coexpression of simple epithelial keratins and vimentin by ... · the mesothelium is the simple...

[CANCER RESEARCH 44, 2991-2999, July 1984]

Coexpression of Simple Epithelial Keratins and Vimentin by HumanMesothelium and Mesothelioma in Vivo and in Culture

Paul J. LaRocca1 and James G. Rheinwald2

Division of Cell Growth and Regulation, Dana-Farber Cancer Institute and Department of Physiology and Biophysics, Harvard Medical School,Boston, Massachusetts 02115

ABSTRACT

We have determined the intermediate filament proteins presentin normal and malignant mesothelium in vivo. Pure sheets ofnormal lung mesothelium were obtained by transfer to agar-coated slides or by gentle scraping and cytocentrifugation. Cy-

toplasmic filament networks in the mesothelium were labeled viaindirect immunofluorescence both by anti-Mr 40,000 keratin andanti-vimentin antisera. Two-dimensional gel electrophoresis ofthe Triton:high-salt-insoluble proteins of normal lung mesothel

ium disclosed the presence of vimentin and all but the largest(M, 55,000) of the four simple epithelial keratins synthesized bymesothelial cells in culture. Samples of three peritoneal and threepleural mesotheliomas were found to contain either all four simpleepithelial keratins or all but the M, 55,000 keratin. Notably, noneof the keratins characteristic of stratified and many glandularepithelia and their malignant forms was present in these meso-,theliomas. Two mesothelioma samples from which the tumorcells could be obtained free of other cell types were found tocontain vimentin as well as simple epithelial keratins and tosynthesize these same proteins during short-term culture. None

of the mesotheliomas placed in culture grew progressively inmedium optimal for the growth of normal mesothelial cells. Thesedata demonstrate that malignant mesothelial cells preserve thenormal pattern of intermediate filament protein synthesis, including coexpression of simple epithelial keratins and vimentin, andsuggest the use of this characteristic as an aid in the identificationof cells of mesothelial origin.

INTRODUCTION

The mesothelium is the simple squamous epithelium that linesthe pleural, pericardial, and peritoneal cavities and covers theouter surfaces of the organs contained therein (6, 7). Unlike mostother epithelia, the mesothelium is of embryonic mesodermalorigin. Perhaps because of this the malignant form, mesothelioma, can adopt either sarcomatous or carcinomatous forms,often making differential diagnosis difficult (9, 18, 37).

This laboratory has recently determined the cell type-specific

requirements of normal, diploid human mesothelial cells for rapidgrowth in long-term, serial culture (4). Mesothelial cells are

different from most other normal epithelial cell types in that theyexpress in culture high levels of 2 different types of intermediatefilament proteins: keratins and vimentin (42,44). Because normalmesothelial cells that detach from the mesothelium and float inascites fluid also contain both vimentin and keratins (4), we

1Recipient of National Research Service Awards T32 CA09361 and

IF32AG05303 Â¡npartial support of this investigation.1 Recipient of grants from the National Cancer Institute and the National Institute

on Aging.Received July 25,1983; accepted March 30,1984.

wanted to determine whether vimentin is present in the cells ofthe intact mesothelium also, or whether it is only acquired aftercell detachment or cultivation, as has been suggested by others(12, 21, 32, 39). The set of keratin proteins synthesized bynormal human peritoneal mesothelial cells in culture is differentfrom that of stratified squamous epithelial and many glandularepithelial cell types (22, 27, 38, 43). We wished to determinewhether mesotheliomas express any keratins different fromthose of normal mesothelium, inasmuch as squamous cell carcinomas often synthesize keratins inappropriate for their cell typeof origin (23, 43, 44). We report here that normal mesothelialcells in the intact tissue coexpress keratins and vimentin, thatthe largest keratin M, 55,000 appears to be induced by conditionsof culture, and that mesotheliomas express vimentin and thesame set of keratins as does normal mesothelium.

MATERIALS AND METHODS

Preparation of Mesothelial Tissues. Sheets of normal lung mesothelium were obtained free of underlying connective tissue by a modificationof the imprinting method (42). A molten solution of 1% agarose (SeaKem) in water was applied to glass slides and dried overnight. Thesurface of freshly excised human fetal lung was rinsed in isotonic PBS3

and exposed to the air. Just as the mesothelium began to appear dry,an agarose-coated slide was pressed against the lung for several sec

onds. When the slide was pulled away, parts of the mesothelial cell layeradhered to the agar and were peeled off. Slides were examined underan inverted microscope with phase contrast optics to identify areascontaining large sheets of mesothelium. These were rinsed with PBS toremove any blood cells. Slides were then either briefly air-dried andimmersed in cold (-20Â°) methanol to fix for immunofluorescence micros

copy, or the agarose and cell material were scraped into Triton:high-salt

extraction buffer and stored frozen for future electrophoretic analysis.Tissues to be used as electrophoresis samples were collected on slidesthat had been coated with electrophoretically purified agarose in orderto reduce the levels of contaminants which would be stained in thesubsequent silver staining procedure. As an alternate method for obtaining small sheets and clusters of mesothelial cells for Â¡mmunofluores-

cence, the lung surface was gently scraped with a rubber policeman intoPBS. The cell suspension was then diluted 1:1 with serum and appliedto a glass slide by spinning in a cytocentrifuge. The slides were then airdried and fixed in cold methanol. For these experiments, lungs wereobtained from 18 to 22-week-old human fetuses or from adult rats.

