ADHESION, GROWTH AND MORPHOLOGY OF HUMAN MESOTHELIAL CELLS ON

J. Cell Sci. 85, 133-147 (1986) 133Printed in Great Britain © The Company of Biologists Limited 1986

ADHESION, GROWTH AND MORPHOLOGY OF

HUMAN MESOTHELIAL CELLS ON EXTRACELLULAR

MATRIX

M. J. NIEDBALA1, K. CRICKARD2 AND R. J. BERNACKI1*1 Department of Experimental Therapeutics, Roswell Park Memorial Institute, Buffalo,NY 14263, USA2Department ofGynecology and Obstetrics, Buffalo General Hospital, Buffalo, NY 14203,USA

SUMMARY

Human mesothelial cells (HMC) cover a variety of serosal surfaces and have been shown to restupon an underlying subcellular basement membrane in vivo. Bovine corneal endothelial cellsproduce an extracellular matrix (ECM) in vitro that mimics HMC subcellular basement membraneand was found to modulate HMC adhesion, morphology and proliferation in vitro. Our resultsindicated that within minutes after plating, a high percentage (>80%) of HMC firmly attached toECM. Active cellular migration and subsequent proliferation were observed leading to theformation of a well-organized closely apposed cell monolayer. However, when cells were plated onplastic, the rate of cell attachment was much lower and the proliferative rate of HMC grown onplastic also was strikingly lower (exponential doubling time 4-3 days) than that of cells grown onECM (exponential doubling time 2-4 days). Cells upon reaching confluency on plastic weremarkedly enlarged as compared to confluent cells grown on ECM. These observations corroborateddifferences in final cell density where it was noted that HMC cultured on ECM demonstrated a 10-fold greater final cell density as compared to cells grown on plastic. Results from these studiesillustrate the fact that phenotypic expression as well as proliferative responsiveness of HMC can bemodulated by adhesive interactions with preformed ECM.

INTRODUCTION

The serous membranes lining the major body cavities consist of a single layer ofhighly flattened mesothelial cells resting upon an underlying subcellular basementmembrane. Mesothelial cells provide a non-adhesive surface that facilitates themovement of visceral organs (Andrews & Porter, 1973) and acts as a permeabilitybarrier, which regulates fluid and solute transport between the systemic circulationand body cavities (Cunningham, 1926; Courtice & Simmonds, 1954). The basementmembrane separates mesothelial cells from the underlying supportive tissue andfunctions as an extracellular scaffold that maintains orderly tissue structure (Vracko,1974). The composition of the basement membrane present beneath mesothelial cellsin vivo has not been characterized. However, on the basis of studies with basementmembranes from a variety of cell types (Hay, 1981), it is most probably composed ofcollagens, glycosaminoglycans, glycoproteins and proteoglycans arranged in a bio-chemically complex and intricately arranged three-dimensional structure.

* Author for correspondence.

Key words: human mesothelial cells, extracellular matrix, adhesion.

134 M.jf. Niedbala, K. Crickard and R. jf. Bernacki

At present, few in vitro studies have been undertaken on isolated humanmesothelial cells (HMC). The long-term maintenance of these cells is hampered bythe inability to culture human mesothelial cells (HMC) easily on plastic on a routinebasis (Castor & Naylor, 1969; Singh et al. 1978). Limited growth of these cells hasbeen achieved by using specific growth factors (Connell & Rheinwald, 1983),conditioned medium and fibroblast feeder layers (Wu et al. 1982), as well as specificattachment factors (Lechner et al. 1985).

Cell-substratum interactions have been reported to be important for cell growthand differentiation of both normal and malignant epithelial cells (Hay, 1981;Vlodavsky et al. 1980). Such cells generally attach, grow and maintain theirdifferentiated functions better when extracellular matrix (ECM) or isolated ECMcomponents are used as the substratum rather than plastic (Wilde et al. 1984; Jozanet al. 1982). The relevance of cell-substratum interactions in HMC growth andmorphology has not been fully ascertained. This paper demonstrates the effect ofECM-coated culture dishes, produced by bovine corneal endothelial cells (BCEC),on the cellular adhesion, morphology and growth of isolated HMC. It has beendemonstrated that the ECM produced by BCEC is composed of interstitial collagentypes I—III and basement membrane collagen types IV and V (Tseng et al. 1981,1983). Also present are proteoglycans comprised primarily of heparan sulphate anddermatan sulphate (Robinson & Gospodarowicz, 1983; Nevoef al. 1984), as well asglycoproteins such as elastin, fibronectin and basement-membrane-specific laminin(Gospodarowicz et al. 1981). It has also been shown by Gospodarowicz et al. (1977)that the ECM produced by BCEC in vitro is analogous to the ECM associated withthese cells in vivo.

