Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Fibrosis and malignancy are the major clinical consequences of asbestos in the pleural space, occurring between 10 and 40 yr after exposure (1). In each condition there is extensive remodeling of the pleura, involving proteolysis of the extracellular matrix, cell proliferation and migration, and angiogenesis (4). In considering these changes, we asked whether the urokinase-type plasminogen activator (uPA) receptor (uPAR), a receptor implicated in each of these processes, could be involved in the pathogenesis of asbestos-induced pleural disease. Although uPAR has been identified on the surface of mesothelial cells (8), it is not known whether asbestos exposure might alter uPAR expression.

The uPAR is a glycosyl-phosphatidylinositol (GPI)- linked membrane protein that binds uPA and thereby localizes proteolytic activity at the cell surface (9). In addition to its role in proteolysis, uPAR mediates diverse cellular functions, including cell migration, differentiation, mitogenesis, and angiogenesis (8). uPAR, along with various integrin receptors, also binds vitronectin (VN), an adhesive protein shown to adsorb to the surface of crocidolite asbestos and increase its uptake by mesothelial cells (12). uPAR expression is known to be increased by several cytokines, including interleukin (IL)-1, transforming growth factor (TGF)-, and basic fibroblast growth factor (bFGF) (9, 13). On the basis of these observations, we speculated that exposure to asbestos could alter mesothelial cell uPAR expression, either directly or indirectly via products of asbestos-exposed monocytes.

In this study we determined whether asbestos induced uPAR expression in cultured rabbit or human (MeT5A) pleural mesothelial cells, either by direct exposure or by incubation with conditioned media from asbestos-exposed monocytes. We tested whether VN or serum adsorption onto asbestos fibers altered the effect of asbestos on the mesothelial cells or monocytes and, using antibody neutralization techniques, whether the effects were mediated by selected cytokines implicated in the pathogenesis of pleural inflammation and repair. Expression of uPAR was measured by the binding of radiolabeled uPA both to intact cells and to the receptor in the membrane extracts (ligand blotting) as well as by immunochemical detection in intact cells. uPAR messenger RNA (mRNA) was assayed to determine whether uPAR expression was associated with increased message. Our results demonstrate that chrysotile and crocidolite asbestos increase uPAR expression of cultured mesothelial cells either directly or via exposure to mediators from asbestos-stimulated monocytes. By contrast, neither asbestos nor wollastonite exposure induced uPA expression, although wollastonite but not asbestos appeared to localize exogenous uPA activity to the cell surface. These findings suggest that altered expression of uPAR by the mesothelium could influence local repair in asbestos-induced pleural disease.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Mesothelial Cells

Rabbit mesothelial cells were freshly isolated from preparations of rabbit pleura obtained from New Zealand white rabbits (female, 6 to 7 pounds, specific pathogen-free) as described (12). Primary cell cultures were maintained in growth medium consisting of RPMI with 10% fetal calf serum (FCS), 1% glutamine, and 2% antibiotic-antimycotic (GIBCO BRL, Grand Island, NY) at 37°C in 5% CO₂. MeT5A cells, an immortalized human pleural mesothelial cell line, were obtained from Dr. Brenda Gerwin (National Cancer Institute, Bethesda, MD). MeT5A cells were maintained in growth medium at 37°C in 5% CO₂, as previously reported (8).

Particulates

Chrysotile (length, 1 to 5 µm) and crocidolite (length, 2 to 10 µm) asbestos were obtained from NIEHS, Research Triangle Park, NC. Wollastonite, a nonfibrogenic fiber, was obtained from NYCO, Willsboro, NY. Before use, all particulates were freshly dispersed in Hanks' balanced salt solution (HBSS) by 10 passages through a syringe fitted with a 23-gauge needle. All particulates were tested using the Limulus assay (Sigma, St. Louis, MO) and found to be free of endotoxin contamination (< 0.01 EU/ml).

For adsorption experiments, chrysotile and crocidolite asbestos or wollastonite were preincubated with either serum (100%) or VN (50 µg/ml) (GIBCO BRL) for 30 min at 37°C and then washed twice before being added to mesothelial cells. VN was determined to be free of endotoxin by means of the Limulus assay described previously. Human serum was collected under sterile conditions and frozen in 1-ml aliquots at 20°C until use. All serum was used within 14 d of collection.

