Virchows Archiv B Cell Pathol (1991) 61:81-87

ViwhowsArchivB Cellttn/Mogy l~t~i~ M o t ~ e d ~ v

9 Springer-Verlag1991

Keratin and carcinoembryonic antigen (CEA) in human melanoma cells* Anil Om, Tarun Ghose, and Geoffrey Rowden Department of Pathology, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4H7 Received September 17, 1990 / Accepted May 12, 1991

Summary. H u m a n melanomas are known to contain vimentin intermediate filaments but there has been some dispute about their expression of cytokeratins. The cytoplasm o f human M21 melanoma cells maintained in culture reacted with a rabbit anti-keratin antibody and two monoclonal anti-keratin antibodies AE1 and AE2. Cells derived directly from subcutaneous xenografts of M21 melanoma in nude mice, however, failed to express cytokeratins. The presence of keratin filaments in cultured M21 cells was confirmed by electronmicroscopic and immuno-electronmicroscopic examinations of cell extracts. Polyacrylamide gel electrophoresis (PAGE), revealed 46 K D keratin proteins in cultured M21 cells. Small amounts of these low molecular weight keratins were detected by P A G E in M21 melanoma xenografts even though immunofluorescence and immunoperoxidase assays failed to demonstrate keratin at the light microscopic level. Immunofluorescence revealed keratin and carcinoembryonic antigen (hitherto undetected in human melanomas) first on the 9th day of culture of xenograftderived M21 cells. The appearance of keratin and CEA in M21 melanoma cells in vitro was not affected by inhibition o f cellular proliferation or as a result of exposure to methotrexate or adriamycin. However, adriamycin altered the cytoplasmic distribution of keratin. Key words: H u m a n M21 melanoma xenograft - Cultured melanoma cells - Carcinoembryonic antigen Keratin - Methotrexate - Adriamycin

Introduction Tissue and tumor markers are being increasingly used for elucidating the histogenesis of tumors, especially when the lesions are metastatic and undifferentiated. To be diagnostically useful, the expression o f these markers should not be affected by cancer-chemotherapeutic Offprint requests to: G. Rowden, Departmeat of Pathology, llth FI., Sir Charles Tupper Medical Bldg, Dalhousie University, Halifax, Nova Scotia, Canada B3H 4H7

agents or the status o f proliferation of tumor cells. Cytokeratins are intermediate filaments associated with epithelial cells and carcinoembryonic antigen (CEA) is frequently used to establish the epithelial origin of tumors (Franke et al. 1978a). Epithelial tumor cells synthesize CEA (Kanazawa et al. 1982) and with some minor exceptions (Denk et al. 1982; Franke et al. 1981) continue to express the same intermediate filament protein both in vitro and in vivo (Sun and Green 1978a; Sun et al. 1979; Sun et al. 1984). H u m a n melanoma cells in excised tumor specimens contain the vimentin class of intermediate filament (Caselitz et al. 1983) and are not generally accepted as expressing CEA (Goldenberg et al. 1978 and see Urmacher 1990). We report here the appearance of low molecular weight keratins and CEA in the M21 line of human melanoma cells when these cells were maintained in vitro for long periods. However, both CEA and keratin became undetectable by the staining assays when the cultured M21 cells were grown as subcutaneous xenografts in nude mice. Exposure of M21 cells to methotrexate (MTX) or adriamycin did not interfere with the expression of CEA and keratin.

Materials and methods Tumor cell lines. The human M21 melanoma cell line was from Dr. Soldano Ferrone, New York Medical College, Valhalla, New York. The human cervical carcinoma cell line, Hela and the human kidney carcinoma line Caki-2 were from the American Type Culture Collection, Rockville, MD. The human colon adenocarcinoma cell line, SK-CO-1 was from Dr. J. Fogh, Sloan Kettering Institute for Cancer Research, New York and the human lymphoblastoid cell line was obtained in our laboratory by the Epstein-Barr virusinduced immortalization of peripheral blood lymphocytes from a chronic lymphocytic leukemia patient. Tumors. Formalin fixed paraffin-embedded blocks of the various tumors examined were from the files of the Victoria General Hospital, Halifax, Nova Scotia. Antibodies. The anti-CEA antibody (DAKO, Cedarlane Laborato-

ries Ltd., Hornby, Ontario) was obtained from rabbits immunized with human colonic adenocarcinoma. In addition to CEA it also