HMC cultured on ECM may provide a relevant model that may help to elucidatefurther the mechanisms by which human tumour cells interact with the mesotheliumto form metastatic implants (Niedbala et al. 1985). In addition, mesothelial cellscultured in this manner may aid in investigating the growth, repair and trans-formation of the mesothelium.

MATERIALS AND METHODS

Isolation of human mesothelial cellsHuman mesothelial cells were derived from peritoneal effusions of an ovarian cancer patient,

since it is well known that ascites fluid is an excellent source of large quantities of viable HMC(Cunningham, 1926; Domagala & Koss, 1979). Ascitic fluid containing a high percentage ofmesothelial cells, as determined by cytological analysis of cytospin preparations, was centrifuged(500£) for lOmin. The cell pellet was resuspended in Roswell Park Memorial Institute (RPMI)1640 culture medium supplemented with 20% foetal calf serum (FCS), fungizone (2-5/igmP1),and gentamycin (SO^gmF1) (Gibco, Grand Island, NY). Cells were plated onto 10cm plasticculture dishes (Falcon, Becton, Dickinson and Co., Cockeyville, MD) for 24h at 37°C (95 % air/5 % CO2). During this time interval a high percentage of HMC adhered to the plastic, while themajority of tumour cells and other inflammatory cell types in general did not become adherent.Tumour cells and other cell types were easily removed from the adherent HMC by gentle washingwith warm medium (37 °C). HMC were subsequently dissociated from the plastic substrate by abrief exposure (2-3 min) to 0-5 % trypsin and 0-02% EDTA at 37°C. Cells were washed, pelletedby centrifugation, replated onto plastic culture dishes and incubated for 6h at 37°C. Adherent

Mesothelial cell culture on extracellular matrix 135

HMC were then maintained as described by Wu et al. (1982) with the following modifications: a1:1 (v/v) mixture of Dulbecco's modified Eagle's medium (DMEM) supplemented with 20 %FCS, hydrocortisone (0"4/igml~') and cell-free medium, that had been conditioned by 3T3 mousefibroblasts in exponential growth for 48-72 h, was used. Under these selective culture conditions itwas observed, by phase-contrast microscopy, that HMC grew progressively as primary cultures.

Preparation of ECM-coated culture dishesPrimary cultures of bovine corneal endothelial cells were established from calf eyes and were

maintained in culture as described by Gospodarowicz et al. (1977). After 6-8 days, post-confluentBCEC monolayers were solubilized by washing cultures with sterile distilled water and brieflyexposing cells to 0 - 0 2 M - N H 4 O H . This resulted in the lysis of BCEC and exposure of theunderlying, structurally intact ECM, which was firmly and uniformly attached to the culturedishes. ECM-coated culture dishes were carefully washed with PBS and used immediately, orstored in PBS containing antibiotics at 4°C for no longer than 2 weeks.

Adhesion assay

Serially passaged human mesothelial cells (P2) at a concentration of 2xlO6 to SxlO6cellsml~'were prelabelled with 0-25 mCi of 51Cr-labelled sodium chromate as described previously(Crickard et al. 1983; Niedbala et al. 1985). Radiolabelled HMC were added to either 16mmECM-coated or plastic culture dishes. At selected times non-adherent cells were removed byaspiration and the cultures were carefully washed three times with warm PBS (37°C). Theremaining, firmly adhering cells were released by trypsinization and cell-associated radioactivitywas measured using a gamma counter (Searle model 1185, TM Analytic Inc., Elk Grove Village,IL).

Karyotyping of cellsHMC grown on ECM-coated culture dishes were exposed to colcemid (lO^gml"1) and

incubated for 8h at 37°C (95% air /5% CO2). Cells were dissociated from culture dishes bytrypsinization, washed and pelleted. Cells were processed for karyotypic analysis, as described byCrickard et al. (1983).

Cytological analysisCytology slides were prepared using a Cytospin centrifuge (Shandon Southern Instruments,

Inc., Sewichley, PA) and immediately fixed in 70% ethanol for staining with haematoxylin andeosin.