Culture Conditions

Mesothelial cells were grown in T-75 plastic tissue culture flasks (Falcon; Becton Dickinson, Franklin Lakes, NJ) for ¹²⁵I-uPA binding and in T-125 plastic tissue-culture flasks (Becton Dickinson) for ligand blotting, as described later. At confluence, growth medium was removed and replaced with fresh medium (control) or with media containing chrysotile or crocidolite asbestos at concentrations indicated in the figures. Mesothelial cells were cultured in the presence of asbestos for 18 h at 37°C in 5% CO₂ and then used for specific measurements of uPAR and uPAR mRNA.

Monocytes

In a separate series of experiments, we studied the indirect effect of asbestos on mesothelial cells by exposing mesothelial cells to supernatants from asbestos-exposed monocytes. Monocytes were isolated as previously described using differential centrifugation followed by adherence to plastic culture dishes (16). The monocytes were cultured with media alone (control), with uncoated chrysotile asbestos, or with asbestos preincubated with serum or VN for 18 h at 37°C in 5% CO₂. Monocyte supernatants were collected and centrifuged at 500 × g for 5 min to remove debris and particulates, and then added to mesothelial cells in a 1:1 dilution with media for 18 h at 37°C. Mesothelial cells were then collected to measure uPAR expression as described later. In antibody neutalization experiments, we tested the ability of antibodies to tumor necrosis factor (TNF)- or TGF- to block the ability of monocyte-conditioned medium to stimulate uPAR expression, assessed through the binding of radiolabeled uPA to the surface of MeT5A cells, as described previously. In these experiments we treated the monocyte-conditioned media with 25 µg/ml of the immunoglobulin (Ig) G fraction of the neutralizing monoclonal antibodies (mAbs) to the cytokines or control mouse IgG for 1 h at 37°C, after which the ability of the media to induce uPAR expression in MeT5A cells after 18 h of exposure was determined by binding of radiolabeled uPA, as described later.

¹²⁵I-uPA Cell Binding Assay

Cell-surface expression of uPAR was determined by measurement of the binding of ¹²⁵I-uPA to intact cells. The iodination of ¹²⁵I-uPA and the binding assay were performed as previously described by Shetty and colleagues (17). After being cultured in the presence of asbestos or control conditions, mesothelial cells were washed once in buffer containing glycine-HCl to remove endogenous uPA bound to uPAR. To do this, we used the technique originally described by Stoppelli and colleagues (18). We previously confirmed that this technique reverses the binding of exogenous uPA using MS-1 mesothelioma cells, and found that acid-treated cells bound radiolabeled uPA to the same extent as did the cells before acid treatment (8). After two washes in serum-free media, the cells were incubated with iodinated uPA for 120 min at 4°C. The cells were then washed four times and lysed for measurement of radioactivity. Nonspecific binding was determined by incubation of the cells with iodinated uPA in the presence of a 500-fold excess of unlabeled uPA. Results are reported as picomoles of uPA per well, as previously reported (17). The average nonspecific binding for assays in which MeT5A cells were used was 32%, whereas that of assays in which rabbit pleural mesothelial cells were used was 27%. In preliminary experiments, we determined that the presence of 10% FCS did not stimulate uPA binding to MeT5A cells.

Ligand Blotting

The expression of uPAR was next studied by measurements of the binding of ¹²⁵I-uPA to uPAR separated from other cell-membrane constituents by sodium dodecyl sulfate (SDS) gel electrophoresis under nonreducing conditions (8). This ligand-blotting approach may be made more specific by isolating the receptor from membrane-associated fibers that might bind ¹²⁵I-uPA. These experiments required relatively large numbers of cells to generate the requisite membrane preparations and were therefore only performed using MeT5A cells. After culturing in the presence of asbestos (uncoated or protein-adsorbed), wollastonite, or no particulate for 18 h, the cells were acid-treated as described previously to remove endogenous uPA. Membrane fractions of acid-treated cells were prepared according to the procedure of Shetty and associates (8, 19), separated by SDS gel electrophoresis under nonreducing conditions and electroblotted to a nitrocellulose membrane. The membrane was then blocked with nonfat milk (5%), incubated overnight with ¹²⁵I-uPA (300 k cpm) at 4°C, washed three times with Tris-buffered saline, air-dried, and autoradiographed.