82 reacted with the non-specific cross-reacting antigens (Von Kleist et al. 1972; Tsutsumi et al. 1984). The antibody was absorbed with insolubilized human plasma and blood group antigens A2 and B and its reactivity was established by cross immunoelectrophoresis against CEA. Details of production, serologic analysis and specificity of rabbit antimelanoma antibody (AMA) have been described elsewhere (Ghose et al. 1975). After appropriate absorptions to render them specific, the sera reacted only with human melanoma cells and not with human normal skin fibroblasts, peripheral blood lymphocytes or other normal adult human tissues. The monoclonal antibody 225.28 to a high molecular weight melanoma associated antigen (MAA) was a gift from Dr. Soldano Ferrone. The target antigen of this MAB has not been detected in normal tissues and its expression is not associated with the synthesis of melanin. The rabbit antikeratin antibody was from Dr. K.M. Shiekh. The specificity of this antibody was established by immunoprecipitation and by its reactivity with the cytoplasm of cells in all the layers of the normal epidermis on IP staining and is comparable to commercial polyclonal antibodies against human callus produced by DAKO. The monoclonal antikeratin antibodies were from Dr. T.T. Sun, Department of Dermatology and Pharmacology, School of Medicine, New York University, New York. Monoclonal anti-keratin antibody AEI, recognizes 40, 48, 50 and 56.5 KD keratins. AE2 recognizes 56 and 56--57 KD keratins and AE 3 recognizes 52, 56, 58 and 65-67 KD keratins. The specificity of the monoclonal anti-vimentin antibody (DAKO) was confirmed by its reactivity, on IFL assay, with uterine and vascular smooth muscle, vascular endothelium and the lymphoblastoid cell line and its lack of reactivity with epithelial cells.

Extraction and characterization of keratin. Aliquots of 8 x 107 cells obtained from i) M21 cells in culture and demonstrated by IFL and IP to contain keratin, ii) xenografts of M21 melanoma in which intracellular keratin could not be detected and iii) Hela cells in culture and demonstrated by IFL and IP to contain keratin were subjected to the 3 step extraction procedure of Sun (Sun and Green 1978). Supernatants from the third extraction were dialyzed against distilled water for 24 h at 4~ C, the precipitates separated by centrifugation and were then examined at 60 Kev in a Philips 300 electron microscope following negative staining with phosphotungstic acid. For immunoelectron microscopic (IEM) examination, 1 i11 drops of precipitates were placed on coated grids as for the negative staining and then exposed to 5% swine serum followed by a 1:100 dilution of a rabbit antihuman keratin antibody. The grids were subsequently washed with 0.05 M tris buffer containing 0.2% bovine serum albumin and finally exposed for 10 min to a 1:50 dilution of goat anti-rabbit IgG labeled with collodial gold, (20 nm particles, Janssen Life Sciences, Toronto, Ontario) and then washed as for the primary antiserum. In control preparations, either the anti-human keratin antibody was omitted or substituted by preimmune rabbit serum. Keratin precipitates were dissolved in 1% sodium dodecyl sulfate in 10 mM dithiothreitol and then subjected to PAGE and immunodiffusion against the rabbit antihuman keratin antibody.

Results

Drugs. Methotrexate was from Sigma, St. Louis, Missouri and adriamycin was from Adria Laboratories, Mississauga, Ontario.

Reactivity of the antibodies with deparaffinized sections of human tumors

lmmunofluorescence (IFL). Cytoplasmic IFL was performed either

By I P assay, the A M A 225.28 a n d the p o l y c l o n a l A M A r e a c t e d with sections o f 7/10 m a l i g n a n t m e l a n o m a s . T h e

on 5 lim thick cryostat sections of snap frozen ( - 176~ C) human melanoma xenografts or on acetone fixed (1 min at 4~ C) cytocentrifuged smears of melanoma cells using, as appropriate, fluoresceinated goat anti-rabbit globulin or rabbit anti-mouse globulin. Membrane staining was performed using suspension of viable tumor cells. Specificity of staining was validated by the use of appropriate negative controls, including omission of primary antibodies, and substitution of primary antibodies by irrelevant isotypematched immunoglobulins.