Surgical specimens of solid peritoneal and pleural mesotheliomas(kindly provided by Dr. J. M. Corson, Brigham and Women's Hospital)

were collected within 1 h after resection and minced into 1 cu mm pieces.Some fragments from each tumor were placed in culture, and some werefrozen in Triton:high-salt buffer until extracted for electrophoresis. Me

sothelioma cells were also obtained from a pleural effusion of mesothelioma (MS-7). This effusion was found by cytopathological evaluation to

consist of >90% mesothelioma cells. The cells were concentrated by

3The abbreviation used is: PBS, phosphate-buffered saline (137 mm NaCI:3 HIM

KCI:16 mw HajrtPQt:2muKHJPO*, pH 7.3).

JULY 1984 2991

on March 2, 2021. © 1984 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

P. J. La/tocca and J. G. Rheinwald

centrifugation from 200 ml of effusion. One fourth of the cell pellet wasextracted for electrophoretic analysis, and the rest was used to initiatecultures.

Frozen ampuls of cells concentrated from the ascites fluid of a patientwith an ovarian surface epithelial neoplasia were thawed and applied toslides by cytocentrifugation for examination by immunofluorescence. Thiswas the same ascites fluid from which the normal mesothelial strain LP-

9 have been derived (4, 43).Indirect Immunofluorescence Microscopy. Anti-M, 40,000 keratin

antiserum, raised in rabbits, is specific for this particular keratin (43, 44).Anti-stratum corneum keratins antiserum, raised in rabbits, reactsstrongly with epidermal keratins and cross-reacts to various degreeswith all the "mesothelial" or "simple epithelial" keratins (43, 44). An anti-

human fibroblast vimentin antiserum was raised in rabbits. An anti-

vimentin antiserum raised in guinea pigs was kindly provided by Dr.Werner Franke, German Cancer Research Center, Heidelberg, WestGermany. Mouse monoclonal antibody OC125, which recognizes a surface antigen of ovarian carcinoma cells (1), was obtained from Dr. RobertBast of this institute. Rhodamine- and fluorescein-conjugated secondary

antibodies were obtained from Miles and Cappel Laboratories. Theprocedures used for indirect immunofluorescence microscopy and photography has been described (4,44).

Cell Labeling, Extraction, and Electrophoresis. Methods were similar to those previously described (4, 43, 44). Proteins synthesized bycultured cells were labeled by a 4-hr incubation with [^SJmethionine (50

ÃŸd/ml) (specific activity, ~800 Ci/mmol, MEN Chemicals) in otherwisemethionine-free minimal Eagle's medium (Flow Laboratories) supple

mented with 20% dialyzed fetal calf serum. Labeled cultured cells,agarose slide-collected mesothelium, and minced tumor tissue wererinsed with PBS, Dounce homogenized in Triton:high-salt extractionbuffer (20 mw Tris-HCI, pH 7.4:0.6 M KCI:1% Triton X-100:1 ITIMphenyl-methylsulfonylfluoride) at 0Â°, and then sonicated for 45 sec. Lysateswere then centrifuged at 10,000 x g for 45 min at 4Â°.The insoluble

pellets (containing the keratins and vimentin) were then dissolved inO'Farrell 2-dimensional lysis buffer. The proteins were separated in the

first dimension by isoelectric focusing (29) and in the second dimensionby sodium dodecyl sulfate:polyacrylamide gel electrophoresis (10% ac-rylamide) (20). High-purity agarose (Sea Kern) was used to hold the

isoelectric focusing tube gel in place at the top of the sodium dodecylsulfate:polyacrylamide slab gel. The possible presence of basic (isoelectric points 7 to 8) keratins was examined by separating the proteins bynonequilibrium pH gradient electrophoresis for 3.5 hr in the first dimension, according to the method of O'Farrell ef al. (30). Labeled proteins

were detected by scintillation fluor-enhanced (Enhance; MEN Chemicals)

autoradiography.Protein Detection in Gels by Silver Staining. An ultrasensitive silver

method (25) was used with the following modifications. After fixation in10% glutaraldehyde, the gel was rinsed for at least 48 hr to morecompletely remove glutaraldehyde and consequently reduce the background staining. Ethanolamine (5 Â¿J/ml)was added to the citric acidfixative to prevent the formaldehyde from overdeveloping the silver (10).Wet gels were then photographed over a light box with Kodak TechnicalPan film (ASA 50), which was developed in a 1:36 dilution of HC-110 for

8 min. Gels were placed in a solution of 0.02% sodium carbonate andstored in sealed plastic bags for future examination.

Cell Culture Methods. A population of normal, diploid, human pleuralmesothelial cells (strain HPM-2) was cultured from pleural effusion fluid

withdrawn from an adult female. The normal, diploid, human peritonealmesothelial strain LP-9 has been described previously (4, 43). LP-9 andHPM-2 were grown as described (4), in either M199 or M199/

MCDB202:15% fetal calf serum (Sterile Systems):hydrocortisone (0.4Mg/ml; Calbiochem):epidermal growth factor (10 ng/ml; CollaborativeResearch). In order to maximize keratin synthesis (Fig. 3, b to d), cellswere grown to confluence during one passage in the absence of epidermal growth factor.

