Asbestos-associated mesothelial cell autoantibodies promote collagen deposition in vitro

Abstract

Fibrosis, characterized by excessive collagen protein deposition, is a progressive disease that can fatally inhibit organ function. Prolonged exposure to pathogens or environmental toxicants such as asbestos can lead to chronic inflammatory responses associated with fibrosis. Significant exposure to amphibole asbestos has been reported in and around Libby, Montana due to local mining of asbestos-contaminated vermiculite. These exposures have been implicated in a unique disease etiology characterized predominantly by pleural disorders, including fibrosis. We recently reported the discovery of mesothelial cell autoantibodies (MCAAs) in the sera of Libby residents and demonstrated a positive and significant correlation with pleural disease; however, a mechanistic link was not determined. Here we demonstrate that MCAAs induce pleural mesothelial cells to produce a collagen matrix but do not affect production of the pro-inflammatory cytokine tumor growth factor-β. While autoantibodies commonly induce a pro-fibrotic state by inducing epithelial-mesenchymal transition (EMT) of target cells, we found no evidence supporting EMT in cells exposed to MCAA positive human sera. Although implicated in other models of pulmonary fibrosis, activity of the protein SPARC (secreted protein, acidic and rich in cysteine) did not affect MCAA-induced collagen deposition. However, matrix formation was dependent on matrix metalloproteinase (MMP) activity, and we noted increased expression of MMP-8 and -9 in supernatants of mesothelial cells incubated with MCAA positive sera compared to control. These data suggest a mechanism by which MCAA binding leads to increased collagen deposition through altering MMP expression and provides an important mechanistic link between MCAAs and asbestos-related, autoimmune-induced pleural fibrosis.

Figure 1

Binding of serum antibodies to human mesothelial cells indicates the presence of mesothelial cell autoantibodies (MCAAs) as determined by cell-based ELISA as previously described (Marchand et al., 2012). (A) Binding by MCCA +ive samples was significantly higher compared to the MCAA −ive samples, though no differences were seen in MCAA binding of individual serum samples compared pooled sera. Mean ±SE, n =at least 6, p =0.002. (B) Binding in pooled sera samples previously identified as MCAA +ive was significantly higher than pooled samples of MCAA −ive sera. MCAAs were not detected in normal human sera (NHS) obtained from subjects with no known asbestos exposure. Removal of IgG by Protein G precipitation (cleared serum) resulted in loss of MCAA binding compared to NHS, p =0.001. Mean ±SE, n =6, *p<0.001 by one-way ANOVA.

Figure 2

MCAAs induced deposition of extracellular type I collagen proteins. (A) Collagen detection significantly increased following exposure to MCAA +ive sera compared to MCAA −ive sera or NHS. Clearance of IgG reduced collagen detection to background levels. Mean ±SE, n =8, *p<0.001 by one-way ANOVA. Collagen deposition by MeT5A cells was determined using cell-based ELISA following a 4-day incubation of cells with human sera. (B) No significant differences in collagen detection were observed when cells were exposed to pooled MCAA +ive sera compared to individual samples, but cell exposure to both pooled and individual MCAA +ive samples significantly increased collagen compared to respective MCAA −ive samples. Mean ± SD, n =3, p<0.05 by one-way ANOVA with Tukey post hoc test.

Figure 3

MCAA exposure does not induce a mesenchymal transition in cultured human pleural mesothelial cells. (A) SMA expression was measured in cells exposed to human sera or TGF-β (5 ng) as a positive control for 4 days. Cells were fixed and stained with primary anti-SMA and secondary FITC conjugated antibodies and mean fluorescence measured by flow cytometry. TGF-β ( ) induced SMA expression compared with exposure to secondary antibody control ( , ), MCAA +ive ( , ), or MCAA −ive ( ) sera. Exposure to MCAA +ive or −ive sera resulted in increased SMA expression over secondary control (x-axis) but expression did not increase to the same degree as TGF-β exposure. TGF-β was detected in MCAA +ive and −ive sera (B) and supernatants (C) of cells exposed to these sera samples. TGF-β was detected using a cytokine bead array analysis and concentrations suggest that the TGF-β detected in supernatants was added upon addition of sera and not produced by the mesothelial cells, mean ± SD, n =3). (D) Total soluble collagen concentrations in cell supernatants were determined using a Sircol Soluble Collagen Assay. Incubation with sera from asbestos-exposed subjects did not significantly affect soluble collagen compared to NHS.

Figure 4

SPARC expressed by human mesothelial cells associates with extracellular matrix but does not contribute to matrix formation. (A) Immunoblot analysis was used to detect endogenous SPARC in lysates of MeT5A cells incubated with human sera. (B) Densitometry analysis using Image J software revealed no differences in expression levels between cells exposed to MCAA +ive sera or controls, mean ±SD, n =2. (C) Using a cell-based ELISA, detection of SPARC increased significantly in MeT5A cells exposed to MCAA +ive sera compared to controls, mean ±SE, n =4, *p<0.001 as determined by one-way ANOVA. Cells were incubated with human sera for 4 days and SPARC detected using a rabbit polyclonal IgG anti-SPARC, H-90 antibody. (D) Addition of anti-SPARC antibody ( ) did not significantly alter detection of collagen proteins compared to no treatment (■), indicating that SPARC does not directly affect MCAA-induced collagen deposition, as determined by two-way ANOVA, mean ±SE, n =4.

Figure 5

MCAA-induced collagen deposition is dependent on MMP activity. (A) Cell-based ELISA was used to detect extracellular collagen deposited by MeT5A cells exposed to human sera for 4 days in the absence (■) or presence ( ) of the non-specific MMP inhibitor EDTA (5 mM). Collagen deposition was significantly inhibited by EDTA in cells exposed to MCAA +ive sera (*p =0.018) but not to NHS (p =0.28) or MCAA −ive sera (p =0.32) as determined by two-way ANOVA. (B) Collagen zymography revealed the presence of MMP-8 and MMP-9 in supernatants of cells exposed to human sera for 4 days. MMPs were identified by their reported molecular weights and bacterial collagenase (200 pg) was used as a positive control. (C) Immunoblot analysis of MMP-8 confirmed its identity. (D) Densitometry analysis of the immunoblot revealed significantly more MMP-8 present in supernatants of cells exposed to MCAA +ive sera compared to NHS. However, exposure to MCAA −ive sera also resulted in increased MMP-8, although this increase was not significant compared to NHS. Statistical significance is indicated by different letters at p<0.01 by one-way ANOVA, mean ±SD, n =3.