Immunoperoxidase (IP). The indirect immunoperoxidase method described by Nakane and Pierce (1966) was used. After blocking of tissue endogenous peroxidase, deparaffinized sections were incubated with an appropriate concentration of the primary antibody (anti-keratin antibody - 1 : 50, anti-CEA antibody - 1 : 300, antivimentin antibody - 1 : 100, monoclonal AMA - 1 : 100, polyclonal AMA --1:100) for 60 min, washed and then exposed, as appropriate, to peroxidase labelled anti-rabbit or anti-mouse antibody (DAKO). The stain reaction was demonstrated by the amino-ethyl carbazole cytochemical procedure. Assay of cell proliferation. Cells from melanoma xenografts were cultured in 10% FCS supplemented RPMI (control), 10% FCS supplemented RPMI containing 1 g MTX/mL, or 5 g adriamycin/ mL or in RPMI without FCS. Cells were counted daily for 5 days using a Coulter counter (Coulter Electronics Inc., Hialeah, Florida). Inhibition of proliferation was expressed as percent inhibition i.e. 100

[number of cells in treated preparation x 100] [number of cells in control preparation]

Tumor xenografts. Trypan blue impermeable M21 cells (107 cells per mouse) were inoculated subcutaneously into (nu/nu) BALB/c mice (Harlan Sprague Dawley, Inc., Madison, Wisconsin). Fifty days later the mice were killed and tumors removed for examination.

Fig. 1. Iris melanoma surgical specimen. Stained for vimentin. x 150

83

"

9

Ib

i

tb.

t

,1 \i ;

Fig. 2. Iris melanoma stained for cytokeratins (AE1 and AE3). x 400

polyclonal AMA reacted with sections of 4/8 nevi and with none of the 8 squamous cell carcinomas and 8 lymphomas examined. The monoclonal AMA reacted with 5/8 nevi, 3/8 squamous cell carcinomas and with none of the 8 lymphomas examined. All the skin melanomas reacted with the anti-vimentin antibody, and none reacted with any of the anti-keratin antibodies9 However, a single primary melanoma of the iris examined stained both for vimentin and cytokeratins (Figs9 1, 2).

Fig. 3. Cultured M21 melanoma cells 9 days. Stained for cytokeratin with polyclonal antibodies, x 150

!

IFL and IP analysis of melanoma M21 cells Up to the eighth day in culture, cells derived from human M21 xenografts reacted only with the anti-melanoma antibodies but not with anti-keratin and anti-CEA antibodies9 However, from the 9th day of culture approximately 10% of cells reacted with the polyclonal anti-keratin antibody (Fig. 3) and more strongly after 21 days with the monoclonal anti-keratin antibodies AE1 and AE2 (Fig. 4) and with the anti-CEA antibody (Fig. 5) but not with the monoclonal anti-keratin antibody AE3. The intensity of staining for these antibodies gradually increased from the 9th day and reached a plateau on the 18th day of culture when approximately 60% of cells reacted with these antibodies9 On all days examined the cells reacted with the anti-vimentin antibody and the anti-melanoma antibodies9 M21 cells that were expressing keratin, vimentin and CEA in vitro were transplanted subcutaneously into

Fig. 4. Cultured M21 melanoma cells 21 days. Stained for cytokeratin with monoclonal AE1/AE2 antibodies, x 400

84

~

e

:~, ..

Fig. 6. Cultured M21 melanomacells 9 days of adriamycin exposure. Globularmasses of cytokeratinfilamentsare evident (arrows). Monoclonal AEI antibody staining, x 400 Fig. 5. Cultured M21 melanoma cells 21 days. Stained for CEA. x 400

nude mouse. Melanoma cells harvested from the s.c. tumor on the 50th day after transplantation did not react with the anti-keratin and anti-CEA antibodies, but reacted with both the AMA and the anti-vimentin antibody. All Hela cells from culture reacted with the antikeratin antibodies irrespective of when they were harvested.