Ribonuclease Protection Assay

The steady-state expression of uPAR mRNA by asbestos-exposed or control untreated cells was determined by ribonuclease (RNase) protection assay (RPA). After exposure of cells to asbestos, total RNA was extracted using a standard procedure with Tri-Reagent (20). Total RNA was incubated with a ³²P-labeled cellular RNA probe for uPAR (8). After hybridization, the RNA was digested and washed. Fragments protected by binding to the probe were separated by electrophoresis on a Tris borate EDTA (TBE)-polyacrylamide/urea gel and then autoradiographed.

Immunohistochemical Analysis of uPAR

Rabbit mesothelial cells and human MeT5A cells cultured overnight on laminin-coated coverslips in 24-well plates were exposed to crocidolite (3 µg/cm²), chrysotile (3 µg/ cm²), or wollastonite (6 µg/cm²) for different times (30 min, 2 h, 4 h, 6 h, and 18 h). After exposure, the cells were washed in phosphate-buffered saline (PBS), fixed in freshly prepared paraformaldehyde (2%) for 10 min at room temperature, washed three times in PBS, and permeabilized in 0.1% Triton for 5 min. After blocking in 5% bovine serum albumin (BSA) for 10 min, the cells were incubated with the primary antibody, mouse antihuman uPAR (American Diagnostics, Greenwich, CT), at 1:100 for 1 h at room temperature, washed three times, and incubated with a secondary antibody, a fluorescein isothiocyanate (FITC)-labeled sheep antimouse antibody (Cappel, Aurora, OH) at 1:100 for 1 h. To exclude nonspecific fluorescent staining in control experiments, the primary antibody only was omitted. After washing, the cells were sealed in Slowfade (Molecular Probes, Eugene, OR) to maintain fluorescence, and were examined and photographed under fluorescence confocal microscopy (MRC-600; Bio-Rad, Hercules, CA).

Measurement of uPA-Related Plasminogen Activation and uPA Antigen Expression by Fiber-Treated MeT5A Cells

Activation of plasminogen by fiber-treated cells was assessed using our previously described technique in which samples are assayed for liberation of plasmin by measurement of thioesterase activity (8). Plasmin concentrations of the samples are determined on the basis of the activity generated by plasmin standards. In experiments to test the ability of fiber pretreatment to localize uPA activity to the MeT5A cell surface, cells were initially treated with 10 µg/ cm² for 18 h. The cells were then glycine-treated to remove endogenous uPA from the cell surface, after which they were exposed to a final concentration of 2 µg/ml of two-chain uPA in 10 mM N-2-hydroxyethylpiperazine-N'-ethane sulfonic acid, 137 mM NaCl, 4 mM KCl, 11 mM glucose, 0.03% Na azide, 0.5% BSA, and 5 mM CaCl₂ for 2 h at 4°C. The cells were then washed five times with 10 mM Tris, 150 mM NaCl, and 5 mM CaCl₂, pH 7.45, after which they were rinsed with HBSS. The cells were then maintained in HBSS supplemented with 25 µg/ml plasminogen, and plasminogen activator activity was then determined as described previously.

We used SDS-polyacrylamide gel electrophoresis (PAGE) and Western blotting to measure the antigenic uPA level of the cells. Cell lysates were prepared as described elsewhere (21) and the proteins were separated by SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was blocked with 1% BSA for 1 h at room temperature, followed by overnight hybridization with a monoclonal uPA antibody (American Diagnostics) and the blot was developed using enhanced chemiluminescence.

Lactate Dehydrogenase and Total Protein Assays

Asbestos cytotoxicity was approximated by release of lactate dehydrogenase (LDH) at 18 h and compared between treatment groups and control. LDH was measured using the Sigma LDH optimized LDH 1.1.1.27 ultraviolet-method kit and is reported as units per liter per well. Total protein determinations were made using a spectrophotometric technique using bicinchoninic acid (Pierce Chemical Co., Rockford, IL). In experiments designed to test the ability of fibers to adsorb serum proteins, 1 mg of each fiber type was used to adsorb proteins from 1:500 or 1:750 dilutions of normal pooled plasma (George King, Overland Park, KS) in normal saline solution. The fibers were exposed to the serum dilutions for 1 h at 37°C and the residual protein concentrations of the diluted serum were determined to compare the proteins adsorbed to equivalent amounts of chrysotile, crocidolite, and wollastonite. Determinations of serum proteins were made in triplicate.