Cells from a pleural effusion mesothelioma (MS-7) and from a solidperitoneal mesothelioma (MS-10) which could be dislodged as clusters

of pure tumor cells were placed into culture in the growth mediumdescribed above.

Keratin Nomenclature. Keratins 40KM, 44KM, 52KM and 55KM, the"mesothelial keratins" described in Wu ef a/. (43), are keratins 19,18, 8,and 7, respectively, independently identified as "simple epithelial keratins"

by Moll ef a/. (22).

RESULTS

M, 40,000 Keratin Immunofluorescence of Mesothelia. Normal mesothelium was transferred from the outer surface of afetal human lung to an agarose-coated slide by an imprinting

technique described by Whitaker et al. (42). Cell shape, nuclearposition, and intermediate filament content was examined byphase-contrast microscopy (Fig. 1a) and indirect immunofluorescence with anti-40KM keratin. Favorable areas of the trans

ferred mesothelium were found to contain this keratin in a fibrousnetwork (Fig. 1b) similar in appearance to that of cultured epithelial cells. Heterogeneity of staining presumably resulted fromvariability in the amount of cytoskeleton transferred to the slide.In these preparations, the apical portion of the plasma membraneadheres to the agarose, and the cells apparently fracture suchthat variably amounts of the cytoplasmic contents are removedto the slide, even though the nucleus (or its outline) can be seenin each transferred cell (Fig. 1a).

Small clusters to large sheets of mesothelial cells could alsobe dislodged by scraping the lung surface with a rubber policeman and depositing the PBS:serum suspension of these cells ona slide with the aid of a cytocentrifuge (Fig. 1c). Clusters of easilyidentified mesothelial cells stained positively for the 40KM keratinand also revealed details of the cytoplasmic filament network(Fig. 1d). This was consistent with the results of our preliminaryexamination of paraffin sections of mesothelium by immunoper-

oxidase staining (44). The mesothelium was also stained stronglyby anti-total stratum corneum keratins, an antiserum whichcross-reacts with most of the keratins (43, 44) (data not shown),

as had been reported previously (5, 36).The M, 40,000 keratin was also detected by immunofluores

cence in secondary cultures of mesotheliomas MS-7 (Fig. 1ft)and MS-10 (data not shown). The mesothelioma cells had a

morphology in culture (Fig. 1g) distinct from that of normalmesothelial cells (Fig. 1e). MS-7 and MS-10 could not growprogressively in the culture medium permissive for long-term

growth of normal pleural and peritoneal mesothelial cells andceased dividing in secondary culture. We initiated 5 other culturesof solid pleural and peritoneal mesotheliomas, but the only cellsthat grew had morphologies and growth requirements of eithernormal mesothelial cells or of normal connective tissue fibro-

blasts and were not tumorigenic in nude mice. Apparently themesothelioma cells present in those 5 tumor specimens couldnot grow at all under the conditions we used.

Detection of Vimentin in Normal Mesothelial Cells by Immunofluorescence. This laboratory previously had reported thatnormal mesothelial cells present among the cells in an ascitesfluid specimen, which had been allowed to attach to a coverslip,contained both keratin and vimentin (4). We wished to confirmthis coexpression in ascites fluid mesothelial cells that never hadbeen exposed, even briefly, to culture conditions, and also todetermine whether the cells of intact mesothelium contain vimentin. We applied human ascites fluid cells and intact rat lungmesothelium (dislodged by scraping with a rubber policeman) to

2992 CANCER RESEARCH VOL. 44



Intermediate Filament Proteins of Mesothelium and Mesothelioma

Table 1Keratin types and vimentin coexpression in normal and malignant mesothelia

Keratins

(40KM/19) (44KH/18) (46K>,) (52IW8) (55KM/7) VimentinNormal mesothelium8

Pleural v6 +Â° + - ++ +Pleural (HPM-2) c + ++/-++ +Peritoneal (LP-9) c + + +/- ++ +

Malignant mesotheliumPleural (MS-2)Pleural (MS-11)Pleural (MS-7)

Peritoneal(MS-4)Peritoneal (MS-5)Peritoneal(MS-10)

" Data summarized from Fig. 3 and Refs. 4 and 43.v, in vivo',c, in culture.

0 +, moderate amounts; ++, relatively high amounts; +/-, variable amounts among experiments; -, none

present." Data summarized from Figs. 3 and 4.eAmount in mesotheliomacells indeterminatebecause of connective tissue present in tumor specimen.' Relative amounts unknown due to partial proteolysis.

slides by cytocentrifugation and double labeled them with anti-vimentin and anti-keratin antisera. As shown in Fig. 2, a to c, all

mesothelial cells in rat lung mesothelium coexpressed vimentinand keratin. Consistent with the previous report from this laboratory (4), some of the ascites fluid cells coexpressed vimentinand keratin (Fig. 2, d to f). Vimentin-positive, keratin-negative

cells in the ascites fluid were assumed to be macrophages (3,11); these cells attached and spread under culture conditionsbut did not divide (4). The keratin-positive, vimentin-negative

cells were presumed to be neoplastic cells of ovarian surfaceepithelial origin because cells in the sample which stained withthe ovarian surface-specific monoclonal antibody OC125 (1) were

vimentin negative (Fig. 2, g to /) and keratin positive (data notshown). Thus, keratin and vimentin coexpression is a normalcharacteristic of mesothelial cells and is retained by cells thatbecome detached from the mesothelium.