Effect of FCS concentration and drugs When M21 ceils were grown in RPMI without FCS, there was 60% inhibition of proliferation on the 3rd day and 50% on the 5th day compared to M21 cells grown in 10% FCS supplemented RPMI. Exposure of M21 cells to MTX and adriamycin also inhibited proliferation by 40 to 60% from the 3rd day onward. Irrespective of whether RPMI was supplemented with FCS and whether the medium contained MTX or adriamycin from the 9th day and onward a proportion of cultured M21 cells reacted with both anti-keratin and anti-CEA antibodies. The adriamycin-treated melanoma cells revealed globular masses in the cytoplasm that reacted with the anti-keratin antibodies on immunofluorescence assay (Fig. 6).

Characterization of extracts of M21 cells

Fig. 7. Electron micrograph of extracts of M21 melanoma cells. Negative stained with phosphotungsticacid. Intermediatefilaments are evident (arrows). • 70000

EM and IEM examinations of extracts derived from M21 cells that were maintained in vitro and had shown the presence of CEA and keratin by IFL revealed large numbers of filaments (Fig. 7) that reacted with the antikeratin antibody. However, a very small number of similar filaments were also seen in the extracts derived from

xenograft-derived M21 cells. On PAGE, the extract from the M21 cells maintained in vitro showed a prominent wide band in the region of 45-46 KD. A faint thin band could also be seen in this region when the extract from xenografted M21 cells was subjected to PAGE. When

85

A

B

/

\

Fig. 8. Double diffusion assay between extracts of cultured M21 melanoma cells A and HeLa cells B and polyclonal antikeratin antibody (center well). Line of identity is evident

extracts from cultured M21 melanoma and HeLa cells were subjected to immunodiffusion against the rabbit anti-keratin antibody a single precipitin line showing reaction of identity developed between the antibody and the two extracts (Fig. 8). Discussion

Immunohistochemistry has found application in the diagnosis of metastatic melanoma. Because of the remarkable heterogeneity of morphology typical of melanomas and its propensity to metastasize to all organs of the body, melanoma appears in the differential diagnosis of almost all poorly differentiated anaplastic metastases. Besides their expression of S100 protein which still remains as the best initial screen for melanoma (Nakajima et al. 1982) other antibodies are now available for its detection. These include HMB-45/50 (Gown et al. 1986), NK1/C3 (Vennegoor et al. 1985), IKH-1 (Ishihara et al. 1989) and HMSA-1/2 (Maeda et al. 1988) among them. A recent comparison of these antibodies which all work effectively on retrospective material demonstrates the utility especially of NK1/C3 and HMB-45/50 (Thomson and Mackie 1989). It has become almost universally accepted that melanomas do not express cytokeratins (Caselitz et al. 1983; Huszar etal. 1983; Miettinen et al. 1983 and Ramaekers et al. 1983). This dogma has been accepted and has led to an entrenched position concerning unusual staining results with respect to other intermediate filaments. However, several workers have sounded a note of caution in that cytokeratins have been detected in a small percentage of clinical specimens (Gatter et al. 1985; Lambe et al. 1988). Whether this represents true expression of cytokeratins or is due to unexplained cross-reacting epitopes has been examined by Western bloting of extracted material (Zarbo et al. 1989). A recent study shows that if frozen sectioned specimens are stained a significant number of both primary and secondary melanomas stain for cytokeratin with antibodies CAM 5.2, MAK-6 and AE1 (Miettenen and Franscilla 1989). These antibodies detect, among the cytokeratin especially, the low molecular weight or acidic

