism of the Inc., 1958. reaction of ninhydrin with alpha amino18. Mooma, S. AND STEIN, W. H.: Photometric acids. J. Biol. Chem., 186: 1—12, 1950. ninhydrin method for use in the chro- 12. VANSLYKE,D.D.,DILLON,R.T., MACFADYRN, motography of amino acids. J. Bml. Chem., D. A. AND HAMILTON, P.: Gasometric de- 176: 367—388,1948. termination of carboxyl groups in free 19.KLIGMAN, M.ANDMEscoN, H.: The periodic- amino acids. J. Bid. Chem., 141: 627—669, acid-Schiff stain for the demonstration of 1941. fungi in animal tissue. 4. Bacteriology, 60: 13. CRI5T0L, P., CaA5TE5 DR PATJLRT, A., ET No. 4, October, 1950. CRI5TOL, P.: Action de l'acide nitreux, de la 20. HALE, A. 4.: Histochemistry of polysac- chioramine T et de la Ninhydrine sur les charides. Int. Rev. Cytology VI: 193—263, glycyl, alanyl, et leucyl . Bull. Soc. 1957. Chim. biol., 38: 639—47, 1956. 21. BEAUN-FALCO, 0.: The Histochemistry of the 14. MACFADYEN, D. A.: Determination of amino Hair Follicle. In the Biology of Hair Growth, acids in plasma by the ninhydrin-carbon dioxide reaction without removal of proteins. p. 70. Montagna, W. and Ellis, H. A., J. Biol. Chem., 145: 387—403, 1942. Editors, New York, Academic Press Inc., 15. SANGEa, F.: The free amino groups of insulin. 1958. Biochem. J., 39: 507—15, 1945. 22.WEIss,L. P., Tsou, K. C. AND SELIGMAN, 16. OLcOTT, H. S. AND FRAENKEL-CONRAT, H.: A. M.: Histochemical demonstration of Specific group reagents for proteins. Chem. protein-bound amino groups. 4. Histochem. Reviews, 41: 151—197, 1947. & Cytochem., 2: 29—49, 1954.

DISCUSSION Dx. STEPREN ROTHMAN (Chicago, Ill.): I Soft keratin behaves differently from hard would like to ask Dr. Potter if he can explain the keratinin other procedures as well as those for seemingly paradoxical situation that in the epi-free amino groups. The hair medulla resembles dermis compact horny layers stain brighter thanthe soft keratin of the stratum corneum of the the loose ones while in the hair the medulla withepidermis, while the cortex and cuticle represent a loose keratin structure stains more than thehard keratin. Amino groups of hard keratin are cortex which is built of hard compact keratin. evidentlynot present in the free state. Dat. Lou1s H. WINER (Beverly Hills, Calif.): Insections of lipoid proteinosis and of eolloid Dr. Potter's paper is stimulating in that it sug-milium, these amorphous substances stain equally gests a means of studying lipoid proteinosiswith ninhydrin-Schiff; consequently they cannot (hyalinosis cutis et mucosa) and differentiating itbe distinguished by this method. Histologic differ- from colloid milium. ential diagnosis between them can be made by the In addition this study shows quite clearly thatpresence of hyaline material around sweat glands granular cell myoblastoma is of muscle originin lipoid proteinosis. This does not occur in rather than of neurogenic origin even thougheolloid milium. morphologically and structurally it is frequently I have not worked on other tissues, but granti- in juxtaposition to a nerve. larcell myoblastoma cytoplasm staths much Dn. BRIAN POTTER (in closing): I would likemorebrightly with ninhydrin-Sehiff than does to thank Dr. Rothman and Dr. Winer for theirneurofibroma or normal nerve sheath present in remarks. the skin.