Cell proliferationExponentially grown HMC maintained under previously defined conditions, were harvested by

trypsinization, washed and pelleted. Cells were plated at an initial density of 1X104 cells/35 mmculture dishes with or without an ECM. Triplicate cultures were trypsinized and cell number wasdetermined electronically with a Coulter Counter (Coulter Electronics Inc., Hialeah, FL) everyother day. The morphological appearance of these cultures was observed by phase-contrastmicroscopy.

DNA-flow cytometryCell suspensions containing 2 x l 0 5 cells ml"1 in RPMI 1640 were stained with a DNA

fluorochrome, 4'-6-diamidino-2-phenylindole (1 ̂ gml"1), in the presence of 0-2% Triton X-100(Taylor, 1980). Human blood leucocytes were used as a diploid standard. The DNA ploidy indexwas calculated as the ratio of the G\ peak channel of cell suspension to that of the internal diploidcell standard. Flow cytometry was carried out using a ICP 22A flow cytometer (Ortho Instru-ments, Westwood, MA) as described by Frankfurt (1983).

136 M. jf. Niedbala, K. Crickard and R. J. Bernacki

Indirect immunofluorescence studiesCells were cultured on glass coverslips in the presence and absence of ECM, fixed in absolute

methanol at — 20°C for 5 min and then in acetone for 5 min. After fixation, the cells were stained forindirect immunofluorescence, as described by Asch etal. (1981). Experiments were undertakenwith a mouse monoclonal human anti-keratin (AE-3), described by Woodcock-Mitchell etal.(1982), and goat antisera raised against mouse mammary fibroblast vimentin, recently shown to becross-reactive with human fibroblast vimentin (Asch & Asch, 1986); fluorescein-conjugated sheepanti-mouse and rabbit anti-goat immunoglobulin (IgG) were used as secondary antibodies.Coverslips were mounted with phosphate-buffered glycerol (pH 8-6) containing />-phenylenedia-mine. Cells were examined for indirect immunofluorescence with a Zeiss epifluorescence photo-microscope II and a 40x plan Neofluor objective with oil immersion. Photomicrographs weremade using Kodak Tri-X film exposed at ASA1200.

RESULTS

Characterization of mesothelial cell primary cultures

Under the selective culture conditions used in this study, a relatively homogeneousmesothelial cell population was isolated as characterized, according to the followingcriteria.

Cytological analysis. Cytospin preparations of dispersed mesothelial cells werestained with haematoxylin and eosin for cytological analysis. In all cases, typicalcytological features of normal cells were noted and it was concluded from cellpreparations that a single population of proliferating cells was present. Occasionally,multinucleated reactive mesothelial cells were observed but the percentage of thesecells in primary cultures was limited to less than 5 %.

Flow cytometry and karyotypic analysis. A diploid population of cells constitutedthe majority of the proliferating cell population as determined by cytogenetic analysis(Fig. IB). A modal chromosome number of 46 with a modal range of 42-48,equivalent to 85 % of the total proliferating cell population, was observed. DNA flowcytometry indicated the presence of a single population of diploid cells using humanleucocytes as an internal diploid standard (Fig. 1A). A total of 30000 to 50000 cellswas analysed for each histogram and a ploidy index equivalent to unity was observed,corroborating results from cytogenetic studies.

Immunofluorescence microscopy. HMC, cultured to a semiconfluent state oneither glass or ECM-coated glass coverslips, were found to contain large amounts ofkeratin organized in a dense, filamentous network, as defined by reaction with keratinantibodies. All of the cells reacted positively, regardless of the substrates used(Fig. 2). Keratin filaments were distributed throughout the cells despite changes incell shape and size, which were dependent on the growth substrate. Cells showedbundles of keratin filaments, which were organized concentrically in perinuclearregions (Fig. 2C,D). The positive reactivity of these cells to a variety of keratin-specific antibodies demonstrates their epithelial nature. In contrast, human meso-thelial cells when cultured on plastic or ECM-coated culture dishes expressedrelatively low levels of immunoreactivity to goat antiserum directed against vimentin(Fig. 3A,B).