isoforms (Moll et al. 1982; Sun et al. 1984). It is significant that melanomas showing epithelioid histology tend to be most heavily stained. There is, however, considerable heterogeneity in the staining pattern and varying percentages of the tumor cells actually decorate with antibodies against keratins. These findings raise a note of caution with respect to the interpretation of immunostaining for intermediate filaments when based on single reagents. The need for the application of batteries of reagents is clear. The results of our PAGE, immunodiffusion, EM and IEM studies show that a small amount of low molecular weight keratin (undetectable by IFL and IP) is present in M21 melanoma cells grown as a subcutaneous solid tumor in nude mice and that the amount of keratin is increased to detectibility by immunoeytochemical methods when M21 cells are maintained in vitro for 9 days or more. The reactivity of almost 100% of M21 cells with MAB 225.28 and the rabbit anti-melanoma antibody (both of which have high melanoma specificity) at all times in vitro, demonstrate that M21 cells were not contaminated and overgrown by other cells during our study. The only other demonstration of keratin proteins in human melanoma cells either growing as a xenograft in nude mice or maintained in culture is that of Trejdosniewicz and associates in 1986. In examination of six human melanoma and two murine cell lines only one, namely M5 consistently showed cytokeratins of the low molecular weight variety. Other studies on cultured human melanoma cells have, however, failed to show cytokeratin expression (Lombardi and Castellucci 1990). Although we could not detect cytokeratin by IP in a number of formalin-fixed specimens of both primary and metastatic melanomas of the skin, AE1/AE3 positive cytokeratin was demonstrated in the single specimen of the melanoma of the iris examined by us. PAGE of the cellular extracts showed the major component of the cytokeratin in M21 cells to be a 46 KD protein. It is interesting that a 46 KD keratin has been observed to be substantially increased in both proliferating normal and neo-plastic epithelial cells (Weiss et al. 1988) e.g. human esophageal epithelium and mouse mammary epithelium. Consistent with the pattern of expression of the 46 KD cytokeratin in M21 cells, the normal profile of keratin expression was restored after xenotransplantation of cultured human esophageal cells in nude mice. In mouse mammary epithelium the expression of the 46 KD cytokeratin was also independent of the proliferative status of the epithelium. Although vimentin is regarded as the intermediate filament typical of human melanoma has been routinely demonstrated in melanoma specimens removed from patients, the presence of keratin in human melanoma M21 cells is not surprising because keratin is known to be a structural component of all cells of ectodermal origin (Franke et al. 1978b). Melanomas are of neuroectodermal origin as they originate from melanocytes which in their part are derived from migrated neural crest cells. Thus the detection of cytokeratin in unpigmented tumor cells especially those in malignant effusion should not exclude the diagnosis of malignant melanoma.

86

It is widely accepted that the expression of distinctive intermediate filaments is a stable function of given cell types. However, co-expression ofcytokeratins with other members of the intermediate filament family has been seen in a number of situations and this apparent lineage infidelity raises questions about the validity of diagnoses dependent on single marker stains. For example, cytokeratins are expressed with vimentin in normal and neoplastic smooth and cardiac muscle cells (Huitfeld and Brandtzaeg 1985; Brown et al. 1987; Norton et al. 1987; Gown et al. 1988), in rhabdomyosarcomas (Coidre et al. 1988), a variety of epithelial neoplasms including renal cell carcinoma (Holthofer et al. 1983), anaplastic thyroid carcinoma (Bolen and McNutt 1987) lung carcinoma (Gatter et al. 1986) spindle cell carcinoma of the urinary bladder (Ro et al. 1990, 1991) and sarcomatoid carcinoma of the breast (Meis et al. 1987). Mesotheliomas (Churg 1985), meningiomas (Meis et al. 1986), chordomas (Salisbury and Isaacson 1985), synovial sapcomas (Leader et al. 1987), epithelioid sarcomas (Manivel et al. 1987), malignant fibrous histiocytomas (Weiss et al. 1988) and small cell epithelial tumors of the peritoneum (Ordonez et al. 1989), similarly show dual intermediate filament expression. Plasmocytomas express both types of intermediate filaments (Sewell et al. 1986) and extrafollicular reticulum cells in lymph nodes, although seemingly of fibroblastic nature have also been shown to express cytokeratins (Iuzzlino and Bontempini 1989). Further proof of the great flexibility in phenotypic expression of intermediate filaments is in the presence of cytokeratins in cultured fibroblastic cells (Zackroff et al. 1984). Keratins have also been shown to be expressed along with nurofilaments in neuroectodermal cells (Blobel et al. 1985) and with desmin and vimentin in mesothelial cells (Van Muijen et al. 1987) as well as in the smooth and striated muscle tumors previously cited. The precise mechanism underlying the increase in the amount of keratin and the appearance of CEA in cultured melanoma cells remains to be defined. The comparatively rapid rate of cell proliferation in vitro and/or the altered milieu associated with growth in vitro could have contributed to the increased synthesis. However, the state of proliferation of M21 cells does not appear to play a significant role because reductions in the rate of proliferation of M21 cells brought about either by decreasing the amount of FCS in the culture medium or by exposure to MTX or adriamycin did not prevent the expression of keratin and CEA. Exposure to adriamycin did, however, alter the distribution of keratin in M21 cells. The keratin was aggregated in a globular mass that impinged on the nucleus. The aggregation of keratin could have been caused by adriamycin-induced formation of electrostatic and hydrophobic bonds in and between cytokeratin filaments.

