The Influence of Protease Digestion and Duration of Fixation on the Immunostaining of Keratins.

Transcription

1 /86/$3 30 The Journal of Histochemistry and Cytochemistry Copyright 1986 by The Histochemical Society, inc Vol 34, No 8, pp , 1986 Printed in U S A Original Article The Influence of Protease Digestion and Duration of Fixation on the Immunostaining of Keratins A Comparison of Formalin and Ethanol Fixatio& HECTOR BATFIFORA2 3 and MARY KOPINSKI Division ofanatomic Pathology and the Sylvia Cowan Laboratory ofsurgical Pathology, City ofhope National Medical Center Duarte, CA Received for publication October 1, 1985 and in revised form January 20, 1986; accepted February 6, 1986 (5A0563) The effect of three proteases - trypsin, pepsin, and pronase - on the immunohistochemical staining of keratins with a broad-spectrum monodonal antibody was investigated in paraffin sections of formalin and ethanolfixed tissues by means of the peroxidase-antiperoxidase method Both the length ofexposure to the fixative and the duration of proteolysis were varied over a wide range Ethanol-fixed tissues showed excellent preservation of the antigenicity of keratins, and no appreciable differences in immunostaining related to the length of fixation were found The use of proteolytic enzymes did not improve these results; on the contrary, it caused rapid tissue disintegration Formalin-fixed epitheial tissues stained weakly or failed to stain unless they were treated with a pro- teolytic enzyme The optimal length of proteolysis varied with the degree offixation; tissues that were fixed for long periods of time in formalin required longer exposure to a proteolytic enzyme and were more resistant to digestion than were tissues that were fixed briefly No significant advantage of one protease over another was found in this study We condude that a proteolytic step must precede immunostaining for keratins if the tissue is fixed in formalin, but that the digestion period must be adjusted according to the length ofexposure to the fixative The superiority of alcohol over formalin fixation for the preservation of the antigenicity of keratins is confirmed by this study KEY WORDS: Kenatins; Cytokeratins; Immunohistochcmistny; Proteases; Fixation Introduction The detection of keratins in neoplastic tissues by immunohistochemical methods has numerous diagnostic and investigative applications (1-6,14,15,17) There are discrepancies in the literature, however; some investigators report detection of keratin in fewer epithelial tissues and neoplasms (16,18) than others (1,2,4) Although some ofthe discrepancies may result from the use of antibodies with different specificities for the various polypeptides that make up the keratin family (13), variations in the sensitivities of the methods employed may also play an important role (4) There is ample evidence that formalin fixation has a deleterious effect on the immunostaining of keratins and other antigens (3,4,14) Alcohol fixation, on the other hand, has been reported to yield better preservation of the antigenicity of keratins (1,2) 1 This investigation was supported by PHS grant number ROl CA awarded by the National Cancer Institute, DHHS 2 Recipient of the Governor Richard A Bryan Research Fellowship 3 Correspondence to: Hector Battifora, MD, Division of Anatomic Pathology, City of Hope National Medical Center, 1500 E Duarte Rd Duarte, CA Predigestion of tissue sections with proteases enhances the immunostaining of several antigens, including keratins; for this rcason, trypsinization or digestion with another protcase is now frequently used for detection of keratins in formalin-fixed tissues (7,8) In our consulting practice, however, we have observed several cases in which kenatin-containing, formalin-fixed neoplasms failed to stain with several monoclonal antibodies to keratins after routine 30 mm trypsin digestion, but stained readily following prolonged (2-3 hr) digestion In many of these cases we had received wet, formalin-fixed tissue, and thus the specimen had been exposed to formalin for periods ranging from a few days to several weeks For this reason, we decided to study the influence of the duration of formalin fixation on the immunostaining of keratins and the effect of various periods of digestion with trypsin, pepsin, and pronase, three of the proteolytic enzymes most commonly used in immunohistochemistry We also carried out a parallel study on the effect of prolonged fixation with absolute ethanol Methods Tissues and fixation From a patient on whom an autopsy was per- 1095