Mesothelial cell culture on extracellular matrix 137

In addition to the studies carried out above we have used a monoclonal antibody(F36/22) that recognizes an antigen determinant expressed by ovarian carcinomacells but not by mesothelial cells (Croghan et al. 1984). This monoclonal antibodywas used to screen primary mesothelial cell cultures as well as subsequent cultures,periodically, for the presence of tumour cells. The results were compared with thosefrom studies of the reactivity of cells to an irrelevant control monoclonal antibody ofthe same subclass using indirect immunoperoxidase staining. In all cases cultures

0-8'

0-7

0-6

I 0-5

8 0-4

•5 0-3

0-2

01

11

10

9

8

53 7

I 6

* 5

3 4321

(a)

A

(b)

JK.In n

Fluorescence intensity

(3) (3)

n. .n38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

Chromosome number

Fig. 1. DNA flow cytometric and karyotypic analysis of primary cultures of humanmesothelial cells. Cells grown on plastic culture dishes were processed for DNA flowcytometry (A) or karyotypic analysis (B) as described in Materials and Methods. Cellsuspensions for flow cytometry were prepared by trypsinization and assayed (a) withoutor (b) with human leucocytes as an internal diploid standard. Karyotypic analysis showsthe number of chromosomes per metaphase plate in histogram form. Filled bars indicatethe modal range while numbers in parentheses indicate percentages.

138 M. J. Niedbala, K. Crickard and R. J. Bernacki

Fig. 2. Indirect immunofluorescent staining patterns of human mesothelial cells culturedon glass (A,C) and ECM-coated glass coverslips (B,D). Cells were stained with humananti-keratin mouse monoclonal AE-3 as described in Materials and Methods. Bars: A,B,

i; C,D, 3jUm.

the same subclass using indirect immunoperoxidase staining. In all cases cultureswere consistently negative for the expression of F36/22 antigen determinant withstaining levels similar to that of irrelevant controls. The sensitivity of this probe forovarian carcinoma cells has been recently described (Niedbala et al. 1985) in thepresence of mixed cultures of mesothelial cells grown on ECM following the additionof ovarian tumour cells. In these studies single ovarian tumour cells could be readilyidentified by the intense immunoreaction product associated with tumour cells.Finally, mesothelial cells when injected subcutaneously into nude mice were notobserved to be tumorigenic with initial inocula up to SxlO7 cells per animal,following 5 months of observation.

Rate of cell attachment

In order to quantify the attachment of HMC to ECM versus plastic, HMC wereprelabelled with 51Cr and seeded onto these respective substrata (Fig. 4). Time-dependent attachment to these substrata was determined by measuring the radio-activity associated with adherent HMC as a function of total cellular radioactivity

Mesothelial cell culture on extracellular matrix

Fig. 3. Indirect immunofluorescence staining pattern of human mesothelial cellscultured on ECM-coated glass coverslips (A) or on glass (B). Cells were stained with goatantiserum directed against vimentin as described in Materials and Methods. Bars, 10 (im.

140 M.jf. Niedbala, K. Crickard andR.J. Bernacki

1001—

50

u

I „ I0-5 1

Time (h)1-5

Fig. 4. Rate of attachment of mesothelial cells to plastic or ECM-coated culture dishes.Mesothelial cells were prelabelled with NaslCr04 and seeded onto duplicate 16 mm platesin the presence of ECM ( • • ) or plastic culture dishes (O O) at a density of5 x l 0 4 c e l l s m r ' . Non-adherent cells were removed by aspiration while firmly adherentcells were released by trypsinization. Radioactivity was measured as described inMaterials and Methods. The data are expressed as the mean±s.D. of two separateexperiments.

(adherent and non-adherent fractions). At early times HMC adhered to ECM morerapidly and more firmly than to plastic. Greater than 80% of the seeded HMCattached to ECM after 15 min. In contrast, less than 25 % of HMC attached to plasticwithin 15 min. The amount of time required for the firm attachment of 50% ofmesothelial cells was defined as the adhesion index for 50% of cells (AI50). Forplastic an AI50 of 28 min was observed whereas for ECM it was markedly faster withan AI50 of 9 min. At later times (>6 h) the percentage of cells that became firmlyadherent to plastic approached that of cells exposed to ECM (Fig. 4).

Effect of substratum on cell morphology and culture organization

Marked differences in the morphological appearance of HMC grown on plastic-and ECM-coated culture dishes were observed using scanning electron microscopy(Fig. 5). HMC cultured on plastic became considerably enlarged and a high per-centage (25%) was observed to be multinucleated within 72 h. In addition, HMCcultured on plastic showed a greater degree of cellular overlap in comparison to cellsgrown on ECM. In contrast, HMC plated on ECM adopted a highly flattenedcellular morphology composed of tightly packed non-overlapping cells, whichcovered the entire surface of the culture dish. In some cases, complex intercellularjunctional complexes mediated via interdigitating cell processes were observed whenHMC were cultured on ECM. Cells could be maintained on either substratum for aslong as several months following confluency with minimal alterations in theirrespective cell morphologies and viability, as determined by their subsequentproliferative capacity. Interestingly, upon dissociation of cells grown on plastic and

Fig. 5. Scanning electron micrographs of confluent mesothelial cells maintained for 10days on plastic or ECM-coated culture dishes. Cells maintained on plastic (A) wereconsiderably enlarged and in some cases, binucleate (arrow). While cells maintained onECM (B) were more organized and formed a well-defined monolayer. Complex inter-cellular junctions were observed and seemed to be mediated by cell processes (arrows).Numerous surface blebs and centrally located microvilli were present regardless ofsubstrate. Bar, 10 jum.