Acknowledgments. The authors thank Drs. S. Ferrone (monoclonal antimelanoma antibody), K.M. Sheikh (polyclonal anti-keratin antibody) and T.T. Sun (monoclonal anti-keratin antibodies) for providing various antibodies, Mrs. M. Mammen, Mr. D. Sadi and Mrs. D. Dean for their excellent technical assistance and Ms. H. Maxner for her typing assistance. This work was supported by

Grant Nos. MT 6922 (T.G.) and MA 6800 (G.R.) from the Medical Research Council of Canada.

References Blobel GA, Gould VE, Moll R, Lee I, Huszar M, Geiger B, Franke WW (1985) Co-expression of neuroendocrine markers and epithelial cytoskeletal proteins in bronchopulmonary neuroendocrine neoplasms. Lab Invest 52:39-51 Bolen JW, McNutt MA (1987) Cytoskeletal intermediate filaments. Practical applications of intermediate filament analysis. Ultrastruct Pathol 11:175-189 Brown DC, Theaker JM, Banks PM, Gatter KC, Mason DY (1987) Cytokeratin expression in smooth muscle and smooth muscle tumors. Histopathol 11 : 477-486 Caselitz J, Janner M, Breitbart E, Webber K, Osborn M (1983) Malignant melanoma contain only the vimentin type of intermediate filament. Virchows Arch [A] 400:43-51 Churg A (1985) Immunohistochemical staining for vimentin and keratin in malignant mesothehoma. Am J Surg Pathol 9:360365 Coidre JM, DeMascarel A, Trojani M, DeMascarel I, Pages A (1988) Immunohistochemical study of rhabdomyosarcoma. Unexpected staining of S-100 protein and cytokeratin. J Pathol 155:127-132 Cosgrove M, Fitzgibbons PL, Sherrod A, Chanrasona PT, Martin SE (1989) Intermediate filament expression in astrocytic neoplasms. Am J Surg Pathol 13:141-145 Denk H, Krepler R, Lakinger R, Artilibe U, Franke WW (1982) Biochemical and immunocytochemical analysis of the intermediate filament cytoskeleton in human hepatocellular carcinomas and in hepatic nodules in mice. Lab Invest 46:584-596 Franke WW, Moll R (1987) Cytoskeleta| components of lymphoid organs. Synthesis of cytokeratin 8 and 18 and desmin in subpopulation of extrafollicular, reticulum cells of human lymph node, tonsils and spleen. Differentiation 36:145-163 Franke WW, Schmid E, Osborn M, Weber K (1978a) Different intermediate-sized filaments distinguished by immunofluorescent microscopy. Proc Natl Acad Sci USA 75:5034-5038 Franke WW, Weber K, Osborn M, Schmid E, Freudenstein C (1978b) Antibody to prekeratin. Decoration of tonofilamentlike arrays in various cells of epithelial character. Exp Cell Res 116: 429-445 Franke WW, Mayer D, Schmid E, Denk H, Boren-Freund E (1981) Differences in expression of cytoskeletal proteins in cultured rat hepatocytes and hepatoma cells. Exp Cell Res 134:345-365 Gatter KC, Ralfkiaer E, Skinner J, Brown D, Heryet A, Pulford KAF, Hou-Jensen K, Mason DY (1985) An immunocytochemical study of malignant melanoma and its differential diagnosis from other malignant tumors. J Clin Pathol 38:1353-1357 Gatter KC, Dunnill MS, Van Muijen GNP, Mason DY (1986) Human lung tumors may coexpress different classes of intermediate filaments. J Clin Pathol 39:950-954 Ghose T, Norvell ST, Guclu A, MacDonald AS (1975) Immunotherapy of human malignant melanoma with chlorambucil-carrying antibody. Eur J Cancer 11:321-326 Goldenberg DM, Sharkey RM, Primus FJ (1978) Immunocytochemical detection of carcinoembryonic antigen in conventional histology specimens. Cancer 42:1546-1553 Gown AM, Vogel AM, Hoak D, Gough F, McNutt MA (1986) Monoclonal antibodies specific for melanocytic tumors distinguish subpopulations of melanocytes. Am J Pathol 123:195203 Gown AM, Boyd HC, Chang Y, Ferguson M, Reichler B, Tippens D (1988) Smooth muscle cells can express cytokeratins o f " simple" epithelium. Immunocytochemical and biochemical studies in vitro and in vivo. Am J Pathol 132:223-232 Holthofer H, Miettenen A, Paasivuo R, Lehto VP, Linder E, Alfthan O, Virtanen I (1983) Cellular origin and differentiation of renal carcinomas. A fluorescent microscopic study with kid-

