Expression and Function of β1 and β3 Integrins of Human Mesothelial Cells in vitro

Experimental and Molecular Pathology 66, 131–139 (1999) Article ID exmp.1999.2252, available online at http://www.idealibrary.com on

Expression and Function of b1 and b3 Integrins of Human Mesothelial Cells in vitro

Lothar Tietze, Jana Borntra¨eger, Bernd Klosterhalfen, Baffour Amo-Takyi, Stefan Handt, Kalle Gu¨nther, and Sabine Merkelbach-Bruse Institute of Pathology, Aachen University of Technology, Aachen, Germany

Received January 8, 1999, accepted February 7, 1999

Mesothelial cells (MC) and extracellular matrix (ECM) components are thought to play a pivotal regulatory role during the inflammatory– reparative response of serosal membranes. Integrins are known to serve as cellular ECM receptors, but mesothelial integrin expression and its function, particularly its role for attachment to different ECM components, remain to be elucidated. The aim of the present study was to characterize the integrin expression of human omentum majus derived MC (HOMC) in vitro by immunohistochemistry and to investigate their functional significance with regard to HOMC adhesion to fibronectin (fn), vitronectin (vn), collagen IV (coll IV), and laminin (ln). Mesothelial cells in vitro strongly expressed b1, b3, a2, a3, a5, and av chains. A weak reactivity was found for a1 and a6, but no a4 reactivity was detectable. Compared to the control, fn, vn, coll IV, and ln caused a significant 2.6-, 2.2-, 2-, and 1.6-fold increase of HOMC adhesion, respectively. Inhibition studies revealed that HOMC attachment to fn is mediated by a5b1, avb1, and avb3, with a synergistic effect of a5b1 and avb3. Adhesion to vn is mediated by avb1 and avb3. Integrins a1b1, a2b1, and a3b1 mediate adhesion to coll IV and ln. We suggest that the integrin expression and function of mesothelial cells described here play an important role in the interaction of MC with the ECM, particularly during the acute and chronic inflammatory–reparative response of serosal membranes. q 1999 Academic Press Key Words: Mesothelium, peritoneum, integrins, extracellular matrix.

actively participate in the pathogenesis of diseases of serosal membranes, particularly peritonitis, fibrous adhesions, and metastasis (1–5). Generally, the inflammatory–reparative response of serosal membranes caused by different stimuli is a complex, interactive process involving mesothelial cells (MC), inflammatory cells, humoral factors, and extracellular matrix (ECM) components. Cell–cell and cell–ECM interactions are in part mediated by a group of ab heterodimeric glycoproteins, known as integrins. There are at least 8 known b subunits and 14 a integrin subunits. Many a subunits can associate with only a single b subunit, so subfamilies with shared b subunits may be defined. Each integrin displays a unique and often cell-specific binding pattern to ECM molecules. In addition to their role as adhesion receptors, integrins function as signaling receptors and have been shown to modulate cellular migration, survival, growth, and differentiation (6–9). In the specific context of the inflammatory–reparative tissue response of serosal membranes, cell–cell interactions of mesothelial cells with inflammatory cells have been characterized in part. Initially mesothelial cells are stimulated by activated peritoneal macrophages to liberate a variety of inflammatory mediators (PGE2, PGI2, IL-1, IL-6, IL-8, TGFb, and MCP-1) and also to express cell–cell adhesion molecules, which participate in the local recruitment of macrophages and granulocytes, particularly ICAM-1 and VCAM1 (1–5,10,11).

INTRODUCTION

The mesothelium is a finely regulated cellular monolayer lining serosal cavities. There is growing evidence that mesothelial cells are not a passive semipermeable barrier, but

0014-4800/99 $30.00 Copyright q 1999 by Academic Press All rights of reproduction in any form reserved.