Electrophoretic Analyses of Mesothelial and MesotheliomaCytoskeletal Proteins. The Triton:high-salt-insoluble protein

fraction extracted from normal and malignant mesothelial samples were analyzed by 2-dimensional gel electrophoresis to

determine precisely the intermediate filament proteins present(Table 1). Fetal human pleural mesothelium collected by strippingwith an agar-coated slide, precluding contamination with under

lying connective tissue, contained vimentin, keratins 40KM and44Km, and, in greatest abundance, keratin 52KM (Fig. 3a). Keratin55KM, synthesized by adult peritoneal mesothelial cells in culture(4, 43) (see Fig. 3, c and d) was not detectable. Normal pleuralmesothelial cells (HPM-2) synthesized vimentin and all 4 simple

epithelial keratins in culture, negating the possibility that cells ofpleural origin are unable to synthesize keratin 55KM. A rapidlyturning over, basic M, 46,000 keratin which is expressed atvariable levels by mesothelial cells growing in culture (4, 43) (seeFig. 3cO,also was not detected in the in vivo mesothelium.

A relatively pure cell suspension of pleural mesothelioma (MS-

7) and clusters of cells teased from a solid peritoneal mesothelioma (MS-10) were extracted directly and their intermediate

filament proteins compared with those synthesized by cells fromthese same tumors after 1 to 2 weeks in culture. MS-7 contained

vimentin and keratins 40KM, 44KM, and 52KM in vivo and also

synthesized these same proteins in secondary culture (Fig. 3, eand g). MS-10 contained vimentin, keratins 40KM and 44KM, asmall amount of keratin 52KM, and what appeared to be proteo-lytic digestion products of 52KM (Fig. 3f). Cultured MS-10 cells

synthesized vimentin and all 4 simple epithelial keratins (Fig. 3h).Thus, malignant mesothelial cells may also increase their expression of 55KM in culture.

Samples of 3 other solid peritoneal mesotheliomas and 4 otherpleural mesotheliomas were extracted and analyzed by 2-dimen

sional gel electrophoresis. Three samples gave no interpretablespots due to failures of isoelectric focusing or to paucity ofmesothelioma cells in a predominantly fibrous tumor. Four samples could be scored (Fig. 4). Keratins 40KM, 44KM, and 52KMwere present in all samples, with keratin 55KM present in all butone sample (Fig. 4o). One tumor (Fig. 4d) also contained aprotein migrating in the position of keratin 46KM. Vimentin wasdetected in 3 of the samples, but since these contained connective tissue, the vimentin content of the tumor cells themselveswas indeterminate.

Proteolysis of vimentin (14, 28) and of keratin 52KM (35) wasextensive in some tumor extracts (Fig. 3f; Fig. 4, a, b, and d),precluding quantitative estimation of these proteins when itoccurred. Evaluation of gels made from extracts of small tissuesamples was also complicated because the silver stainingmethod used to detect very low levels of protein also stainedminor impurities in the buffers and in the agarose used toimmobilize the tube gel during the second dimension of electrophoresis. These impurities were most common in the M, 56,000to 70,000 range (eg., cf. Fig. 3 c and d, and see Fig. 3 a and Fig.4 b, c, and d). Despite the staining background, no proteinsmigrating at the position of any known keratin (22,43) other thanthe simple epithelial keratins were identified in any of the mesothelioma extracts.

DISCUSSION

We previously had identified vimentin in the normal mesothelialcells present in ascites fluid and rising to higher levels in thesecells during serial culture (4). Nevertheless, it was necessary to

JULY 1984 2993



P. J. LaRocca and J. G. Rheinwald

examine mesothelial cells in situ, since it has been proposed thatvimentin may become expressed as a consequence of loss ofcell-cell contact after detachment from an epithelial sheet (21,

32). It is also known that some epithelial cell types (12, 39) butnot others (4, 34) abnormally synthesize vimentin under certainconditions of culture.