Statistics

Results for uPA binding and LDH release were analyzed using INSTAT 2.0 on a personal computer. Evaluation of comparisons between treatment groups and control was performed by analysis of variance with appropriate post-testing comparisons. Data are represented as means ± one standard deviation. For each result, statistical significance is defined as a P < 0.05 for the described comparisons and is indicated in figures by an asterisk (*).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cell-Surface Binding of ¹²⁵I-uPA

We initially tested the ability of asbestos fibers to influence expression of uPAR at the mesothelial cell surface. In rabbit mesothelial cells exposed to asbestos for 18 h, there was a significant increase in uPAR expression as measured by specific binding of ¹²⁵I-uPA (P < 0.05, n = 8 experiments) (Figure 1). In human MeT5A mesothelial cells exposed to asbestos, there was also a significant increase in uPAR expression over that in cells treated with media alone (P < 0.05, n = 12 experiments) (Figure 2). In all experiments there was a significant increase in the binding of radiolabeled uPA to MeT5A compared with rabbit mesothelial cells. These differences are attributable to an apparent increase in the number of uPAR per cell in the MeT5A (data not shown). In additional experiments (performed in triplicate), we determined that serum exerted little effect on the basal expression of uPAR by MeT5A cells. Binding of radiolabeled uPA averaged 36.2 pmol/10⁵ cells in cells pretreated for 18 h with 10% FCS in RPMI medium versus 39.8 pmol/10⁵ in cells that were not exposed to serum over the same interval but were studied under otherwise identical conditions.

View larger version (28K): [in this window] [in a new window]   Figure 1.   Effect of asbestos on rabbit mesothelial cell uPAR expression. Rabbit mesothelial cells were cultured with increasing concentrations of crocidolite (open bars) and chrysotile (hatched bars) asbestos for 18 h, after which 125I-uPA binding was measured as described in MATERIALS AND METHODS. *P < 0.05 compared with media alone (control, solid bar). Data illustrated are representative of the findings in eight separate experiments.

View larger version (26K): [in this window] [in a new window]   Figure 2.   Effect of asbestos on MeT5A human mesothelial cell uPAR expression. MeT5A were cultured with crocidolite (open bars) and chrysotile (hatched bars) asbestos for 18 h, then 125I-uPA binding was measured. *P < 0.05 compared with media alone (control, solid bar). Data illustrated are representative of the findings in 12 separate experiments.

uPAR expression on mesothelial cells exposed to VN or serum-coated asbestos was greater than on those exposed either to uncoated asbestos or to no particulate (P < 0.05, n = 12 experiments) (Figure 3). There was no increase in uPAR expression in mesothelial cells exposed to uncoated wollastonite or to VN- or serum-coated wollastonite.

View larger version (22K): [in this window] [in a new window]   Figure 3.   Effect of VN and serum adsorption onto asbestos and wollastonite on MeT5A uPAR expression. Crocidolite asbestos (open bars) and wollastonite (hatched bars), uncoated or coated with VN or serum, were incubated with MeT5A for 18 h. 125I-uPA binding was then measured. *P < 0.05 compared with media alone (control, solid bar), **P < 0.01 compared with control, and P < 0.05 compared with uncoated crocidolite asbestos. Data illustrated are representative of the findings in 12 separate experiments.

We next sought to determine whether products elaborated by asbestos-treated monocytes could likewise influence expression of mesothelial cell uPAR. The conditioned media from monocytes exposed to uncoated chrysotile asbestos (25 µg/cm²) significantly increased uPAR expression in mesothelial cells compared with that of cells cultured with media alone. Conditioned media supernatants from monocytes exposed to VN-coated chrysotile increased uPAR expression significantly more than did supernatants from monocytes exposed to uncoated asbestos (P < 0.05, n = 8 experiments) (Figure 4).