131

132 Acute changes of the ECM during peritonitis are characterized by deposits of a fibrin-rich exudate, including plasmaderived fibronectin and vitronectin, which may be organized into granulation tissues, leading to fibrous adhesions (12,13). Repetitive exposure to irritating agents, particularly dialysis solutions, may lead to peritoneal fibrosis and failure of the peritoneal membrane, which is morphologically characterized by subserosal accumulation of collagen fibers (14,15). It has been suggested that MC stem cells may contribute to peritoneal fibrosis by proliferation and overproduction of ECM components (15). However, cell–ECM interactions of MC have been little investigated and most studies have focused on the ability of MC to regulate plasminogen activation, a crucial factor for dissolution of the fibrin-rich exudate (16–18). MC cell behavior in vitro is regulated by different ECM components, particularly fibronectin, collagen, and laminin, which cause a RGD-dependent increase in MC attachment and proliferation (19,20). As recently demonstrated by an in vitro scratch assay, migrating mesothelial cells at the wound edge are accompanied by the de novo synthesis of several ECM proteins, particularly collagen IV, laminin, and fibronectin (21). It has been concluded that the described ECM effects are mediated by (RGD-dependent) receptors of MC, but these remain to be determined (20,21). Immunohistochemical studies revealed that MC in vivo and in vitro express b1 and b3 integrins, which are known receptors for the ECM proteins mentioned above. However, the results with regard to the expression of a subunits were not consistent and the knowledge about the functional significance of MC integrin expression is limited to the role of b1 integrins. No data are available regarding the role of b3 integrins and a subunits expressed by MC for adhesion to ECM components (22,23). Thus, the present study was designed to characterize mesothelial integrin expression in vitro by immunohistochemistry and to investigate their functional significance with regard to mesothelial adhesion to fibronectin (fn), vitronectin (vn), collagen IV (coll IV), and laminin (ln).

MATERIALS AND METHODS Materials Tissue culture reagents were purchased from Gibco (Karlsruhe, FRG). Monoclonal antibodies (mAb) against pancytokeratin, factor VIII associated antigen, CD 14, CD 68 (clone PGM 1), and the Ulex europaeus lectin came from

TIETZE ET AL.

Dako (Hamburg, FRG); mAb against cytokeratins 7, 8, 18, and 19 were from Sigma (Deisenhofen, Germany). Monoclonal antibodies for integrin immunostaining and inhibition experiments came from Dianova (Hamburg, FRG). The clones used for immunohistochemistry were as follows: CD 29 (b1), clone K 20; CD 61(b3), clone SZ 21; CD 49 a (a1), clone HP2B6; CD 49 b (a2), clone Gi9; CD 49 c (a3), clone M-KID2; CD 49 d (a4), clone HP2/1; CD 49 e (a5), clone SAM 1; CD 49 f (a6), clone GoH3, and Cd 51 (av), clone AMF 7. The same mAb were used for inhibition experiments except for inhibition of b1 and av for which clone Lia 1/2 and clone 69.6.5 were used, respectively. ChromPure rabbit IgG (Dianova) was used as control antibody for inhibiting experiments. Biotinylated secondary antibodies and strepavidin-CY3 conjugate were obtained from Dianova. Purified collagen type IV from human placenta and vitronectin from human plasma were ordered from Biomol (Hamburg, FRG). Purified fibronectin from human plasma and laminin, extracted from mouse EHS–tumor matrix, were purchased from Boehringer (Mannheim, FRG). RGD peptides and EDTA came from Sigma. Human Omentum Majus Mesothelial Cell Culture (HOMC) Primary human mesothelial cells were derived from pieces of omentum majus as described previously (18). Briefly, a fresh piece of omentum majus (5 3 5 cm) obtained from patients undergoing elective abdominal surgery was transferred from the operation room in ice-cold buffer (10 mM Hepes, 140 mM NaCl, 4 mM KCl, 11 mM D-glucose, 1% human serum albumin, 100 IU penicillin, and 0.1 mg/ml streptomycin). The tissue was rinsed three times in phosphate-buffered saline (PBS, 378C). After digestion with 0.125% (w/v) trypsin in PBS at 378C for 10 min, the tissue was removed and 20 ml of RPMI 1640 containing 10% pooled human serum protein, L-glutamine, penicillin, and streptomycin was added. After centrifugation for 10 min at 200g, the cells were washed in 20 ml PBS (378C). After further centrifugation for 10 min at 200g, the cells were resuspended in 10 ml RPMI 1640 containing 10% (v/v) pooled human serum protein, L-glutamine, penicillin, and streptomycin and plated in a 0.2% gelatin-coated tissue culture flask (75 cm2, Falcon, supplied by Becton DickinsonGambil, Heidelberg, Germany). Cell cultures were incubated at 378C in a humidified atmosphere of 5% CO2 in air. Cells were passaged after treatment with 0.125% (w/v) trypsin for 2 min. The cells were pelleted after centrifugation for 10 min at 200g, suspended in 20 ml PBS, and pelleted. The cell pellet was resuspended in 30 ml complete medium and