142 M. J. Niedbala, K. Crickard and R. y. Bernacki

reseeding on ECM, HMC immediately adopted the morphological characteristicsdescribed above for HMC maintained originally on ECM.

Cell proliferation studies

Comparative studies were undertaken to determine the growth rate of HMC platedat a low density on plastic versus that of ECM-coated culture dishes (Fig. 6). Cellsmaintained on ECM divided more rapidly with a mean exponential doubling time of2-4 days. Cells grown on plastic did proliferate under the conditions used but at asignificantly slower rate (mean doubling time 4-3 days). Differences in final celldensity were also observed between the two substrata after 10 days of growth. HMCgrown on ECM had a final cell density 10-fold higher than that of cells grown onplastic. These results were consistent with the morphological appearance of con-fluent cultures maintained on plastic versus ECM (Fig. 5). Cells grown on ECMwere composed of tightly packed closely apposed cells, whereas those grown onplastic were strikingly larger and exhibited considerably more cell overlap.

DISCUSSION

The role of cell—substratum interactions in regulating cell proliferation andmorphogenesis both in vitro and in vivo has been demonstrated in a number of

106

| l O 5

10"4 6Time (days)

10

Fig. 6. Proliferation of HMC when maintained on plastic versus ECM-coated culturedishes. Mesothelial cells were plated at 1X104 cells/35 mm culture dish in the presence( • • ) or absence (O O) of ECM. Triplicate cultures were counted every otherday following release of cells by trypsinization. Mean ± s.D.

Mesothelial cell culture on extracellular matrix 143

laboratories (Hay, 1981; Kleinmanei al. 1981; Vracko, 1974). Mesothelial cells havebeen shown to rest upon a basement membrane in vivo (Odor, 1954). Cellularcontact with the basement membrane may directly modulate the morphogenesis andproliferation of the mesothelium. Likewise, in order to proliferate and express theirnormal phenotype, mesothelial cells maintained in tissue culture may require notonly nutrients and growth factors, but also an appropriate physical substratum towhich they can attach and then proliferate.

This study takes advantage of the fact that bovine corneal endothelial cells(BCEC) retain their ability to secrete basally an extensive ECM in vitro. This ECM,although bovine in origin, has been shown to conform to the chemical criteria(Kefalides, 1980) and biological criterion (Kleinman et al. 1981) of a basementmembrane, in general, and serves as a substratum to which cells can attach, and onwhich they can migrate and divide both in vitro and in vivo. BCEC ECM has beenfound in this study to support HMC growth in vitro and to serve as a substitute forthe basement membrane underlying HMCin vivo. In addition, HMC may modulatethe preformed BCEC ECM by producing their own collagens and mucopoly-saccharides, as described for other diploid cell types (Madri et al. 1983; Kato &Gospodarowicz, 1985), and thus become capable of modulating their own growthand morphology by the production of a hybrid matrix. Preliminary studies indicatethat the HMC used in this investigation are capable of depositing extracellular matrixcomponents onto plastic. These studies were carried out using monospecific anti-bodies directed against basement-membrane-specific laminin and collagen type IV inconjunction with immunofluorescence (unpublished data). These observations cor-roborate the findings of Rennard et al. (1984), which demonstrated that rat pleuralmesothelial cells can produce matrix components in vitro.

In this study it was observed that HMC attached rapidly to ECM as compared toplastic. Cellular attachment to ECM involves a complex series of interactions with avariety of adhesive and polyionic structural macromolecules present in the ECM(Yamada, 1983; Kefalides, 1980). Interestingly, the ability of HMC to attach toplastic approached that of ECM at later times, indicating that perhaps adherentHMC can condition the plastic culture surface, making it more favourable tosupporting further attachment.