2 1096 BATFLFORA, KOPINSKI :s;,,: ; ;, 1 :w JI- #{149} _, #{149} - #{182} - C 4 : #{149} : e : - e #{149} #{149} % * #{149},u 3 lbb 1114:-, 1 SI, S 5 - o,,, - #{182} -1 -,,- : #{149}#{149} A i : : -,14 _- q- #{149} r j #{149}: t ts - 4 _, - I : #{231}r L!,::% : 4 a -: : : #{149}, -,c : #{149} Figure munostained with antibody AE1 without protease digestion Strong staining of virtually the entire thickness of the mucosa is present Bar -200 pm L2 A1 #{149},,;; I, formed shortly after death, we obtained samples of brain, kidney, esophagus, stomach, pancreas, lung, prostate, skeletal muscle, and skin Each sample was diced into 12 nearly equal-sized aliquots Eight of the 12 samples were placed in buffered fonmalin and the remainder in absolute ethanol One sample from each formalin-fixed tissue type was removed from the formalin at 24 hr and others at 1, 3, and 6 weeks; all were processed routinely for paraffin embedding without further exposure to formalin in the tissue processor Alcohol-fixed tissues were processed similarly, except that the tissues were placed directly in xylcnes and embedded in paraffin A single paraffin block was prepared from each tissue set A similar expeniment was carried out with a portion of a typical infiltrating ductal carcinoma ofthe breast except that 3, 8, 9, and 12 hr fixation times in fonmalin were added Proteases and digestion procedure For protease digestion, we used the following enzymes: trypsin (1-300, ICN, Cleveland OH), 01% in phosphate-buffered saline (PBS) ph 74; pronase (XIV P5147, Sigma Chemical, St Louis, MO), 01% in 05 M Tnis buffer, ph 74; and pepsin (P-7012, Sigma Chemical, St Louis MO), 04% in 001 N HCI To avoid losing the tissue sections during the digestion, we placed them on slides coated with a 3% solution of casein white glue (Elmer s Glue, Borden, Inc, Columbus, OH) and dried them overnight at 60 C Sections were deparaffinized with xylene and ethanol prior to immunostaining Endogenous peroxidase was blocked by incubating the sections in a fresh solution of08% hydrogen peroxide in absolute methanol for 30 mm at room temperature After a brief wash in distilled water and three short washes in PBS, ph 74, the sections were incubated in each of the enzyme solutions for periods ranging from 5 to 180 mm Control sections were incubated in PBS All of the sections were then quickly washed in two changes of distilled water and placed in PBS for 5 mm before immunostaining Antibody and immunohistochemical methods Monoclonal antibody AE1 (obtained from Hybritech Inc, San Diego, CA, now distributed by Boehninger Mannheim Biochemicals, Indianapolis, IN) was used for immunostaining of keratins This antibody recognizes an epitope shared by most keratins of the A family (4) It has been used for several years in our laboratory as a sensitive and specific broad spectrum detector of keratins However, similar results can be obtained with several commercially available monoclonal antibodies to low-molecular-weight keratins (4) H4 #{149} #{149}i#{149} t,p ::-L #{149} lh t?14;1: f; : ;;4 #{149}#{248} p S d P ; a 0 i 4j #{149}, #{231}i 41, Figure 2 Sections of a sample of breast carcinoma fixed in formalin for 24 hr and immunostained with antibody AE1 Virtually no staining is noted when trypsinization is omitted (A), whereas staining is optimal after 10 mm digestion with trypsin (B) Loss of cell detail and reduction in the intensity of the immunoreaction are evident after 30 mm of trypsinization (C) Bars = 20 pm The unlabeled-antibody avidin-biotin peroxidase complex (ABC) method was used in these experiments Briefly, tissue sections were treated with 5% normal goat serum (Cappel, Malvern, PA) for 20 mm Excess se-

3 PROTEASES IN IMMUNOHISTOCHEMISTRY OF KERATINS 1097 rum was drained off and replaced with hybnidoma supennatant of mouse monoclonal antibody AE1 at 1:50 dilution of PBS, and incubated overnight at room temperature in humid chambers Purified mouse IgG (Cappd) was used in appropriate dilutions in control sections The sections were washed three times in PBS and incubated in a 1:100 solution of biotinylated horse anti-mouse immunoglobulin G (IgG) (Vector, Burlingame, CA) for 60 mm After three more washes in PBS the sections were covered with ABC, 1:80(avidin-biotin peroxidase kit, Vector, Burlingame, CA) for 60 mm Sections were washed three times in PBS, and color was developed with diaminobcnzidine tetrahydnochloride (DAB) (Aldrich, Milwaukee, WI) as a substrate After light countenstaining