View larger version (34K): [in this window] [in a new window]   Figure 4.   Effect of supernatants from asbestos-exposed monocytes on MeT5A uPAR expression. Human blood monocytes were isolated and cultured for 18 h in media alone (control, open bar) or in media containing 25 µg/cm2 chrysotile asbestos that was uncoated or coated with VN or serum (hatched bars). Supernatants were removed and applied to MeT5A cells for 18 h followed by 125I-uPA binding. *P < 0.05 compared with control, **P < 0.01 compared with control, and P < 0.05 versus uncoated asbestos. Data illustrated are representative of the findings in eight separate experiments.

The LDH concentrations of the conditioned media of asbestos-treated rabbit mesothelial or MeT5A cells did not differ from those of controls after 18 h (data not shown). For all conditions, there was an increase in cell number over 18 h. In the case of cells exposed to media alone or wollastonite, there was an average increase in cell number of 60% at 18 h compared with 0 h. For cells exposed to bare chrysotile or crocidolite asbestos, there was an average increase in cell number of 48%. For serum-coated asbestos, the average increase was 52% over 18 h. Therefore, with chrysotile or crocidolite treatment there were, on average, approximately 10% fewer cells at 18 h than with media- or wollastonite-treated cells.

Ligand Blotting of uPAR

We next used a complementary technique: ligand blotting, to assess asbestos-induced changes in mesothelial cell uPAR expression. By measurement of ¹²⁵I-uPA binding to uPAR separated from membrane proteins by gel electrophoresis, there was a reproducible increase in uPAR expression in the MeT5A exposed to uncoated chrysotile asbestos, and a further increase in MeT5A cells exposed to either VN- or serum-coated chrysotile asbestos (Figure 5A). Similar responses were observed in cells exposed to crocidolite asbestos (data not shown). There was no change in uPAR expression in cells exposed to wollastonite, whether uncoated, VN-coated, or serum-coated (data not shown). These data independently confirm the data using ¹²⁵I-uPA binding to intact cells and indicate that asbestos treatment increases mesothelial cell uPAR expression.

View larger version (32K): [in this window] [in a new window]   View larger version (28K): [in this window] [in a new window]   Figure 5.   Ligand blotting of MeT5A cultured with asbestos or with supernatants of monocytes exposed to asbestos. (A) Autoradiograph of ligand blot and accompanying densitometry readings for mesothelial cells cultured with media alone or with chrysotile asbestos (25 µg/cm2) uncoated or coated with serum or VN for 18 h. These data are representative of n = 6 separate experiments. (B) Autoradiograph of ligand blot and accompanying densitometry readings for mesothelial cells cultured with supernatants from monocytes cultured in media alone or in chrysotile asbestos (25 µg/cm2) uncoated or coated with serum or VN. These data are representative of n = 6 separate experiments.

By ligand blotting, media from monocytes exposed to uncoated chrysotile did not induce a detectable increase in uPAR expression in MeT5A cells, but media from monocytes exposed to VN- or serum-coated induced a significant increase (Figure 5B). In follow-up experiments, we tested the ability of neutralizing antibodies to either TNF- or TGF- to block the upregulation of uPAR by the media of monocytes exposed to serum-coated chrysotile. In these experiments, performed in triplicate, we found that only the antibody to TGF- attenuated binding of radiolabeled uPA (average = 77.9; range, 60.4-94.2 pmol/10⁵ cells) to the MeT5A cell surface. By contrast, cells treated with media alone bound 148.3 (range, 135.4-165.9 pmol/10⁵ cells); cells treated with media plus IgG bound 118.5 (range, 110.1-127.1 pmol/10⁵ cells); and cells treated with antibody to TNF- bound 141.7 (range, 94.6-173.9 pmol/10⁵ cells). These data suggest that TGF- could contribute to the upregulation of uPAR at the MeT5A cell surface.

Immunohistochemical Analysis of uPAR

We used a third independent technique, immunohistochemistry, to substantiate asbestos-induced uPAR expression at the surface of mesothelial cells. As seen in Figure 6, fluorescent staining of uPAR of mesothelial cells treated with crocidolite and chrysotile asbestos at both 6 and 18 h was increased compared with control cells and those treated with wollastonite. The fluorescent activity is shown to be specific for uPAR by a failure of staining with omission of the primary antibody.