133

MESOTHELIAL INTEGRIN EXPRESSION AND FUNCTION

seeded with a splitting ratio of 1:3. Only cells from the first or second passage were used for the experiments described below. Immunofluorescence Microscopy Immunofluoresence staining of mesothelial cells was performed on monolayers that had been grown on coverslips coated with fibronectin. The monolayers were fixed in 70% (v/v) ethanol for 30 min at 2208C, rinsed, and stored in PBS at 48C. Cells were incubated with primary antibodies diluted in PBS with 0.5% (w/v) bovine serum albumin (BSA) and 0.5% (v/v) Tween 20 for 45 min at room temperature, and rinsed three times with PBS/0.5% (w/v) BSA. After incubation with the secondary biotin-labeled antibody for 30 min at room temperature, cells were again rinsed three times with PBS/0.5% (w/v) BSA and subsequently incubated with streptavidin/Cy3. The cells were examined with a confocal laser scan microscope at a wavelength of 568 nm using a Zeiss LSM 410 equipped with an ArKr Laser. Coating with ECM Components Flat-bottom polystyrene microtiter plates (Becton Dickinson, Heidelberg, FRG) were coated for adhesion experiments according to the manufacturer’s instructions. Briefly, the purified ECM components were dissolved in PBS (each component in concentrations ranging from 10 to 1 3 1024 mg/ml), added to the wells, and incubated for 24 h at 48C, except for laminin, which was incubated at 378C for 45 min in a humidified atmosphere of 5% CO2 in air. Nonspecific protein binding of the coated wells was blocked by adding 1% (w/v) BSA (48C, 12 h) and rinsing the wells with Hepes buffer. Adhesion Assay For adhesion experiments HOMC were detached with collagenase I (15 min, 378C; Worthington, Freehold, USA), washed once in RPMI 1640, centrifuged (200g for 10 min), resuspended in RPMI 1640, and preincubated for 30 min in a humidified atmosphere of 5% CO2 in air (378C). Twenty thousand cells were seeded into each well. Quantification of adherent cells was performed using crystal violet staining according to the method described by Aumailley et al. (24). Briefly, after 15, 30, 60, and 120 min the supernatant with nonadherent cells was removed by two washes with prewarmed RPMI 1640. Attached cells were fixed with 30% (v/v) methanol/ethanol for 15 min at room temperature. Cells

were stained with 0.1% (w/v) crystal violet (Sigma, Hamburg, FRG), extensively washed with distilled water, and dried at room temperature. The dye was resuspended with 50 ml of 0.2% (v/v) Triton X-100/well and color yields were then measured in an ELISA reader at 590 nm (Titertek Multiscan Plus MKII, Flow Laboratories GmbH, Meckenheim, FRG). Optical density (OD) showed a linear relationship to the cell count between 1 3 103 and and 2 3 104 cells per well, as determined by dilution series. Negative controls with BSA/polystyrene (without cells) showed OD readings in the range between 0 and 0.07. These background values were subtracted from those obtained with the ECM components used. Inhibition of HOMC Attachment to ECM Components To determine the integrin-dependent binding of HOMC to each ECM component, inhibiting mAbs against different a and b subunits were added in concentrations ranging from 5 3 10 to 1 3 1021 mg/ml. Controls included incubations with an irrelevant control antibody in the same concentration and buffer as the relevant antibody. RGD mediated adhesion was evaluated by adding RGD peptides (0.5 mM). EDTA (5 mM) was used as positive control to block Ca21-dependent adhesion. Adhesion of HOMC to the various ECM components in the presence of inhibiting mAb was related to the adhesion rate in the presence of the control antibody (relative adhesion 5 ODx/ODcontrol). Statistical Analysis All experiments were performed at least three times in triplicate. Statistical analysis was performed using the Student t test for paired data with SPSS for MS Windows. All data are presented as mean values 6 SEM. Statistically significant differences are indicated by asterisks (*P , 0.05; **P , 0.01; ***P , 0.001).

