Oxidative damage and TGF-β differentially induce lung epithelial cell sonic hedgehog and tenascin-C expression: implications for the regulation of lung remodelling in idiopathic interstitial lung disease

Abstract

Idiopathic interstitial lung diseases (iILDs) are characterized by inflammation, hyperplasia of Type-II alveolar epithelial cells (AECs) and lung remodelling often with progressive fibrosis. It remains unclear which signals initiate iILD and/or maintain the disease processes. Using real-time RT-PCR and immunohistochemistry on archival biopsies of three patterns of iILD (usual interstitial pneumonitis/UIP, non-specific interstitial pneumonitis/NSIP and cryptogenic organizing pneumonia/COP) we investigated whether hedgehog signalling (previously associated with lung damage and repair) was functional and whether the damage associated extracellular matrix protein tenascin-C was present in activated Type-II AECs in all three iILDs. Using tissue culture, protein and mRNA detection we also determined how two signals (oxidative damage and TGF-β) associated with iILD pathogenesis affected Sonic hedgehog (SHH) and tenascin-C production by a Type-II AEC cell line. We report that SHH pathway and tenascin-C mRNA and proteins were found in UIP, NSIP and COP. SHH signalling was most active at sites of immature organizing fibrous tissue (fibroblastic foci) in UIP. In vitro Type-II AECs constitutively secrete SHH but not tenascin-C. Oxidative injury stimulated SHH release whereas TGF-β inhibited it. TGF-β and oxidative damage both upregulated tenascin-C mRNA but only TGF-β induced synthesis and release of a distinct protein isoform. SHH signalling is active in Type-II AECs from three types of ILD and all three express tenascin-C.

Figure 1

GLI1 and tenascin-C in mRNA extracted from paraffin sections. (a) Real-time PCR as evidence of expression of genes in COP, UIP, NSIP and microscopically normal lung sections (three cases of each condition were analysed by determining the mean value per case from three separate 10 μm sections). Data are expressed as means ± SEM relative to GAPDH levels multiplexed in the same samples. (b) Percentage of disease involved tissue on biopsy sections. COP biopsies (n = 7 cases) had significantly less involved tissue per section than UIP (n = 7 cases) or NSIP (n = 3 cases) sections.

Figure 2

Immunohistochemistry of iILDs. Pictures representative of 6 UIP, 6 COP and 3 NSIP cases are shown. (a–c) Tenascin-C staining in fibroblastic foci of UIP (a); in buds of organizing exudate in COP (b); and, focal expression within thickened, fibrotic alveolar walls in NSIP (c). Black bars = 200 μm. (d–o) SHH signalling components in Type-II AECs in iILD. SHH (d–g), PTC1 (h–k) and GLI1 (l–o) staining in fibroblastic foci of UIP (d, h, l), non-fibroblastic foci of UIP (e, i, m), COP (f, j, n) and NSIP (g, k, o). Note the GLI1 positive nuclei in i–o. Black bars = 50 μm.

Figure 3

Nuclear GLI1 is increased in active fibrosis. (a) No statistically significant difference in the percentage of type-II AECs showing nuclear staining for GLI1 was detected between the UIP, COP and NSIP cases studied. (b) There was however a significant difference between the % of type-II AECs expressing nuclear GLI1 overlying fibroblastic foci (FF) compared with those overlying mature fibrous tissue (non-FF) in UIP biopsies (P = 0.0152, Mann–Whitney U-test, n = 6 cases of UIP).

Figure 4

Effects of oxidative damage on A549 cells. (a) RT-PCR representative of seven separate experiments for SHH (211 bp, lane 3), PTC1 (206 bp, lane 4), SMO (140 bp, lane 5), GLI2 (200 bp, lane 6) and GLI1 (244 bp, lane 7). Lane 1= DNA markers, lane 2 = mastermix only. See Table 1 for details of primers. (b–d) Data pooled from three separate experiments. (b) MTT assay showing that H₂O₂ induces damage to A549 cells. (c, d) ELISA estimation of SHH in A549 cell supernatants, dotted line = limit of detection, *P < 0.05, **P > 0.005, compared to values for cells cultured in medium alone, (Student’s t-test); (c) Constitutive secretion of SHH by A549 cells. (d) SHH release induced by H₂O₂. Monolayers of A549 cells in 24-well plates were washed and exposed to different concentrations of H₂O₂ for 90 min, washed and fresh medium added. Supernatants collected after 6, 24, 48 and 72 h. (e) Tenascin-C protein isoforms in cell lysates after 72 h culture in medium only (lanes 1 and 2) or following exposure to 2.5 mM H₂O₂ (lanes 3–5, three separate experiments); equal protein loading was confirmed by probing with anti-actin antibody. Gel representative of five separate experiments. (f) Real-time PCR for tenascin-C mRNA after exposure to 2.5 mM H₂O₂. Data pooled from four separate experiments [normalized to medium only control = 1 at each time point], *P = 0.0407, unpaired t-test with Welch’s correction.

Figure 5

Effect of TGF-β exposure on tenascin-C and SHH. (a) Mean ± SEM fold increase (normalized to a value of 1 for untreated cells) in tenascin-C mRNA in A549 extracts after 48 h exposure to 10 ng/ml acid-activated TGF-β compared to cells in acid control medium only. All samples were normalized to a value of 1 for untreated cells in medium only; data pooled from four separate experiments, normalized to 18s (similar results observed with GAPDH normalization). **P =0.0043, unpaired t-test with Welch’s correction. (b) Western blot (representative of four separate experiments) for tenascin-C protein and β-actin loading controls in cell lysates (25 μg/lane) and tenascin-C protein in supernatants (100 μg/lane) of A549 cells cultured for 72 h in TGF-β (lanes 1 and 3) or control medium (lanes 2 and 4) showing induction of a lower molecular weight isoform released from TGF-β treated cells. (c) SHH protein in supernatants of A549 cells (data pooled from four separate experiments) after 72 h culture with TGF-β. Results are normalized to the total amount of protein in the supernatants. *P =0.0253, unpaired t-test with Welch’s correction. (d) Western blot (representative of four separate experiments) of SHH protein and β-actin loading controls in A549 cell lysates after 72 h culture in medium only (lane 1), in medium with acid diluent control (lane 2) or with TGF-β (lane 3). (e) Mean ± SEM fold increase (normalized to a value of 1 for untreated cells) in SHH mRNA after 48 h exposure to 10 ng/ml acid-activated TGF-β compared to cells in acid control medium only.