We have found by indirect immunofluorescence microscopy ofscraped mesothelium and by 2-dimensional gel electrophoreticanalysis of pure, agar slide-transfer preparations of mesothelium

that this tissue contains both keratins and vimentin in vivo.Previous reports of an absence of detectable vimentin in cross-

sections of chick (36) and mammalian (33) mesothelium byimmunofluorescence microscopy may have been due to therelatively low vimentin content in mesothelial cells compared tothe high levels present in fibroblasts and endothelial cells andbecause the mesothelium is very thin and easily damaged duringhandling and fixation. Mesothelial cells may express keratins andvimentin because they are derived from the embryonic meso-

derm (this subject is reviewed in Refs. 9, 18, and 37), inasmuchas most other cell types of mesodermal origin, including mes-enchyme-derived cells and cells of some hematopoietic lineages,

express vimentin (2, 8, 12, 13, 15, 31, 36).The types of keratins in normal mesothelial tissue are restricted

to the simple epithelial keratins, as we had predicted from theprevious results of this laboratory on peritoneal mesothelial cellsin culture (4, 43). Samples of human pleural mesothelium contained the 40 KM, 44 KM, and 52 KM simple epithelial keratins,but very little or no keratin 55 KM.This lack of keratin 55 KMwasnot simply a regional difference between pleural and peritonealmesothelial cells, however, because pleural mesothelial cellsexpressed the keratin 55 KM in culture. We do not know themechanism of this specific regulation of keratin 55 KMbut it couldbe the result of very rapid cell division or the abnormal culturesubstratum. We had found previously that synthesis and contentof all 4 keratins are controlled coordinately in mesothelial cells inculture as a function of growth rate (4, 34).

Solid and effusion mesotheliomas and 2 short-term cultured

mesothelioma populations contained 3 or 4 of the simple epithelial keratins and vimentin. Six of 9 surgical samples were successfully analyzed by 2-dimensional gel electrophoretic separa

tion and silver staining. Qualitative determination of the intermediate filament proteins was possible by this method, butquantitative estimation of the M, 52,000 simple epithelial keratinand vimentin was often prevented by substantial amounts ofproteolysis. These 2 intermediate filament proteins are particularly susceptible to digestion by an endogenous protease insome tissue types and cell lines (14, 28, 35). Furthermore, thepresence of vimentin in the mesothelioma cells of most samplescould not be confirmed by electrophoretic analysis because ofthe presence of connective tissue, including vimentin-containingfibroblasts and endothelial cells. Immunofluorescence or immu-

noperoxidase staining with antivimentin will undoubtedly be amore generally useful method for screening tumors for vimentinexpression.

We tried to culture mesothelioma cells in order to comparetheir growth requirements with those of normal mesothelial cellsand to identify selective markers for malignant transformation ofthis cell type. Only 2 of 7 tumors placed in culture yielded aviable population of cells other than normal fibroblasts andmesothelial cells, and these did not grow progressively in a

medium permissive for rapid, long-term growth of normal me

sothelial cells. This suggests that mesothelioma cells have significantly altered growth requirements which are difficult to satisfy in culture. We suspect that this is the reason that we havebeen unable to find any published reports of mesothelioma celllines. Normal mesothelial cells are anchorage independent, forming large colonies in semisolid medium.4 Thus, there is no basis

at present for culturing mesothelioma cells or selectively identifying them.

In addition to the coexpression of keratins and vimentin inmesothelioma, the specific subset of keratins expressed bymesotheliomas could be used to aid in the differential diagnosisof tumors in the pleural and peritoneal cavities. Tumors ofstratified epithelial and mammary epithelial origin contain 2 ormore keratins from a set of more basic (M, 58,000 and 56,000)and more acidic (M, 52,000, 50,000, and 46,000) keratins (keratins 5 and 6 and keratins 13, 14, and 17, respectively) (22-24,

26, 44). Carcinomas of the colon, stomach, and ovary would notbe distinguishable from mesothelioma by the keratin patternalone because these tumors also are restricted to simple epithelial keratins (22-24). Nevertheless, mesotheliomas are carci-

noembryonic antigen negative or very weakly positive (5,19, 40,41 ), while adenocarcinomas of the colon are strongly carcinoem-bryonic antigen-positive (17). Ovarian epithelial neoplasms can

be recognized by a specific antibody against a surface antigen(1). Our results indicate that vimentin is likely to be present inmesotheliomas, in contrast to its absence in tumors of all otherepithelial cell types (31). Vimentin has also been found recentlyin several renal carcinomas (16) which originate from anothermesoderm-derived epithelial cell type. The data we have pre

sented here suggest that coexpression of simple epithelial keratins and vimentin is a useful marker for distinguishing normaland malignant cells of mesothelial origin from those of most othercell types.

ACKNOWLEDGMENTSWe are very grateful to Dr. J. M. Corson, Brigham and Women's Hospital, for

providing mesothelioma samples and pathology reports and for helpful commentson the manuscript. Susan Rehwoldt and Terese O'Connell provided excellent

technical assistance with cell culture and preparation of vimentin antiserum. Wethank Dr. Oswaldo Alberti, Brigham and Women's Hospital, for providing fetal

human lung samples, Dr. Werner Franke, German Cancer Research Institute,Heidelberg, West Germany, for his gift of guinea pig anti-vimentin, and Dr. RobertBast of this institute for a sample of OC125 antibody. We appreciate Lynne Dillon's

skillful processing of the manuscript.

REFERENCES

1. Bast, R. C., Feeney, M., Lazarus, H., Nadler, L. M., Colvin, R. B., and Knapp,R. C. Reactivity of a monoclonal antibody with human ovarian carcinoma. J.Clin. Invest., 68.- 1331-1337, 1981.

2. Bennett, G. S., Fellini, S. A., Croop, J. M., Otto, J. J., Bryan, J., and Holtzer,S. Differences among 100-A filament subunits from different cell types. Proc.Nati. Acad. Sci. USA, 75: 4364-4368,1978.