RESULTS Characterization of HOMC The HOMC cultures were characterized with the help of phase-contrast microscopy and immunohistochemistry. The epitheloid cells obtained grew as a monolayer, reaching confluence 4 to 10 days after initial seeding, and showed the typical cobblestone-like growth pattern. The cells could be

134 passaged up to five times. However, after the third passage some cells showed a morphological change: the cell size increased and multinucleated cells appeared. Immunohistochemistry revealed a strong cytoplasmic reactivity for pancytokeratin, cytokeratin 8, cytokeratin 18, and vimentin, whereas reactions for vWF and U. europaeus were negative, an indication that no relevant contamination with endothelial cells was present. Antibodies against myelomonocytic-associated antigens were tested to determine the contamination with macrophages. Contrary to macrophage populations, HOMC did not show CD 14 and CD 68 reactivity. Immunohistochemical Integrin Pattern of HOMC Indirect immunofluorescence analysis of HOMC grown on fibronectin coated polystyrene revealed a specific and strong reactivity for mAb against b1, b3, a2, a3, a5, and av (Fig. 1). A weak reactivity was found against a1 and a6. No anti-a4 reactivity was detectable. The b1, a2, and a3 chains were expressed at a moderate intensity and diffuse pattern at the basal cell–substratum interface and accumulated in a linear pattern along the cell–cell borders. The b3, a5, and av chains were clustered in focal plaques at the basal and peripheral cell membrane. Adhesion of HOMC to Fibronectin, Vitronectin Collagen IV, and Laminin To assess the functional relevance of the mesothelial integrin pattern, we determined the ability of mesothelial cells to bind to different extracellular matrix components. We chose extracellular matrix components with broad reactivity for several integrins, particularly collagen type IV and laminin as basement membrane components, and plasma-derived vitronectin and fibronectin as proteins that are known to be important components of the extracellular matrix during wound healing (12,21). Coating with varying concentrations of each ECM component revealed a concentration-dependent effect of each ECM component on HOMC adhesion (data not shown). Maximal adhesion was found after coating with a concentration of 10 mg/ml for fibronectin, vitronectin, and laminin and of 5 mg/ ml for collagen IV. Data of relative adhesion of HOMC to different ECM components are shown in Fig. 2. Compared to the control (uncoated polystyrene), fibronectin, vitronectin, collagen IV, and laminin caused a significant 2.6-, 2.2-, 2-, and 1.6fold increase of adhesion, respectively. Extracellular matrix mediated adhesions were maximal at 15 min, except for

TIETZE ET AL.

vitronectin and laminin, which showed a maximal relative adhesion at 30 min. Extracellular matrix mediated effects on adhesion declined to a minimum at 120 min. In comparison with the control, there were no significant differences in adhesion after an incubation time exceeding 120 min.

Inhibition of HOMC Adhesion to Fibronectin, Vitronectin, Collagen IV, and Laminin To elucidate which integrins participate in adhesion found on the different ECM components, adhesion experiments in the presence of inhibiting mAb against different integrin subunits expressed by MC were performed. Combinations of mAbs are indicated by an slash (e.g., anti-a5 and antib1: a5/b1). Data of the inhibition experiments are shown in Figs. 3A–3D.