Close contact of HMC with ECM may increase HMC responsiveness to growthfactors present in the culture medium, which are not observed when cells aremaintained on plastic (Gospodarowicz et al. 1980). This could explain the obser-vation that HMC plated at a low density on ECM proliferated more rapidly andattained a higher final cell density as compared to cells grown on plastic. The resultsdescribed here also indicate that mesothelial cultures maintained on plastic andpassaged at a low cell density rapidly lose their normal morphological phenotype(Baradi & Hope, 1964), while cultures maintained on ECM retain it. The ECM maytherefore stabilize normal phenotypic expression of HMC, since a simple change ofsubstrate from plastic to ECM may restore the sensitivity of cells to factors present inthe culture medium, resulting in enhanced proliferation and a striking change incellular morphology. Similar results have been observed for a number of other cell

144 M. jf. Niedbala, K. Crickard andR.J. Bernacki

types (Gospodarowicz & Tauber, 1980; Betheae£ al. 1982; Kato & Gospodarowicz,1985).

The mechanism by which the ECM exerts its 'permissive' effect on HMCproliferation is still a subject for hypothesis. Gospodarowicz et al. (1978) havespeculated that cell shape may be instrumental in making cells responsive to serumfactors and mitogens to which they cannot respond unless they adopt an appropriateshape. Alternatively, the ECM may contain small amounts of growth factors andother mitogens, which may be sequestered within its macromolecular structure,thereby resulting in enhanced cellular proliferation as compared to other substrates.Recently, Gospodarowicz et al. (1983) attempted to address this question byexposing the ECM to a variety of conditions that are known to destroy the activity ofa number of mitogens. In short, the growth-promoting effect of the treated ECM wasequivalent to that of untreated ECM, suggesting that endogenous growth factors maynot be responsible for its growth-promoting effect and that other mechanisms relatedto its structural composition may be involved.

Recent studies by Connell & Rheinwald (1983) have indicated that EGF whenadded to HMC in the presence of serum resulted in the induction of cellularproliferation and in the acquisition of a fibroblastoid morphology, which wasreversible upon removal of EGF. Interestingly, HMC when cultured in the presenceof EGF were observed to have increased amounts of vimentin and decreased absoluteamounts of cytokeratins. In our studies HMC adopted an epithelial cellular mor-phology, with high levels of cytokeratins and low levels of vimentin being observedwhen ECM was used as a growth substrate despite an increase in cellular proliferativecapacity. Our observations are in agreement with in situ observations in which HMCwere shown to possess epithelial morphology, high keratin and low vimentin content(LaRocca & Rheinwald, 1984).

Preliminary studies on EGF treatment of HMC over a wide range of concen-trations (0-5—50ngml~') have failed to demonstrate considerable morphologicaland/or proliferative differences when HMC are maintained on plastic versus ECM.For this reason it is believed, on the basis of our results, that two independentmechanisms may account for differences in HMC proliferative rates, one beingmediated by the growth substrate and another via soluble factors.

The results presented in this study demonstrate the importance of humanmesothelial cell—extracellular matrix interactions in maintaining normal cell mor-phology, behaviour and proliferative responsiveness in vitro. In addition, we havebeen successful in isolating four other mesothelial cell lines from peritoneal effusionsof ovarian cancer patients. In all cases the observations were similar to thosedescribed here. Thus, the culture system described in this paper may represent amodel system in which to investigate the mechanisms by which human tumour cellsinteract with the mesothelium to form metastatic foci (Niedbala et al. 1985). It alsomay permit the study of growth, repair and transformation of the mesotheliumin vitro.

Mesothelial cell culture on extracellular matrix 145

The authors thank Dr E. Nava for her assistance with cytological analyses, Dr B. Asch for helpwith immunofluorescence studies, and Dr O. Frankfurt for performing flow cytometry. This workwas supported in part by grants from the National Cancer Institute (USA), CA-13038, CA-42898,CA-24538 and CA-32767.

REFERENCES

ANDREWS, P. M. & PORTER, K. R. (1973). The ultrastructural morphology and possible functionalsignificance of mesothelial microvilli. Anat. Rec. Yll, 409-426.

ASCH, B. B., BURNSTEIN, N. A., VIDRICH, A. & SUN T.-T. (1981). Identification of mousemammary epithelial cells by immunofluorescence with rabbit and guinea pig antikeratin antisera.Proc. natn.Acad. Sci. U.S.A. 78, 5643-5647.

ASCH, B. B. & ASCH, H. L. (1986). Structural components as markers of differentiation andneoplastic progression in mammary epithelium. In Cellular and Molecular Biology of MammaryCancer (ed. D. Medina, W. Kidwell, G. Heppner & E. Anderson). New York: Plenum (inpress).