87 ney-specific antibodies anti-intermediate filament antibodies and lectins. Lab Invest 49:317-325 Huitfeld HS, Brandtzaeg P (1985) Various keratin antibodies produce immunohistochemical staining of human myocardum and myometrium. Histochemistry 83: 381-389 Huszar M, Halkin H, Herczeg E, Bubis J, Geiger B (1983) Use of antibodies to intermediate filaments in the diagnosis of metastatic amelanotic melanomas. Humn Pathol 14:1006-1008 Ishihara M, Hashimoto K, Taniguiche Y, Cho KH, Setoyama M, Richards J, Benedetto LG, Pietruk T, Mehregan AH (1989) Anti-melanoma monoclonal antibodies which have high specificity (IKH-1) in formalin-fixed paraffin-embedded tissue. J Invest Dermatol 92: 450 Iuzzolino P, Bontempini L (1989) Keratin immunoreactivity in extrafoUicular reticular cells of lymph node. Am J Clin Pathol 92:239-240 Kanazawa S, Morikawa S, Morada T, Yasuhira K, Matsubara Y, Ikeda S, Soeda S, Fujimura M (1982) CEA producing cultured cell lines enable detection of autoantibodies in sera from patients with gastrointestinal cancer. Cancer 50:1755-1782 Lambe CDK, Cartun RW, Knibbs DR, Ricci A Jr (1989) Immunocytochemical reactivity with cytokeratin monoclonal antibodies in malignant melanoma. Further evidence dictating cautious interpretation of immunostaining results. Lab Invest 60: 50A Leader M, Patal J, Collins M, Kristen H (1987) Synovial sarcomas. True carcinosarcomas. Cancer 59:2096-2098 Lombardi T, Castellucci M (1990) Bowes human melanoma cell line. An immunocytochemical study. Virchows Archiv [B] 58:181-183 Maeda K, Maeda K, Jimbow K (1988) Positive reactivity of dysplastic melanocytes with a monoclonal antibody against melanoma melanosomes Moab HMSA-2. J Invest Dermatol 91 : 247-250 Manivel JC, Wick MR, Dehner LP, Sibley RK (1987) Epithelioid sarcoma. An immunohistochemical study. Am J Clin Pathol 87:319-326 Meis JM, Ordonez NG, Bruner JM (1986) Meningiomas. An immunocytochemical study of 50 cases. Arch Pathol Lab Med 110: 934--937 Meis JM, Ordonez NG, Gallager HS (1987) Sarcomatoid carcinoma of the breast: an immunohistochemical study of six cases. Virchows Arch [A] 410:415-421 Miettinen M, Lehto VP, Virtanen I (1983) Presence of fibroblasttype of intermediate filaments (vimentin) and absence of neurofilaments in pigmented nevi and melanomas. J Cutan Pathol 10:188-192 Miettinen M, Franssila K (1989) Immunohistochemical spectrum of malignant melanoma: the common presence of keratins. Lab Invest 61 : 623-628 Moll R, Franke WW, Schiller DL, Geiger B, Krepler B (1982) The catalogue of human cytokeratins. Patterns of expression in normal epithelia, tumours and cultured cells. Cell 31 : 11 24 Nakajima T, Watanabe S, Sato Y, Kameya GT, Hirota T, Shimosato Y (1982) An immunoperoxidase study of S100 protein distribution in normal and neoplastic tissue. Am J Surg Pathol 6:715-727 Norton AJ, Thomas JA, Isaacson PG (1987) Cytokeratin-specific monoclonal antibodies are reactive with tumors of smooth muscle derivation: an immunocytochemical and biochemical study using antibodies to intermediate filament cytoskeletal protein. Histopathol 11 : 487-499 Ordonez NG, Zirkin R, Bloom RE (1989) Malignant small-cell epithelial tumor of the peritoneum coexpressing mesenchymaltype intermediate filaments. Am J Surg Pathol 13:413-421