3. Cain, H., Kraus, B., Krauspe, R., Osbom, M., and Weber, K. Vimentin filamentsin peritoneal macrophages at various stages of differentiation and with alteredfunction. Virchows Arch. B Cell Pathol., 42: 65-81, 1983.

4. Connell, N. D., and Rheinwald, J. G. Regulation of cytoskeletal composition inmesothelial cells: reversible loss of keratin and increase in vimentin duringrapid growth in culture. Cell, 34: 245-253.1983.

5. Corson, J. M., and Pinkus, G. S. Mesothelioma: profile of keratin proteins andcarcinoembryonic antigen: an immunoperoxidase study of 20 cases and comparison with pulmonary adenocarcinomas. Am. J. Pathol., 108: 80-88,1982.

6. Cunningham, R. S. On the origin of the free cells of serous exÃºdales. Am. J.

* P. J. LaRocca and J. G. Rheinwald, manuscript in preparation.





Physiol., 59: 1-36,1922.

7. Cunningham, R. S. The physiology of the serous membranes. Physio). Rev.,6: 242-280, 1926.

8. Dellagi, K., Vainchenker, W., Vinci, G., Paulin, D., and Brauet, J. C. Alterationof vimentin intermediate filament expression during differentiation of humanhemopoietic cells. EMBO J., 2: 1509-1515, 1983.

9. Donna, A., and Betta, P. G. Mesotheliomas: a new embryological approach toprimary tumors of coelomic surfaces. Histopathology (Oxf.), 5: 31-44, 1981.

10. Eschenbruch, M., and Burk, R. R. Experimentally improved reliability of ultrasensitive silver staining of proteins in polyacrylamide gels. Anal. Biochem.,r25: 96-99, 1982.

11. Franke, W. W., Schmid, E. A., Breitkreutz, D., Luder, M., Boukamp, P.,Fusenig, N. E., Osbom, M., and Weber, K. Simultaneous expression of twodifferent types of intermediate sized filaments in mouse keratinocytes proliferating m vitro. Differentiation, 14: 35-50,1979.

12. Franke, W. W., Schmid, E., Osbom, M., and Weber, K. Different intermediate-sized filaments distinguished by immunofluorescence microscopy. Proc. Nati.Acad. Sci. USA, 75: 5034-5038, 1978.

13. Franke, W. W., Schmid, E., Winter, S., Osbom, M., and Weber, K. Widespreadoccurrence of intermediate-sized filaments of the vimentin-type in culturedcells from diverse vertebrates. Exp. Cell Res., 723: 25-46,1979.

14. Granger, B. L, and Lazarides, E. Desmin and vimentin coexist at the peripheryof the myofibril Z disc. Cell, 78: 1053-1063, 1979.

15. Granger, B. L., Repasky, E. A., and Lazarides, E. Synemin and vimentin arecomponents of intermediate filaments in avian erythrocytes. J. Cell Bid., 92:299-312,1982.

16. Herman, C., Moesker, 0., Kant, A., Huysmans, A., Vooijs, G. P., and Ramaek-ers, F. C. S. Is renal cell (Grawitz) tumor a carcinosarcoma? Evidence fromanalysis of intermediate filament types. Virchows Arch. B Cell Pathd., 44: 73-83, 1983.

17. Heyderman, E., and Neville, A. M. A shorter immunoperoxidase technique forthe demonstration of carcinoembryonic antigen and other cell products. J. Clin.Pathol. (Lond.), 30: 138-140, 1977.

18. Klemperer, P., and Rabin, C. B. Primary neoplasms of the pleura. Arch. Pathol.,77:385-412,1931.

19. Kwee, W. S., Veldhuizen, R. W., Golding, R. P., Mullink, H., Stan, J., Donner,R., and Boon, M. E. Histologie distinction between malignant mesothelioma,benign lesion and carcinoma metastasis. Evaluation of the application ofmorphometry combined with histochemistry and immunostaining. VirchowsArch. A Pathol. Anat. Histol., 397: 287-299, 1982.

20. Laemmli, U. K. Cleavage of structural proteins during the assembly of the headof bacteriophage T4. Nature (Lond.), 227: 680-685, 1970.

21. Lane, E. B., Hogan, B. L. M., Kurkinen, M., and Garrels, J. I. Coexpression ofvimentin and cytokeratins in parietal endoderm cells of early mouse embryo.Nature (Lond.), 303: 701-704,1983.

22. Moll, R., Franke, W. W., Schiller, D. L., Geiger, B., and Krepier, R. The catalogof human cytokeratins: patterns of expression in normal epithelia. Cell, 37:11-24,1982.

23. Moll, R., Krepier, R., and Franke, W. W. Complex cytokeratin polypeptidepatterns observed in certain human carcinomas. Differentiation 23: 256-269,1983.

24. Moll, R., Levy, R., Czemobilsky, B., Hohlweg-Majert, P., Dallenback-Hellweg,G., and Franke, W. W. Cytokeratins of normal epithelia and some neoplasmsof the female genital tract. Lab. Invest., 49: 599-610, 1983.