BARADI, A. F. & HOPE, J. (1964). Observations on ultrastructure of rabbit mesothelium. Expl CellRes. 34, 33-44.

BETHEA, C. L., RAMSDELL, J. S., JAFFE, R. S., WILSON, C. B. & WEINER, R. I. (1982).

Characterization of dopaminergic regulation of human prolactin secreting cells cultured onextracellular matrix. J. din. Endocr. Metab. 54, 893-902.

CASTOR, C. W. & NAYLOR, B. (1969). Characteristics of normal and malignant human mesothelialcells studied in vitro. Lab. Invest. 20, 437-443.

CONNELL, N. D. & RHEINWALD, J. G. (1983). Regulation of the cytoskeleton in mesothelial cells:reversible loss of keratin and increase in vimentin during rapid growth in culture. Cell 34,245-253.

COURTICE, F. C. & SIMMONDS, W. J. (1954). Physiological significance of lymph drainage of theserous cavities and lungs. Physiol. Rev. 34, 419-448.

CRICKARD, K., CRICKARD, U. & YOONESSI, M. (1983). Human ovarian carcinoma cells maintainedon extracellular matrix versus plastic. Cancer Res. 43, Tn&L—TKfl.

CROGHAN, G. A., WINGATE, M. B., GAMARRA, M., JOHNSON, E., CHU, T. M., ALLEN, H.,

VALENZUELA, L., TSUKADA, Y. & PAPSIDERO, L. D. (1984). Reactivity of monoclonal antibodyF36/22 with human ovarian adenocarcinomas. Cancer Res. 44, 1954-1962.

CUNNINGHAM, R. S. (1926). The physiology of the serous membranes. Physiol. Rev. 6, 242-280.DOMAGALA, W. & Koss, L. G. (1979). Surface configuration of mesothelial cells in effusions.

A comparative light microscopic and scanning electron microscopic study. Virchows Arch. cell.Path. 30, 231-243.

FRANKFURT, O. S. (1983). Increased uptake of vital dye Hoechst 33342 during S phase insynchronized HeLa S3 cells. Cytometry 4, 216-221.

GOSPODAROWICZ, D., DELGADO, D. & VLODOVSKY, I. (1980). Permissive effect of the extracellularmatrix on cell proliferation in vitro. Proc. natn. Acad. Sci. U.SA. 77, 4094-4098.

GOSPODAROWICZ, D., GONZALEZ, R. & FUJII, D. K. (1983). Are factors originating from serum,plasma, or cultured cells involved in the growth promoting effect of the extracellular matrix •produced by bovine corneal endothelial cells. J. cell. Physiol. 114, 191-202.

GOSPODAROWICZ, D., GREENBURG, G. &BIRDWELL, C. R. (1978). Determination of cellular shapeby the extracellular matrix and its correlation with the control of cellular growth. Cancer Res. 38,4155-4171.

GOSPODAROWICZ, D., GREENBURG, G., FOIDART, J. M. & SAVION, N. (1981). The production and

localization of laminin in cultured vascular and corneal endothelial cells. J. cell. Physiol. 107,173-183.

GOSPODAROWICZ, D., MESCHER, A. L. & BIRDWELL, C. R. (1977). Stimulation of corneal endo-thelial cells proliferation in vitro by fibroblast and epidermal growth factors. Expl Eye Res. 25,75-89.

146 M. J. Niedbala, K. Crickard and R. J. Bernacki

GOSPODAROWICZ, D. & TAUBER, J. P. (1980). Growth factors and extracellular matrix. Endocr.Rev. 1, 201-227.

HAY, E. D. (1981). Extracellular matrix. J . CellBiol. 91, 205-223.JOZAN, S., TOURNIER, J. F., TAUBER, J. P. & BAYARD, F. (1982). Adaptation of human breast

cancer cell line MCF-7 to serum free medium culture on extracellular matrix. Biochem. biophys.Res. Commun. 107, 1566-1570.

KATO, Y. & GOSPODAROWICZ, D. (1985). Effect of exogenous extracellular matrices on proteo-glycan synthesis by cultured rabbit costal chondrocytes. jf. Cell Biol. 100, 486-495.

KEFALIDES, N. A. (1980). Chemistry of basement membranes: structure and biosynthesis. Adv.Microcirc. 9, 295-322.