Adhesion of HOMC to Fibronectin Is Mediated by a5b1, avb1, and avb3 Adhesion to fibronectin could be significantly blocked by incubation with combinations of mAbs against the a5/b1, av/b1, and av/b3 subunits to an adhesion rate of 42, 53, and 67% (P , 0.01), respectively. Inhibition of single subunits alone (a1, a2, a3, a5, b1, and b3) and inhibition of a3/b1 did not significantly reduce adhesion to fn. In contrast, a combination of anti-a5 with anti-b3 caused the strongest inhibitory effect (40%, P , 0.001). These data and the immunohistochemical colocalization of av with a5 indicate that a5b1 and avb3 are synergistically active in cellular attachment to fibronectin. Incubation with RGD caused a 50% reduction in adhesion, demonstrating that HOMC adhesion to fn is RGD-dependent (Fig. 3A).

Adhesion of HOMC to Vitronectin Is Mediated by avb1 and avb3 Incubation with mAbs against av/b1 and av/b3, significantly reduced adhesion to vitronectin to 39 and 17%, respectively. Incubation with mAb against b1, b3 and av alone significantly inhibited adhesion to vitronectin. HOMC attachment to vitronectin is strongly RGD-dependent, as demonstrated by incubation in the presence of RGD, which blocked adhesion to vn to 5%, compared with the control (P , 0.001). These data give evidence that avb1 and avb3 expressed by HOMC serve as receptors for vitronectin (Fig. 3B).

MESOTHELIAL INTEGRIN EXPRESSION AND FUNCTION

FIG. 1. Immunohistochemical staining of HOMC grown on fibronectin-coated coverslips (bar represents 10 mm).

135

136

TIETZE ET AL.

FIG. 2. Adhesion of HOMC on fn, vn, coll IV, and ln coated polystyrene surfaces after 15, 30, 90, and 120 min. Bars represent adhesion rates of each component related to the control (relative adhesion equals ODx/ODcontrol; relative adhesion in the presence of uncoated, BSA blocked polystyrene equals 1).

a1b1, a2b1, and a3b1 Mediate Adhesion of HOMC to Collagen IV Adhesion of HOMC to collagen IV was inhibited by combinations of mAb against a1/b1, a2/b1, and a3/b1 to 14, 11, and 43%, respectively. Inhibiting antibodies against b1 alone caused a significant reduction in adhesion to 35%, whereas inhibition of the b3 chain and the a subunits caused insignificant reduction in HOMC attachment. A combination of anti-a5 and anti-b3 caused a weak but not significant reduction to 77% (Fig. 3C).

a1b1, a2b1, and a3b1 Mediate Adhesion of HOMC to Laminin Adhesion to laminin was significantly reduced by mAbs against a1/b1, a2/b1, and a3/b1 to 34, 26, and 43%, respectively. Incubation with anti-b1 alone caused a reduction to 62% compared to the control, whereas mAbs against b3 or a chains alone did not cause any significant effect. A combination of mAb against a6 and b1 did not cause an enhanced inhibitory effect compared to b1 alone (Fig. 3D).

DISCUSSION Integrins are transmembranous ab heterodimers that play a pivotal role in cell–cell and cell–substratum interactions. Integrin binding to ECM molecules is dependent on both the cell type and the matrix molecule (6,9).

In the present study, we show that mesothelial cells express a2, a3, a5, av, b1, and b3 subunits. A weak staining was observed for the a1 and a6 chains. These data qualitatively confirm recent immunohistochemical studies that describe a strong reactivity of pleural and peritoneal MC in vitro for a3, a5, and av, but a weak and focal expression of the a1, a2, and a6 chains (21,22). The b4 chain is not expressed by MC in vitro (21). Data about the integrin profile in vivo are contradictory. Barth et al. (22) found integrin expression of cultivated MC corresponding to that of activated, epitheloid MC during serositis, whereas resting mesothelial cells in vivo did not show expression of subunits a1–6. In contrast, Witz et al. (23) describe a strong reactivity for a2 and a5, but not for av on peritoneal biopsies from the pelvic wall and the uterine peritoneum. This contradiction is perhaps due to different tissue preparations and antibodies used by these investigators. The in vitro integrin profile of HOMC described here corresponds at least to the phenotype of activated MC during serositis. Further comparative studies of pleural and peritoneal biopsies under resting and inflammatory conditions are required to clarify the dynamics of MC integrin expression in vivo. HOMC adhesion to fn is partly mediated by a5b1, known as the fibronectin receptor, as well as by avb1 and avb3, known as vitronectin receptors. These data are in agreement with the described binding characteristics of these integrins (6,8). Notably, blocking the a5 and the b3 chains caused the strongest inhibitory effect on fn and vn mediated adhesion, indicating that MC avb3 and a5b1 integrins act synergistically as fn receptors. A similar synergism has been described