with Mayer s hematoxylin, the sections were cover-slipped with Acrytol (Sungipath, Northbrook, Results IL) The results of the immunostaining were considered to be optimal when 1) an intense immunoreaction was noticed only in epithelial cells known to express the keratins detected by AE1, 2) the cell detail was well maintained, and 3) there was no noticeable background staining of stroma or nonepithelial cells Ethanol-Fixed Tissues Strong immunostaining ofepithelial cells was obtained, regardless of how long the tissues were kept in the fixative (Figure 1) There was virtually no background staining of the stroma or nonepithelial tissues Details of cell structure were very well preserved, and the fibnillary nature ofthe keratin cytoskeleton was readily visible in thin sections at high magnification We detected no differences between tissues fixed for 24 hr and those fixed for three weeks Digestion with proteases did not enhance the keratin staining; on the contrary, even short periods of digestion resulted in disintegration of architectural and cytologic detail and in reduced immunostaining Formalin Fixed Tissues Epithelial tissues, which normally stain with AE1 in frozen sections or in alcohol-fixed, paraffin-embedded sections, stained weakly or not at all when the protease digestion step was omitted In contrast, all three enzymes tested caused a remarkable enhancement of the intensity of the staining, albeit after different digestion intervals In the adenocarcinoma samples that had been fixed for 3-12 hr, a very brief incubation with any one of the three enzymes sufficed to significantly increase immunostaining For cxample, 10 mm incubation with trypsin resulted in a dramatic improvement in the staining of the neoplastic cells in the sample of breast carcinoma fixed in formalin for 24 hr, whereas without trypsin treatment only a weak, focal reaction was noted (Figure 2A and B) It was possible to oven-digest the tissues with any of the enzymes Excessive digestion manifested itself as a progressive detenioration of cytologic detail that affected the cytoplasm most severely It is significant that these changes were accompanied by a reduction rather than an enhancement of the strength of the immunoreaction (Figure 2C) On the other hand, if the tissues had been fixed in formalin for several weeks, no enhancing effect of protease digestion was seen with the short digestion periods, and the point at which optimal results were observed varied in proportion to the period of fixation - the longer the fixation time, the longer the period of digestion required to produce optimal staining (Figure 3) The increased immunoreactivity of keratins varied with the tissue type and the duration of proteolysis, as shown for trypsin in Table 1 Some tissues showed improved staining after brief incubation with the enzyme, whereas others required longer incubation times to reach optimal immunostaining However, for any given tissue, longer digestion was necessary as the duration of fixation was increased There was no noticeable advantage of one proteolytic enzyme oven the others, except that pepsin required shorten incubation to yield optimal results No artifactual staining of nonepithelial tissues was noticed with any of the three proteases tested, irrespective ofthe duration ofincubation However, with long (180 mm) incubation in trypsin, we observed increased background staining of stroma, particularly in collagen-rich tissues Discussion Immunostaining of keratin is one of the most useful immunohistochemical tests currently available for tumor diagnosis (3,4) With the use of one of several wide-spectrum anti-keratin antibodies that are commercially available, undifferentiated carcinoma and other epithelial neoplasms can now be readily distinguished from melanoma, lymphoma, and other nonepithelial tumors, which they frequently resemble (4) However, reproducibility of results among various laboratories continues to be a prob- 1cm, even when comparable monoclonal antibodies arc used Disagreements probably result from variations in the sensitivity of the methods used by the various researchers as well as from the use of the less sensitive heterologous antisera Although some antigens arc permanently destroyed by formalin fixation, many remain well preserved; they may, however, be masked, probably because of cross-linking of proteins around the antigenic site It seems, from the present study, that the process responsible for the masking of the antigenic sites is a progressive one that is