25. Morrisey, J. H. Silver stain for proteins in polyacrylamide gels: a modifiedprocedure with enhanced uniform sensitivity. Anal. Biochem., 777: 307-310,1981.

26. Nelson, W. G., Battifora, H., Santana, H., and Sun, T-T. Specific keratins asmolecular markers of neoplasms with a stratified epithelial origin. Cancer Res.,44: 1600-1603, 1984.

27. Nelson, W. G., and Sun, T-T. The 50- and 58-kdalton keratin classes asmolecular markers for stratified squamous epithelial cell culture studies. J. CellBiol., 97:244-251,1983.

28. Nelson, W. J., and Traub, P. Properties of a Ca2*-activated protease specific

for the intermediate-sized filament protein vimentin in Ehrtich-ascites tumourcells. Eur. J. Biochem., 776: 51-57, 1981.

29. O'Farrell, P. H. High resolution two-dimensional electrophoresis of proteins.

Anal. Biochem., 250: 4007-4021, 1975.30. O'Farrell, P. Z., Goodman, H. M., and O'Farrell, P. H. High resolution two-

dimensional electrophoresis of basic as well as acidic proteins. Cell, 72:1133-1142, 1977.

31. Osborn, M., and Weber, K. Tumor diagnosis by intermediate filament typing:a novel tool for surgical pathology. Lab. Invest. 48: 372-394, 1983.

32. Ramaekers, F. C. S., Hagg, D., Kant, A., Moesker, O., Jap, P. H. K., andVooijs, G. P. Coexpression of keratin- and vimentin-type intermediate filamentsin human metastatic carcinoma cells. Proc. Nati. Acad. Sci. USA, 80: 2618-2622, 1983.

33. Ramaekers, F. C. S., Puts, J. G., Kant, A., Moesker, O., Jap, P. H. K., andVooijs, G. P. Use of antibodies to intermediate filaments in the characterizationof human tumors. In'. Cold Spring Harbor Symposia on Quantitative Biology,Vol. 46, pp. 331 -340. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory,

1981.34. Rheinwald, J. G., O'Connell, T. M., Connell, N. D., Rybak, S. M., Allen-

Hoffmann, B. L., LaRocca, P. J., Wu, Y.^J., and Rehwoldt, S. M. Expressionof specific keratin subsets and vimentin in normal human epithelial cellsâ€”afunction of cell type and condition of growth during serial culture. In: CancerCells 1/The Transformed Phenotype, pp. 217-228. Cold Spring Harbor, NY:Cold Spring Harbor Laboratory, 1984.

35. Schiller, D. L., and Franke, W. W. Limited proteolysis of cytokeratin A by anendogenous protease: removal of positively charged terminal sequences. CellBiol. Int. Rep., 7: 3,1983.

36. Schmid, E. A., Tappscott, S., Bennet, G. S., Croop, J., Fellini, S. A., Hoitzer,H., and Franke, W. W. Differential location of different types of intermediate-sized filaments in various tissues of the chicken embryo. Differentiation, 75:27-40,1979.

37. Stout, A. P., and Murray, M. R. Localized pleural mesothelioma. Arch. Pathol.,34:951-962, 1942.

38. Tseng, S. C. G., Jarvinen, M. J., Nelson, W. G., Huang, J.-W., Woodcock-Mitchell, J., and Sun, T.-T. Correlation of specific keratins with different typesof epithelial differentiation: monoclonal antibody studies. Cell, 30: 361-372,1982.

39. Virtanen, I., Lehto, V. P., Lehtonen, E., Vartio, T., Kurki, P., Wager, O., Small,J. V., Dahl, D., and Badley, R. A. Expression of intermediate filaments incultured cells. J. Cell Sci., 50: 45-63,1981.

40. Walts, A. A., Said, J. W., and Banks-Schlegel, S. Keratin and carcinoembryonicantigen in exfoliated mesothelial and malignant cells: an immunoperoxidasestudy. Am. J. Clin. Pathol., 80: 671-676, 1983.

41. Wang, N. S., Huang, S. N., and Gold, P. Absence of carcinoembryonic antigen-like material in mesothelioma: an immunohistochemical differentiation fromother cancers. Cancer (Phila.), 44: 937-943,1979.

42. Whitaker, D., Papadimitriou, J. M., and Walters, M. N.-l. The mesothelium:techniques for investigating the origin, nature and behaviour of mesothelialcells. J. Pathol., 732: 263-271, 1980.

43. Wu, Y.>J., Parker, L. M., Binder, N. E., Beckett, M. A., Sinard, J. H., Griffiths,C. T., and Rheinwald, J. G. The mesothelial keratins: a new family of cyto-skeletal proteins identified in cultured mesothelial cells and nonkeratinizingepithelia. Cell, 37: 693-703, 1982.

44. Wu, Y.-J. and Rheinwald, J. G. A new small (40kd) keratin filament proteinmade by some cultured human squamous cell carcinomas. Cell, 25: 627-635,1981.