MESOTHELIAL INTEGRIN EXPRESSION AND FUNCTION

137

FIG. 3. Adhesion of HOMC to fibronectin (A), vitronectin (B), collagen IV (C), and laminin (D) in the presence of various inhibiting mAbs against different integrin subunits. Bars represent HOMC adhesion in the presence of the inhibiting mAbs related to the adhesion in the presence of the control antibody (relative adhesion ODx/ODcontrol; relative adhesion in the presence of the control mAb equals 1).

for M 21 melanoma cells, which coexpress a5b1 and avb3, too (25). According to the known binding specificity, avb1 and avb3 are functionally active as vitronectin receptors, and a1b1, a2b1, and a3b1 are active as receptors for collagen IV and laminin (6,8). Immunohistochemically, the a2, a3, and the b1 chains showed a reactivity at the basal cell–substratum interface and a linear accumulation along cell–cell contacts. Since it is known that a2b1 and a3b1 mediate both cell–substratum and cell–cell adhesion, we suggest that these integrins are also involved in cell–cell contacts between MC (26). MC coexpress integrins that otherwise are relatively restricted to epithelial cells (especially a2b1 and a3b1) or mesenchymal cells (especially avb3 and a5b1) (26). In our opinion this coexpression reflects the histogenesis and phenotype of MC, which are mesodermal-derived cells with an epithelial-like differentiation (i.e., close cell–cell contacts, a polar differentiation, and an expression of cytokeratins).

Our functional studies are limited to integrin mediated adhesion to different ECM components, but it is well known that integrins modulate cellular differentiation and activation in terms of proliferation, migration, gene expression, and survival. Recently it has been demonstrated that immobilized fn, collagen (type I and III), and ln increase MC proliferation, whereas the soluble form of these ECM proteins inhibited MC growth. We suggest that these ECM effects, which indicate a finely tuned MC response to changes of the ECM, are mediated by MC integrins, particularly a5b1, avb1, and avb3, which interact with fn, and a1b1, a2b1 and a3b1, which interact with collagen and laminin. Integrins are involved in the regulation of the inflammatory–reparative tissue response. One unique property of serosal wound healing is the observation that it is finished after 10 days, irrespective of the size of the initial injury (13). This observation and experimental data led to the assumption that remesotheliazation is accomplished by attachment of

138

TIETZE ET AL.

FIG. 3. Continued

free MC (in suspension) to the wound surface, by proliferating MC from the opposed peritoneal membrane, or by proliferation of MC “stem cells” in the subserosal stroma (27). Integrity of the basement membrane seems to be crucial for a restitutio ad integrum (28). Autologous transplantation of HOMC after serosal injury significantly reduces adhesion formation, indicating that attachment of free MC to injured sides may contribute to remesotheliazation (29). a1b1, a2b1, and a3b1 may be involved in adhesion of free MC to basement membrane-associated collagen IV and laminin, which are exposed by denudation of MC. It is well documented that lysis of the fibrin-rich exudate by mesothelial cells is a critical event in serosal wound healing. One may speculate that, in particular, the interaction of MC avb1 and avb3 integrins with components of this exudate, especially fibronectin and vitronectin, may modulate mesothelial differentiation in terms of proliferation, migration, and survival (30– 32). Current studies are in progress to investigate these interactions of MC with fibrin(ogen), fibronectin, and

vitronectin. In summary, our results indicate that MC in vitro strongly express a2b1, a3b1, a5b1, avb1, and avb3. We demonstrate that these receptors are in part required for MC attachment to basement membrane-associated ECM molecules (collagen IV and laminin) and to plasma-derived ECM components, particularly fibronectin and vitronectin. We suggest that the described integrins expressed by MC play an important role in the interaction of MC with the ECM, particularly during the acute and chronic inflammatory– reparative response of serosal membranes.