View larger version (103K): [in this window] [in a new window]   Figure 6.   Immunocytochemical detection of uPAR in rabbit mesothelial cells. Rabbit cells exposed to wollastonite (6 µg/ cm2), crocidolite (3 µg/cm2), or chrysotile (3 µg/cm2) for different times were washed, fixed, permeabilized, blocked, and stained with mouse antihuman uPAR 1:100 for 1 h, followed by an FITC-labeled sheep antimouse antibody (1:100 for 1 h). Fluorescent confocal images are seen on the left with corresponding differential contrast on the right. Data are representative of the findings of three separate experiments. In control experiments, the primary antibody was omitted. No increases were detected at 2 and 4 h.

Changes in uPAR mRNA Expression by RNase Protection

We next assessed the ability of asbestos fibers to influence steady-state uPAR message expression by MeT5A cells. uPAR mRNA was increased in MeT5A exposed to crocidolite asbestos for 6 h and further increased in MeT5A exposed to VN-coated crocidolite asbestos (Figure 7). Cells treated with chrysotile asbestos similarly demonstrated increased expression of uPAR mRNA (data not shown).

View larger version (33K): [in this window] [in a new window]   View larger version (33K): [in this window] [in a new window]   Figure 7.   Effect of asbestos on uPAR mRNA expression in MeT5A cells. Autoradiograph and densitometry readings from RPA for uPAR mRNA (A) or actin mRNA (B) in mesothelial cells cultured in media alone or with crocidolite asbestos (25 µg/ cm2) uncoated or coated with VN. These data are representative of the findings of four separate experiments.

Effects of Fiber Treatment on Expression of uPA Activity at the Cell Surface

We then sought to determine the effects of fiber treatments on the expression of uPA by MeT5A cells. We inferred that asbestos fibers would adsorb uPA and that the fibers themselves might localize uPA to the surface of asbestos-coated cells. To address this possibility, we first sought to establish the ability of chrysotile, crocidolite, and wollastonite to adsorb radiolabeled uPA, using the same ¹²⁵I-uPA binding assay conditions described earlier except that the determinations were made in a cell-free system. A total of 50 µg of each fiber type was tested in quadruplicate determinations. Crocidolite and chrysotile adsorbed comparable amounts of radiolabeled uPA: 2,392.1 ± 198.0 pmol versus 2,289.2 ± 142.2 pmol, respectively. Wollastonite adsorbed 58 to 60% of the radiolabeled uPA compared with the asbestos fibers: 1,373.8 ± 360.0 pmol. These data confirm that uPA adsorbs to each of the fiber types, with more adsorption to the asbestos fibers than to wollastonite.

We next tested the specificity of the adsorption of uPA by the fibers. To this end, we measured the ability of the fibers to adsorb proteins from dilutions of serum and found that chrysotile and crocidolite likewise adsorbed comparable amounts of serum proteins; residual protein after adsorption by 1 mg of chrysotile or crocidolite was 0.108 (0.102-0.115 mg/ml) and 0.093 (0.086-0.102 mg/ml, n = 3) versus a dilution of 1:500 of pooled serum in saline, 0.176 (0.167-0.183) mg/ml. Wollastonite adsorbed slightly less protein; residual protein after adsorption was 0.140 (0.134- 0.144 mg/ml, n = 3). These data suggest that adsorption of uPA by these fibers is nonspecific and that differential adsorption may relate to differences in the physical properties of the selected fiber types, the relatively greater surface area of the asbestos versus wollastonite fibers.

We then tested the ability of the asbestos fibers to induce uPA expression by MeT5A cells treated with chrysotile, crocidolite, or wollastonite. To explore this possibility, we first treated MeT5A cells with a range of fiber concentrations we previously found capable of increasing the binding of radiolabeled uPA: 10 or 25 µg/cm² of either chrysotile, crocidolite, or wollastonite. As shown in Table 1, we found no increase of uPA activity in either the cell lysates or conditioned media of the treated cells. By Western blotting, we likewise found no induction of uPA antigen in the fiber-treated cells (Figure 8). These data indicate that induction of uPA activity or antigen expression by asbestos or wollastonite-treated MeT5A cells was not detectable over a range of fiber exposures able to induce increased expression of uPAR.