JULY 1984 2995



â€¢â€¢>> ' ^v^

, . >-v% ; f, ^'.1Â¿v / ?'.' -"i * â€¢Â»il,â€¢#Mâ€¢Ã¤

r4 ' f<4V

Fig. 1. Morphology and M, 40,000 keratin immunofluoresÅ“nce of human mesothelium in vivo and of mesothelioma cells in culture, a, fetal lung mesothelium transferredto an agarose-coated slide; b, immunofluorescence of same field as a; c, scraped and cytocentrifuged fetal lung mesothelium; cf, immunofluorescence of same field as c;e, normal peritoneal mesothelial strain LP-9 in secondary culture; I, Â¡mmunofluorescence of a different LP-9 culture; g, pleural mesothelioma strain MS-7 in secondaryculture; h, immunofluorescence of a different culture of MS-7. a, c, e, and g are phase contrast; b, d, f, and h are anti-M, 40,000 keratin immunofluorescence. Bar in a is50 urn and is the same magnification for 6; bar in c is 50 Â»mand is the same magnification for d; bar in e is 200 urn and is the same magnification for g; bar in f is 50 ^mand is the same magnification for h.

2996




Fig. 2. Vimentin and keratin immunofluorescence of intact lung mesothelium and of mesothelial cells In ascites fluid, a to c, intact mesothelium scraped from rat lung.Double labeling with guinea pig antivimentin and fluorescein (b) and with rabbit anti-M, 40,000 keratin and rhodamine (c). Small, retractile cells in a are erythrocytes. d tof. human ascites fluid cells from patient with ovarian epithelial neoplasia, double labeled with guinea pig antivimentin and fluorescein (e) and with rabbit anti-stratumcomeum keratins and rhodamine (f). g to i, human ascites fluid cells from the same patient, double labeled with rabbit antivimentin and rhodamine (h) and with mousemonoclonal anti-ovarian carcinoma cell surface and fluorescein (/). Bar in a is 50 pm and is the same magnification for b to i. In d, arrows indicate mesothelial cells. In dand g, arrowheads indicate neoplastic ovarian epithelial cells. Vimentin-positive, keratin-negative cells in d to t are presumably macrophages.

JULY 1984 2997



tbt

a

â€¢':v'<Ã¯r--.'.Â¡'.'i.:â€¢â€¢â€¢.-,?'â€¢;."â€¢'"'

v

'T^

b

v

Cb

b' ^1a





K â€”y Â»

a^~

b

c -

Fig. 4. Two-dimensional gel electrophoretic separation of mesothelioma cytoskeletal proteins. Triton:high-salt-insoluble proteins of solid tumor samples separatedhorizontally by charge (loll, basic; right, acidic) and vertically by sodium dodecyl sulfate:polyacrylamide (10%) gel electrophoresis. Gels a and c were focused isoelectricallybetween pH 6.5 and 5.0. Gels b and d were separated by nonequilibrium pH gradient gel electrophoresis to permit detection of more basic proteins, a, peritonealmesothelioma MS-4; Â£>,pleural mesothelioma MS-2; c, peritoneal mesothelioma MS-5; d, pleural mesothelioma MS-11. All gels are silver stained. Lower case letters andarrows within each frame, position of keratins, vimentin, and actin, as in the legend of Fig. 3; arrowheads, proteolytic breakdown products of vimentin and the 52KÂ«keratin. * in frame a, prominent species of size and charge different from that of any known keratin, v, vimentin.

Fig. 3. Two-dimensional gel electrophoretic separation of the cytoskeletal proteins of normal human mesothelium and mesotheliomas. Triton:high-salt-insolubte proteinfractions of tissues, tumors, and cultured cells separated horizontally by their charge (left, basic; right, acidic) and vertically by sodium dodecyl sulfate;polyacrylamide(10%) gel electrophoresis. Gels a, c, e, and g were isoelectrically focused between pH 6.5 and 5.0 in the horizontal dimension. Gels b, d, f, and h were separated bynonequilibrium pH gradient gel electrophoresis in the horizontal dimension in order to permit detection of more basic proteins, a, human fetal lung mesothelium collectedby slide transfer (spots marked with * are agarose contaminants detected by silver staining); b, cultured adult lung mesothelial strain HPM-2; c and d. cultured adultperitoneal mesothelial cell strain LP-9; e, pleural effusion mesothelioma MS-7; f, peritoneal mesothelioma MS-10; g, secondary culture of MS-7; h, primary culture of MS-10. Proteins in a, d, and e were revealed by silver staining, in frame f by Coomassie blue staining, and in b, c, g, and h by autoradiography of [^Slmethionine-proteins.Lower case letters within each frame, positions of keratins and vimentin: a, 40Â«M/19;b, 44KM/18; c, 52KÂ«/8;d, 55KM/7; e, 46KM. Arrows, actin. c' is presumed to be a

proteolytic digestion product of keratin 52KM/8. v, vimentin.

JULY 1984 2999



1984;44:2991-2999. Cancer Res Paul J. LaRocca and James G. Rheinwald

and in Culturein VivoHuman Mesothelium and Mesothelioma Coexpression of Simple Epithelial Keratins and Vimentin by

Updated version

http://cancerres.aacrjournals.org/content/44/7/2991

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/44/7/2991To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



coexpression of simple epithelial keratins and vimentin by ... · the mesothelium is the simple...

Documents