ACKNOWLEDGMENTS

The results were presented partly in abstract form at the 82nd Meeting of the German Society of Pathology 1998 (Pathol. Res. Pract. 194, 225). We acknowledge the excellent technical assistance of Mrs. Leonie Engels and Mrs. Kerry D. Shirazi.

139

MESOTHELIAL INTEGRIN EXPRESSION AND FUNCTION

REFERENCES 17. 1. Lanfrancone, L., Boraschi, D., Ghiara, P., Falini, B., Grignani, F., Peri, G., Mantovani, A., and Pelicci, P. G. 1992. Human peritoneal mesothelial cells produce many cytokines (granulocyte colonystimulating factor [CSF], granulocyte-monocyte-CSF, macrophage-CSF, interleukin-1 [IL-1], and IL-6) and are activated and stimulated to growth by IL-1. Blood 80, 2835–2842. ¨ 2. Topley, N., Jorres, A., Luttmann, W., Petersen, M. M., Lang, M. J., Thierauch, K. H., Mu¨ller, C., Coles, G. A., Davies, M., and Williams, J. D. 1993. Human mesothelial cells synthesize interleukin-6: Induction by IL-1b and TNF-a. Kidney International 43, 226–233. 3. Topley, N., Brown, Z., Jo¨rres, A., Westwick, J., Coles, G., Davies, M., and Williams, J.D. 1993. Human peritoneal mesothelial cells synthezise IL-8: Synergistic induction by interleukin-1b and tumour necrosis factor alpha. American Journal of Pathology 42, 1876–1886. 4. Topley, N., Petersen, M. M., Mackenzie, R., Neubauer, A., Stylianou, E., Kaever, V., Davies, M., Coles, G., Jo¨rries, A., and Williams, J. D. 1994. Human peritoneal mesothelial cell prostaglandin synthesis: Induction of cyclooxygenase mRNA by peritoneal macrophage derived cytokines. Kidney International 44, 900–909. 5. Offner, F. A., Feichtinger, H., Stadlmann, S., Obrist, P., Marth, C., Klingler, P., Grage B., Schmahl, M., and Knabbe, C. 1996. Transforming growth factor-b synthesis by human peritoneal mesothelial cells. American Journal of Pathology 148, 1679–1687. 6. Hynes, R. O. 1992. Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69, 11–25. 7. Albeda, S. M., and Buck, C. A. 1990. Integrins and other cell adhesion molecules. FASEB Journal 4, 2868–2880. 8. Tuckwell, D. S., and Humphries, M. J. 1993. Molecular and cellular biology of integrins. Critical Review of Oncology and Hematology 15, 149–171. 9. Ruoslahti, E., and Reed, J. C. 1994. Anchorage dependence, integrins, and apoptosis. Cell 77, 477–478. 10. Jonjic, N., Peri, G., Bernasconi, S., Sciacca, F. L., Colotta, F., Pelicci, P. G., Lanfrancone, L., and Mantovani, A. 1992. Expression of adhesion molecules and chemotactic cytokines in cultured human mesothelial cells. Journal of Experimental Medicine 176, 1165–1174. 11. Klein, C. L., Bittinger, F., Skarke, C. C., Wagner, M., Ko¨hler, H., Walgenbach, S., and Kirkpatrick, C. J. 1995. Effects of cytokines on the expression of cell adhesion molecules by cultured human omental mesothelial cells. Pathobiology 63, 204–212. 12. Gailit, J., and Clark, R. A. F. 1994. Wound repair in the context of extracellular matrix. Current Opinion in Cell Biology 6, 717–725. 13. di Zerega, G. S. 1997. Biochemical events in peritoneal tissue repair. European Journal of Surgery 163 (Suppl. 577), 10–16. 14. Dobbie, J. W. 1992. Pathogenesis of peritoneal fibrosing syndromes (sclerosing peritonitis) in peritoneal dialysis. Peritoneal Dialysis International 12, 14–27. 15. Renvall, S., Lehto, M., and Pentinnen, R. 1987. Development of peritoneal fibrosis occurs under the mesothelial cell layer. Journal of Surgical Research 43, 407–412. 16. Vipond, M. N., Whawell, S. A., Thompson, J. N., and Dudley,

18.