KLEINMAN, H. K., KLEBE, R. J. & MARTIN, G. R. (1981). Role of collagenous matrices in theadhesion and growth of cells. J . Cell Biol. 88, 473-485.

LAROCCA, P. J. & RHEINWALD, J. G. (1984). Coexpression of simple epithelial keratins andvimentin by human mesothelium and mesothelioma in vivo and in culture. Cancer Res. 44,2991-2999.

LECHNER, J. F., TOKIWA, T., LAVECK, M., BENEDICT, W. F., BANKS-SCHLEGEL, S., YEAGER, H.,

BANERJEE, A. & HARRIS, C. (1985). Asbestos-associated chromosomal changes in human meso-thelial cells. Proc. natn. Acad. Sci. U.SA. 82, 3884-3888.

MADRI, J. A. & WILLIAMS, S. K. (1983). Capillary endothelial cell cultures: phenotypic modul-ation by matrix components.^. Cell Biol. 97, 153-165.

NEVO, Z., GONZALEZ, R. & GOSPODAROWICZ, D. (1984). Extracellular matrix proteoglycansproduced by cultured bovine corneal endothelial cells. Connect. Tiss. Res. 13, 45-57.

NIEDBALA, M. J., CRICKARD, K. & BERNACKI, R. J. (1985). Interactions of human ovarian tumorcells with human mesothelial cells grown on extracellular matrix: an in vitro model system forstudying tumor cell adhesion and invasion. Expl Cell Res. 160, 499-513.

ODOR, D. L. (1954). Observation of the rat mesothelium with the electron and phase microscopes.Am.J.Anat. 95,433-453.

RENNARD, S. I., JOURAND, M.-C, BIGNON, J., KAWANAMI, O., FERRANS, V. J., DAVIDSON, J. &

CRYSTAL, R. G. (1984). Role of pleural mesothelial cells in the production of the submesothelialconnective tissue matrix of lung. Am. Rev. resp. Dis. 130, 267-274.

ROBINSON, J. & GOSPODAROWICZ, D. (1983). Glycosaminoglycans synthesized by cultured bovinecorneal endothelial cells. J . cell. Physiol. 117, 368-376.

SINGH, G., DEKKER, A. & LADOULIS, C. T. (1978). Tissue culture of cells in serous effusionsevaluation as an adjunct to cytology. Ada Cytol. 22, 487-489.

TAYLOR, I. W. (1980). A rapid single-step staining technique for DNA analysis by micro-fluorimetry. J. Histochem. Cytochem. 28, 1021-1024.

TSENG, S. C. G., SAVION, N., GOSPODAROWICZ, D. & STERN, R. (1981). Characterization

of collagens synthesized by cultured bovine corneal endothelial cells, j . biol. Chem. 256,3361-3365.

TSENG, S. C. G., SAVION, N., GOSPODAROWICZ, D. & STERN, R. (1983). Modulation of collagensynthesis by a growth factor and by the extracellular matrix: comparison of cellular response totwo different stimuli. J . Cell Biol. 97, 803-809.

WILDE, C. J., HASAN, H. R. & MAYER, R. J. (1984). Comparison of collagen gels and mammaryextracellular matrix as substrata for study of terminal differentiation in rabbit mammary epi-thelial cells. Expl Cell Res. 151, 519-532.

WOODCOCK-MITCHELL, J., EICHNER, R., NELSON, W. G. & SUN, T. T. (1982). Immuno-localization of keratin polypeptides in human epidermis using monoclonal antibodies. J. CellBiol. 95, 580-588.

Wu, Y. J., PARKER, L. M., BINDER, N. E., BECKETT, M. A., SINARD, J. H., GRIFFITHS, C. T. &

RHEINWALD, J. G. (1982). The mesothelial keratins: a family of cytoskeletal proteins identifiedin cultured mesothelial cells and nonkeratinized epithelia. Cell 31, 693-703.

VLODAVSKY, I., Lui, G. M. & GOSPODAROWICZ, D. (1980). Morphological appearance, growthbehavior and migratory activity of human tumor cells maintained on extracellular matrix versusplastic. Cell 19, 607-616.

Mesothelial cell culture on extracellular matrix 147

VRACKO, R. (1974). Basal lamina scaffold-anatomy and significance for maintenance of orderlytissue structure. Am. J. Path. 77, 314-346.

YAMADA, K. M. (1983). Cell surface interactions with extracellular materials. A. Rev. Biochem. 52,761-799.

{Received 16 April 1986 -Accepted 30 May 1986)