Anti-Heat Shock Protein 70 Autoantibodies from Patients with Idiopathic Pulmonary Fibrosis Epigenetically Enhance Lung Fibroblast Apoptosis Resistance and Bcl-2 Expression

Abstract

IgG autoantibodies to heat shock protein 70 (HSP70) are found in many immune-mediated clinical syndromes, and their presence among patients with idiopathic pulmonary fibrosis (IPF) portends especially poor outcomes. However, pathological effects of IPF anti-HSP70 have not been studied extensively. IPF lung fibroblasts are apoptosis resistant, and this dysregulation contributes to the accumulation of fibroblasts that characterizes the disease. During stress, HSP70 protein is exported extracellularly, where it binds to cognate cell surface receptors that mediate a variety of functional effects, including apoptosis inhibition. We hypothesized anti-HSP70 could engage HSP70-receptor complexes on fibroblasts that alter their apoptosis susceptibility. We found HSP70 is ubiquitously expressed on primary human lung fibroblasts. Treatment with anti-HSP70 isolated from patients with IPF with acute exacerbations increased Bcl-2 expression in human lung fibroblasts and reduced their susceptibility to staurosporine-induced apoptosis. Chromatin immunoprecipitation assays showed Bcl-2 gene promoter regions are enriched with the active histone mark H4 lysine 16 acetylation, and this was increased in the autoantibody-treated fibroblasts. When H4 lysine 16 acetylation was decreased by knocking down its acetyltransferase, MOF (males absent on the first), the anti-HSP70 treatments failed to upregulate Bcl-2. This study describes a heretofore unknown, to our knowledge, pathogenic consequence of autoimmunity in which autoantibodies affect the epigenetic regulation of fibroblast apoptosis. In addition to IPF, this autoimmune process could also have relevance in other immunological syndromes characterized by anti-HSP70 autoimmunity. These findings lend credence to the importance of autoimmunity in IPF and illustrate pathways that could be targeted in innovative therapies for this morbid, medically refractory lung disease.

Figure 1.. Lung fibroblast expression of HSP70.

A) Primary normal lung fibroblasts were cultured to 70% confluence in complete media (10% serum) and then in serum-free media for 24 hours before confocal microscopy staining with commercial polyclonal rabbit anti-human HSP70 and then anti-rabbit IgG f(ab’)2-PE (red) as the secondary antibody. Blue = DAPI. B) Similarly cultured fibroblasts stained with DAPI and secondary antibody only. C) Fibroblasts stained with anti-HSP70 autoantibodies (HSP70-auto-Abs) isolated from IPF patient plasma, followed by mouse anti-human IgG-FITC (green) and DAPI. D) Fibroblasts treated with control normal human IgG and then mouse anti-human IgG-FITC (and DAPI) show specificity of the preparations in C). E) Representative histogram showing staining of serum-starved IPF lung fibroblasts by anti-HSP70-PE vs. secondary antibody control. F) Surface expression of HSP70 detected by flow cytometry (as % positive cells – isotype control) was greater on primary IPF lung fibroblasts, and was more augmented by serum starvation, compared to primary normal lung fibroblasts (n = 10 in each group). G) Expression of HSP70 mRNA was also greater in primary IPF lung fibroblasts, compared to normal preparations, as measured by real-time RT-PCR. Cells in both these populations were serum starved for 24 hours prior to harvest. Results are an average of four normal IPF and four IPF primary lung fibroblasts, repeated three times each. *p<0.05, IPF compared to non-IPF cells.

Figure 2.. Apoptosis in fibroblasts treated with anti-HSP70 autoantibodies.

A) Serum-depleted, normal primary lung fibroblasts treated with 24 hours of anti-HSP70 auto-Abs were resistant to apoptosis induction by one hour incubation in staurosporine (STP), as assessed by Annexin-V assay. NI IgG = pooled normal IgG; anti-HSP70 = anti-HSP70-auto-Abs isolated from IPF patients. Both antibody preparations were used at 1 μg/ml. B) Primary normal lung fibroblasts were prepared similar to A, the cells were collected and subjected to western blots with cleaved caspase-3, caspase 3 and β-actin, with β-actin as loading control. C) Primary normal lung fibroblasts were prepared as in B, and cell lysate were collected by measuring the activity of cleaved caspase-3. The results are averages of at least three independent experiments, with n=3 in each, using anti-HSP70-auto Abs from three different IPF patients.

Figure 3.. Expression of Bcl-2 in lung fibroblasts after antibody treatments.

A) Primary human lung fibroblasts incubated with HSP70-auto-Abs from IPF patients showed a greater time-dependent increase of Bcl-2 mRNA, compared to identical preparations treated instead with normal IgG (NI IgG). Real-Time RT-PCR quantifications of Bcl-2 were based on the 2^−ΔΔCt method, normalized to β-actin. Results are averages of at least 3 independent experiments. B) and C) Bcl-2 protein levels in lysates of fibroblasts similarly treated also showed increases among the cells incubated with HSP70-auto-Abs, compared to control IgG preparations by Western blots (B), and quantifications of these blots by scanning densitometry (C). Data are representative of one lung primary lung fibroblasts from donor responses to one IPF patient anti-HSP70 auto-Ab. Similar results were obtained in at least three different donor derived primary lung fibroblasts, with anti-HSP70-auto-Abs isolated from three different IPF patients (see supplemental Figure 2 A). *p<0.05, HSP70-auto-Ab treated 24h vs NI IgG treated 24h, n=3.

Figure 4.. Bcl-2 expression in lung fibroblast in response to HSP70-auto-Abs is mediated by epigenetic regulation.

A) Following 24-hour treatment with either IgG or anti-HSP70-auto-Abs, primary lung fibroblast DNA was immunoprecipitated with anti-H4K16Ac antibody, prior to quantitative Bcl-2 PCR. Results were calculated by 2^−ΔΔCt method and normalized to input Bcl-2 DNA levels and expressed as fold-changes relative to results in IgG-treated control cells. Results are averages of at least three independent experiments. *: p<0.05 in comparisons to IgG controls. B) Lung fibroblasts were transfected with either nontargeting (NT) or MOF siRNA, and treated with either control IgG or anti-HSP70-auto-Abs prior to Western blots with Mof, Bcl-2 and β-actin (for loading control). C) Bcl-2 mRNA levels in siRNA NT or siRNA Mof transfected fibroblasts that were treated with HSP70-auto-Abs was quantitated by RT-PCR by 2^−ΔΔCt method ratio to β-actin. *p <0.05 compared to siRNA NT. D) Cross-linked DNA from siRNA NT and siRNA Mof transfected fibroblasts after 24 hr treatment with anti-HSP70 autoantibody was immunoprecipitated with H4K16Ac antibody prior to quantitative PCR. Results were normalized to input DNA and expressed as fold changes relative to siRNA NT cells. *p <0.05 vs siRNA NT.