19.

20.

21.

22.

23.

24.

25.

26. 27.

28.

29.

30.

31. 32.

H. A. F. 1990. Peritoneal fibrinolytic activity and intraabdominal adhesions. Lancet 335, 1120–1122. Hinsbergh, V. W. M., Kooistra, T., Scheffer, A. A., van Bockel, J. H., and van Muijen, G. N. P. 1990. Characterization and fibrinolytic properties of human omental tissue mesothelial cells. Comparision with endothelial cells. Blood 75, 1490–1497. Tietze, L., Elbrecht, A., Schauerte, C., Klosterhalfen, B., AmoTakyi, B., Gehlen, H., Winkeltau, G., Mittermayer, C., and Handt, S. 1998. Modulation of pro- and antifibrinolytic properties of human peritoneal mesothelial cells by transforming growth factor b1 (TGF-b1), tumor necrosis factor a (TNF-a) and interleukin 1b (IL-1b). Thrombosis and Haemostasis 79, 362–370. Niedbala, M. J., Crickard, K., and Bernac, R. J. 1986. Adhesion, growth and morphology of human mesothelial cells on extracellular matrix. Journal of Cell Science 85, 133–147. Yen, C. J., Fang, C. C., Chen, Y. M., Lin, R. H., Wu, K. D., Lee, P. H., and Tsai, T. J. 1997. Extracellular matrix proteins modulate human peritoneal mesothelial cell behavior. Nephron 75, 188–195. Yung, S., and Davies, M. 1998. Response of human peritoneal cell to injury: An in vitro model of peritoneal wound healing. Kidney International 54, 2160–2169. Barth, T. F. E., Bru¨derlein, S., Rinaldie, N., Mechtersheimer, G., and Mo¨ller, P. 1997. Pleural mesothelioma mimics the integrin profile of activated, sessile rather than detached mesothelial cells. International Journal of Cancer 72, 77–86. Witz, C. A., Montoya-Rodriguez, I. A., Miller, D. M., Schneider, B. G., and Schenken, R. S. 1998. Mesothelium expression of integrins in vivo and in vitro. Journal of the Society for Gynecological Investigations 5, 87–93. Aumailey, M., Mann, K. H., von der Mark, H., and Timpl, R. 1989. Cell attachment properties of collagen IV and Arg-Gly-Asp dependent binding to its a2 (VI) and a3 (VI) chains. Experimental Cell Research 181, 463–474. Charo, I. F., Nannizzi, L., Smith, J. W., and Cheresh, D. A. 1990. The vitronectinreceptor avb3 binds fibronectin and acts in concert with a5b1 in promoting cellular attachment and spreading on fibronectin. Journal of Cell Biology 111, 2795–2800. Bosman, F. T. 1993. Integrins: Cell adhesives and modulators of cell function. Histochemical Journal 25, 469–477. Whithaker, D., and Papadimitriou, J. M. 1985. Mesothelial healing: Morphological and kinetic investigations. Journal of Pathology 145, 159–175. Davila, R. M., and Crouch E. C. 1993. Role of mesothelial and submesothelial stromal cells in matrix remodeling following pleural injury. American Journal of Pathology 142, 547–555. Bertram, P., Tietze, L., Hoopmann, M., Treutner, K. H., and Schumpelick, V. Intraperitoneal mesothelial cell transplantation. European Journal of Surgery, in press. Leavesley, D. I., Ferguson, G. D., Wayner, E. A., and Cheresh, D. A. 1992. Requirement of the integrin b3 subunit for carcinoma cell spreading or migration on vitronectin and fibrinogen. Journal of Cell Biology 117, 1101–1107. Raghow, R. 1994. The role of extracellular matrix in postinflammatory wound healing and fibrosis. FASEB Journal 8, 823–831. Clark, R. A. F., Tonessen, M. G., Gailit, J., and Cheresh, D. A. 1996. Transient functional expression of avb3 on vascular cells during wound repair. American Journal of Pathology 148, 1407–1421.