Ramaekers FCS, Puts JJG, Moesker O, Kant A, Vooijs GP, Jap PHK (1983) Intermediate filaments in malignant melanomas. Identification and use in surgical pathology. J Clin Invest 71 : 635-643 Ro JY, Ayala AG, Wishnow KI, Ordonez NG (1991) Sarcomatoid bladder carcinoma: a clinicopathologic and immunohistochemical study of 44 cases. Surg Pathol (in press) Salisbury JR, Isaacson PG (1985) Demonstration of cytokeratins and an epithelial membrane antigens in chordomas and human fetal notochord. Am J Surg Pathol 9:791-979 Sewell HF, Thompson WD, King DJ (1986) IgD myeloma/immunoblastic lymphoma cells expressing cytokeratin. Br J Cancer 53 : 695-696 Sun TT, Green H (1978 a) Immunofluorescence staining of keratin fibers in cultured cells. Cell 14:469-476 Sun TT, Green H (1978b) Keratin filaments of cultured human epidermal cells. J Biol Chem 253:2053-2060 Sun TT, Shih G, Green H (1979) Keratin cytoskeleton in epithelial cells of internal organs. Proc Natl Acad Sci USA 76:2813-2817 Sun TT, Eichner R, Schener A, Cooper D, Nelson WG, Weiss RA (1984) In: Levine AJ, Van de Voude GF, Tropp WC, Watson JD (eds) Cancer cells. The transformed phenotype, vol 1. Cold Spring Harbour, New York, p 169 Thomson W, MacKie R (1989) Comparison of five antimelanoma antibodies for identification of melanocytic cells on tissue sections in routine dermatopathology. J Am Acad Dermatol 21 : 1280-1284 Trejdosiewicz LK, Southgate J, Kemshead JK, Hodges GM (1986) Phenotypic analysis of cultured melanoma cells. Expression of cytokeratin-type intermediate filaments by6 the M-5 human melanoma cell line. Exp Cell Res 164:388-398 Tsutsumi Y, Nagura H, Watanabe K (1984) Immunohistochemical observations of carcinoembryonic antigen (CEA) and CEArelated substances in normal and neoplastic pancreas. Pitfalls and caveats in CEA immunohistochemistry. Am J Clin Pathol 82:535-542 Urmacher C (1990) Histology of normal skin. Am J Surg Pathol 14:671-686 Van Muijen GNP, Ruiter DJ, Warnaar SO (1987) Coexpression of intermediate filament polypeptides in human fetal and adult tissues. Lab Invest 57:259-269 Vennegoor VC, Calafat J, Hageman P, Van Buitenen F, Kolk A, Rumke P (1985) Biochemical characterization and cellular localization of a formalin-resistant melanoma-associated antigen reacting with monoclonal antibody NK1/C3. Int J Cancer 35: 287-295 Virtanen I, Miettenen M, Lehto VP, Karniniemi AL, Paasivua R (1985) Diagnostic application of monoclonal antibodies to intermediate filaments. Ann NY Acad Sci 455:635-648 Von Kleist S, Chavanel G, Burtin P (1972) Identification of a normal antigen that cross-reacts with the carcinoembryonic antigen. Proc Natl Acad Sci USA 69: 2492-2494 Weiss SW, Bratthauer GL, Morris PA (1988) Postirradiation malignant fibrous histocytoma expressing cytokeratin. Implications for the immunodiagnosis of sarcomas. Am J Surg Pathol 12:554-558 Zackroff RV, Goldman AE, Jones JCR, Steinert PM, Goldman RD (1984) Isolation and characterization of keratin-like proteins from cultured cells with fibroblastic morphology. J Cell Biol 98:1231-1237 Zarbo RJ, Gown AM, Visscher DM, Crissman JD (1989) Low molecular weight cytokeratin expression by malignant melanoma. Immunohistochemical Western blot analysis. Lab Invest 60:109A

Keratin and carcinoembryonic antigen (CEA) in human melanoma cells.

Human melanomas are known to contain vimentin intermediate filaments but there has been some dispute about their expression of cytokeratins. The cytop...
2MB Sizes 0 Downloads 0 Views