worsened by prolonged immersion of the tissues in formain A number of antigens - immunoglobulins, for example - may be unmasked by protcase digestion of formaldehyde-fixed sections, but the mechanism for this rescue of antigenicity is unknown (10,1 1) If the antigen is susceptible to the digestive action of the proteolytic enzyme, the enhancing effect may not take place, or it may occur only at a critical and possibly variable time However, if the antigen is resistant to the digestive enzyme, as kenatins appear to be, the unmasking action will not only become manifest, but may be optimized by an increase in the digestion period to compensate for prolonged formalin fixation Adherence to a fixed scheme ofenzymatic digestion therefore may be responsible for false-negative results if the length of the fixation period is not taken into account Although all three proteases that we studied had a similar enhancing effect - if adjustments were made for their different rate of digestion - this may not be true for other antigens, which may have different sensitivities to the different enzymes and thus may be improved selectively by a particular protease

4 1098 BATflFORA, KOPINSKI s - :--- Table 1 Immunostaining ofseveral normal epithelial after various periods offormalin fixation and digestion with trypsin Duration of fixation (mm)1 tissues Tissue 1 day 1 week 3 weeks 6 weeks,4_ : - ; r - I - w Squamous mucosa Gastric mucosa Skin Pancreas, ducts Kidney, tubules Prostate Thyroid a The optimal digestion time for each fixation period is given 3B,-, 1 :;: : - #{149}- 1, O: I, Our results show that excessive digestion may cause a reduction rather than an increase in the number of antigenic sites available for immunohistochemical staining We found that the point at which tissues become overdigested is variable and depends not only on the duration of fixation but also on the type of tissue studied This result closely parallels the findings ofmepham (12), who studied the effect of eight fixatives on the immunostaining of IgA in tonsil tissue In the present study, tissues fixed for less than 24 hr were the most susceptible to overdigestion A recent review states that tissues fixed in formalin for a few hours are actually fixed by the alcohol used in the dehydration steps (9) Our results support this viewpoint; tissues fixed for less than 24 hr were much more susceptible to overdigestion, quite like the behavior of ethanol-fixed tissues Even at 24 hr offixation, it was clean that the central portion of each block was more susceptible to overdigestion than was the periphery This is probably the result oflongen exposure ofthc peniphcry of the tissue block to fonmalin, and hence more resistance to proteolysis As stated above, we have found negative immunostaining for keratins in several cpithelial neoplasms even when we used several monoclonal antikenatin antibodies as a cocktail Upon prolonged incubation of these neoplasms with trypsin, however, the antibodies yielded positive results (Figure 4) For some of these neoplasms, information regarding the duration offixation was not readily available In such cases, a practical approach is to incubate additional sections for periods ranging from 1 to at least 3 hr Exccssivc trypsinization can be detected easily because it produces loss ofcytoplasmic detail and decreases the immunorcaction The results ofthe immunostaining should, therefore, be read from the section with longest exposure to trypsin that still maintains good cytologic detail We compared formalin with alcohol fixation because formalin fixation is used almost universally by pathologists and because k?:h 3D 4#{228}: 4 Figure a Sections of formalin-fixed esophagus, all stained with antibody AE1 (A) 6 weeks fixation, 10 mm trypsinization, no immunostaining noted (B) 24 hr fixation, 10 mm trypsinization, partial immunostaining (C) 24 hr fixation, 1 hr trypsinization, optimal immunostaining (D) 6 weeks fixation, 1 hr trypsinization, suboptimal immunostaining Bar = 200 pm