View larger version (29K): [in this window] [in a new window]   Figure 8.   Western blot of uPA expressed by fiber-treated MeT5A cells. Figure shows autoradiographs of the Western-blotted proteins of fiber-treated cells. (a) Proteins of 100-µg aliquots of cell lysates were studied by blotting with an mAb to uPA. (b) Identical samples immunoblotted with a control mouse IgG antibody. In a and b: lane 1, uPA control, 125 ng; lane 2, cells treated with 10 µg/cm2 chrysotile; lane 3, 25 µg/cm2 chrysotile; lane 4, 10 µg/cm2 crocidolite; lane 5, 25 µg/cm2 crocidolite; lane 6, 10 µg/cm2 wollastonite; lane 7, 25 µg/ cm2 wollastonite; lane 8, cells without fiber treatment. The 50- and 30-kD forms of the uPA standard are illustrated in a, lane 1.

Finally, we sought to explore the possibility that fiber treatment of the cells might localize uPA-related plasminogen activator activity to the cell surface. To address this possibility, we treated MeT5A cells with fibers, added exogenous uPA for 2 h at room temperature, removed the uPA, and then washed the cells five times before measuring residual uPA-related plasminogen activator activity, as described in MATERIALS AND METHODS. Pretreatment with either chrysotile or crocidolite followed by exogenous uPA did not increase expression of uPA activity by the cells (Table 2). Interestingly, uPA activity of wollastonite-treated cells was consistently increased. In follow-up neutralization experiments, we found that the increased uPA activity of wollastonite-treated cells could be completely inhibited with either a mAb to uPA or by aprotinin, indicating that the measured proteolytic activity was related to uPA- mediated plasminogen activator activity and most likely involved elaboration of increased amounts of plasmin.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study we used three independent techniques to demonstrate that both crocidolite and chrysotile asbestos induce an increase in the expression of uPAR in mesothelial cells. The asbestos-induced increase in uPAR varies with the amount and type of asbestos used, with crocidolite asbestos upregulating uPAR expression at lower concentrations than seen with chrysotile asbestos. Asbestos also induces uPAR expression in association with increases in uPAR mRNA. In addition, supernatants from monocytes exposed to asbestos, especially when coated with VN or serum, increase uPAR expression on mesothelial cells. Although these increases are reproducible and were confirmed by multiple techniques, they are not as pronounced as the previously described increases due to cytokine exposure of mesothelial cells (8). The smaller magnitude of increases in uPAR expression might be expected considering that exposure of the particulates to the cell surface is not homogeneous. Despite a lower overall magnitude, the increases in uPAR induced by asbestos might suffice to influence the local tissue response to asbestos.

uPAR is expressed on the surface of mesothelial cells and binds uPA in either the active or the precursor, inactive form (9). When the active form of uPA is bound to uPAR, it localizes plasminogen activation to the region near the cell. Plasmin then acts on fibrin and modifies the transitional fibrin matrix that occurs adjacent to the mesothelial cell in the setting of pleural injury. uPA is present in biologic fluids, such as pleural fluid, in concentrations that could allow binding to the expressed receptors (22). We therefore investigated the possibility that asbestos might promote increased uPA-related proteolytic activity as well as uPAR at the cell surface. We found that exposure of MeT5A cells to the same concentrations of asbestos that induced uPAR expression did not induce expression of uPA, as judged by functional or antigenic criteria. In addition, pretreatment of the cells with asbestos (for 18 h) followed by exposrue to exogenous uPA (for 2 h) did not increase plasminogen activator activity expressed by the cells.