5 PROTEASES IN IMMUNOHISTOCHEMISTRY OF KERATINS 1099 I , -, - H \ -,, : >4 4II:P4;a );0o), ; ;,_ ls:4:a ; (_:, -u4e - ;14 1 k b- #{149})ir:) I : :#{149}-kt ; - _, I!: #{149} :!S f? - #{176};: ct_ - - I, t, t -1 #{149}- 5 - )a t1,(,, -,,-,,c L i4t; )-#, k p I P yi l 1LI,/ v- S I, -I, tr,1f, - - :b-, a :- L- -,-, er r #{231}% --? Ic Figure 4 Section of a neuroendocrine carcinoma of the skin (Merkel cell tumor) which had been fixed in formalin for over 3 weeks, immunostained with antibody AE1 No immunostaining is noted after 30 mm incubation in trypsin (A) Many cells show immunoreaction product encircling the nuclei after two hr incubation (B) Endothelial cells (arrows) serve as intrinsic negative control Bar = 10 pm ethanol fixation is a simple procedure that has been shown in previous studies to yield superior preservation ofkeratins (1,2,3,4) Although we have not conducted a controlled experiment with other fixatives, we have found that B5 and Bouin s fixatives are somewhat intermediate between alcohol and formalin in the presenvation of keratin antigenicity, and that the results can be improved with light protcolytic digestion In Mepham s study, howeven, only formalin-fixed tonsil tissue showed enhanced immunoreactivity for IgA following trypsinization (12) Because most users ofbouin s and B5 fixatives pay careful attention to fixation time to avoid overfixation, these fixatives may be better for immunohistochemical studies than formalin because overfixation with them is infrequent, and not because there is an intrinsic chemical advantage Our results confirm previous work done by us and by others illustrating the superiority of absolute ethanol over fonmalin as a I - fixative for the demonstration of kcratins (1,2,3,4) Unpublished data from our laboratory have shown that many other antigens are also well preserved by ethanol Ethanol fixation affords excellent preservation ofccllulan detail, but at the expense ofslightly higher cell shrinkage than occurs with formalin An additional advantage of ethanol fixation, as confirmed by this work, is that the length oftime offixation, at least for periods ofup to 6 weeks, is not ddetenious to keratin antigenicity Furthermore, ethanol-fixed tissues, regardless of how long they have been fixed, yield optimal results for keratin immunostaining without the needforprotease digestion We recommend routine fixation in ethanol of a nepresentative portion of tissue that is to be used for immunohistochemical study Acknowledgments The authors gratefully acknowledge the excellent technical assistance of Parula Mehta and Zamir Warsi Literature Cited 1 Altmannsbergcr M, Osborn M, Holscher A, Schauer A, Weber K: The distribution of keratin type intermediate filaments in human breast cancer Vinchows Arch [Cell Pathol] 37:277, Altmannsberger M, Weber K, Hoscher A, Schauer A, Osborn M: Antibodies to intermediate filaments as diagnostic tools Human gastrointestinal carcinomas express prekaratin Lab Invest 46:520, Battifona H: Recent progress in the immunohistochemistry of solid tumors Semin Diagn Pathol 1:251, Battifora H: Diagnostic uses of antibodies to keratins A review and immunohistochemical comparison of seven monoclonal and three polyclonal antibodies Prog Surg Pathol (in press) 5 Battifora H, Sheibani K, Tubbs R, Kopinski MI, Sun it Antikeratin antibodies in tumor diagnosis Distinction between seminoma and embryonal carcinoma Cancer 54:843, Battifora H, Sun it, Rao 5, Bahu R: Anti-keratin antibodies in tumor diagnosis: thymomas Hum Pathol 11:635, Cunran RC, GregoryJ: The unmasking ofantigens in paraffin sections of tissue by tnypsin Expenicntia 33:1400, Denk H, Radaszkiewicz T, Weirich E: Pronase pretreatment of tissue sections enhances sensitivity of the unlabelled antibody-enzyme (PAP) technique J Immunol Meth 15:163, Fox CH, Johnson FB, WhitingJ, Roller PP: Formaldehyde fixation J Histochem Cytochem 33:845, Hautzer NW, Wittkuhn JF, McCaughcy WTE: Trypsin digestion in immunoperoxidase staining J Histochem Cytochem 28:52, Huang S-N, Minassian H, More JD: Application of immunofluorescent staining on paraffin sections improved by trypsin digestion Lab Invest 35:383, Mepham BL: Influence of fixatives on the immunoreactivity of panaffin sections Histochem J 14:731, Moll R, Franke IT, Schiller DL: The catalog of human cytokenatins: patterns of expression in normal epithelia, tumors and cultured cells Cell 31:11, Nagle RB, McDaniel KM, Clark VA, Payne CM: The use of antikenatin antibodies in the diagnosis of human neoplasms Am J Clin Pathol 79:458, Osborn M, Weber K: Biology of disease: tumor diagnosis by inter-

6 1100 BAUIFORA, KOPINSKI mediate filament typing: a novel tool for surgical pathology Lab Invest 48:372, Schlegcl R, Banks-Schlegel 5, McLeodJA, Pinkus 5: Immunoperoxidase localization of keratin in human neoplasms Am J Pathol 101:41, ThxyJB, Hidvegi DF, Battifora H: Nasopharyngeal carcinoma: antikeratm immunohistochemistry and electron microscopy AM J Clin Pathol 83:320, Wachnen R, Wittekind C, von Kleist S: Virchows Arch (A) 402:415, 1984