It is possible that we could not detect increased plasminogen activator activity because uPA activity was blocked by concurrent expression of plasminogen activator inhibitor-1 (PAI-1) over the 2-h course of exogenous uPA exposure, because PAI-1 was induced as a result of prior asbestos exposure, or because uPA-PAI-1 complexes occupied uPAR and initiated uPA-PAI-1-receptor internalization (8). Given the modest level of uPAR induction by asbestos fibers, it is also possible that small increments of receptor-bound active uPA could have escaped detection by our functional assay. BSA, included in the assay buffer to simulate physiologic exposure of the cells to uPA in a protein mixture, could also have blocked adsorption of active uPA to fibers protruding from the cell surface. Although we did not detect increased mesothelial cell plasminogen activator activity under the conditions we used, others have shown that asbestos exposure increases uPA-mediated proteolytic activity in lung tissue (23, 24), which suggests that other cell types may respond differently. Interestingly, wollastonite, which did not induce uPAR or uPA, appeared to localize exogenous uPA-related plasminogen activator activity to the cell surface. The mechanism is currently unclear but likely involves adsorption and conserved activity of exogenous uPA, inasmuch as direct exposure to wollastonite did not induce uPA expression by the mesothelial cells.

Our findings suggest that the induction of uPAR by asbestos fibers might influence lung remodeling via noncatalytic functions of uPAR that are initiated by the binding of uPA and may also play an important role in the development of asbestos-induced pleural disease. In several cell types, including malignant and nonmalignant mesothelial cells and lung fibroblasts, uPA binding to uPAR promotes mitogenesis (8, 17). It is also possible that increased uPAR expression could influence cellular signaling independent of receptor occupancy, a possibility that is now being explored in other laboratories (R. Sitrin, personal communication). In addition, uPAR appears to be involved in cellular migration and adherence independent of receptor occupancy (10, 25, 26). Lastly, a protracted increase in the division of mesothelial cells could increase the likelihood of DNA damage caused by asbestos (4), potentially influencing pleural remodeling via malignant transformation and the development of malignant mesothelioma.

The mechanism of asbestos-induced modification of mesothelial cell uPAR expression may be due to direct exposure to the fibers or may be indirect. Because VN adsorption to asbestos enhances uPAR expression, we suspect direct involvement of a VN receptor. Integrin receptors, including v5 as well as uPAR itself, have been shown to bind VN and could alter the interaction between the cells and the VN-coated fibers (27, 28). The integrin v5 in particular has been shown to bind and mediate uptake of VN-coated fibers, an effect that might increase uPAR expression by virtue of an increased cellular load of asbestos (12). A VN receptor might also mediate a separate signaling function leading to increased uPAR expression.

In addition to a direct effect of asbestos on uPAR expression, there is evidence for an indirect mechanism related to macrophage-derived, and perhaps mesothelial cell- derived, products. Cytokines, such as IL-1, TGF-, and bFGF, are known to upregulate uPAR expression (13, 29). IL-1 in particular may be important; it is responsible for asbestos-induced release of IL-8 by mesothelial cells and for uPAR expression on epithelial cells in acute lung injury (15, 30). Whereas the direct effects of asbestos on mesothelial cell uPAR could account for local uPAR upregulation, the indirect effect of monocyte conditioned media indicates that asbestos-mediated upregulation of mesothelial cell uPAR can involve the participation of other cell types. Our results indicate that TGF-, but not TNF-, contributes to the increased uPAR expression induced by exposure of MeT5A cells to the conditioned media of serum-coated chrysotile-treated monocytes.

The adsorption of proteins onto asbestos fibers can significantly augment uPAR expression by mesothelial cells. Our data also confirm that uPA and other serum proteins adsorb to asbestos and wollastonite fibers. Studies of the pathogenesis of asbestos have focused on the changes in different cell types that occur when the cells are exposed to "bare" particulates, or particulates in protein-free media. Our present results and those of previous studies (31, 32) suggest that when asbestos exists in vivo, its bioactivity is altered by the proteins present in the environment. VN, a serum protein, is present in pleural and airway fluid (12) and has been shown to adsorb to asbestos and remain functional when fibers are incubated in pleural fluid, bronchoalveolar lavage fluid, or serum (12). It is very likely that asbestos present in the pleural space will have adsorbed VN or other serum proteins and that these proteins will influence the interaction of fiber and cell, leading to increased uPAR expression by the mesothelium.

In conclusion, we have shown that asbestos increases uPAR expression on mesothelial cells in vitro either directly by fiber exposure or indirectly via elaboration of monocyte-derived products. These data suggest a novel mechanism by which asbestos exposure could influence uPAR-mediated cell functions relevant to pleural injury and repair.