Arsenic trioxide inhibits transforming growth factor-β1-induced fibroblast to myofibroblast differentiation in vitro and bleomycin induced lung fibrosis in vivo

Abstract

Background: Idiopathic pulmonary fibrosis (IPF) is a progressive disease of insidious onset, and is responsible for up to 30,000 deaths per year in the U.S. Excessive production of extracellular matrix by myofibroblasts has been shown to be an important pathological feature in IPF. TGF-β1 is expressed in fibrotic lung and promotes fibroblast to myofibroblast differentiation (FMD) as well as matrix deposition.

Methods: To identify the mechanism of Arsenic trioxide's (ATO)'s anti-fibrotic effect in vitro, normal human lung fibroblasts (NHLFs) were treated with ATO for 24 hours and were then exposed to TGF-β1 (1 ng/ml) before harvesting at multiple time points. To investigate whether ATO is able to alleviate lung fibrosis in vivo, C57BL/6 mice were administered bleomycin by oropharyngeal aspiration and ATO was injected intraperitoneally daily for 14 days. Quantitative real-time PCR, western blotting, and immunofluorescent staining were used to assess the expression of fibrotic markers such as α-smooth muscle actin (α-SMA) and α-1 type I collagen.

Results: Treatment of NHLFs with ATO at very low concentrations (10-20nM) inhibits TGF-β1-induced α-smooth muscle actin (α-SMA) and α-1 type I collagen mRNA and protein expression. ATO also diminishes the TGF-β1-mediated contractile response in NHLFs. ATO's down-regulation of profibrotic molecules is associated with inhibition of Akt, as well as Smad2/Smad3 phosphorylation. TGF-β1-induced H2O2 and NOX-4 mRNA expression are also blocked by ATO. ATO-mediated reduction in Smad3 phosphorylation correlated with a reduction of promyelocytic leukemia (PML) nuclear bodies and PML protein expression. PML-/- mouse embryonic fibroblasts (MEFs) showed decreased fibronectin and PAI-1 expression in response to TGF-β1. Daily intraperitoneal injection of ATO (1 mg/kg) in C57BL/6 mice inhibits bleomycin induced lung α-1 type I collagen mRNA and protein expression.

Conclusions: In summary, these data indicate that low concentrations of ATO inhibit TGF-β1-induced fibroblast to myofibroblast differentiation and decreases bleomycin induced pulmonary fibrosis.

Figure 1

ATO inhibits TGF-β1-induced α-SMA and collagen expression. (A-D) Normal human lung fibroblasts (NHLFs) were serum starved overnight and treated with Arsenic trioxide (ATO) for 24 hrs, then treated with TGF-β1 (1 ng/ml) for 24 hrs. α-SMA, Collagen-1, PAI-1 and CTGF mRNA expression levels were evaluated by TGF-β1 and ATO inhibited TGF-β1’s effect. (E) TGF-β1-induced α-SMA and collagen-1 protein expression was blocked by ATO in NHLFs. (F) Immunofluorescent staining showed TGF-β1-induced α-SMA fiber formation was reduced by ATO. (G) An MTT assay showing that ATO does not affect NHLF viability. RT-PCR Data represent results from three independent experiments with duplicate repeats. Western blot and immunofluoresent staining data represent consistent trend in three independent repeats with the best image quality. *P value < 0.05, **P value <0.01, ***P value < 0.001.

Figure 2

ATO inhibits TGF-β-induced fibroblast contractile activity in rat tail type-I collagen gel. (A) NHLFs were cultured in rat tail type I collagen gels (2 mg/ml) and pre-treated with ATO for 24 hrs. Cells were exposed to TGF-β1 (1 ng/ml) for another 48 hrs. TGF-β1treatment decreased the size of the gel while ATO pre-treatment blocked the reduction in gel size. (B) A summation of the percentage of gel surface area in each well of Figure 2A to the control gel surface area. (C) NHLFs were cultured in rat tail type I collagen gels and treated as described above. Gels were digested using type-I collagenase (5 mg/ml) and cells were harvested in SDS loading buffer for western blot. α-SMA protein expression was induced by TGF-β1 and ATO abrogated the induction. Data represent results from three independent experiments with triplicate repeats. *P value < 0.05, **P value <0.01, ***P value < 0.001.

Figure 3

ATO inhibits Smad2/Smad3 and Akt phosphorylation but does not affect p38 phosphorylation. (A-B) NHLFs were pre-treated with ATO for 24 hrs. Then cells were treated with TGF-β1 (1 ng/ml) for 30 mins. TGF-β1 did not affect total Smad2 and Smad3 expression, but did increase their phosphorylation. TGF-β1 induced Smad2 and Smad3 phosphorylation were blocked by ATO. (C) NHLFs were pre-treated with ATO for 24 hrs, and then washed 5 times with PBS to remove ATO from the media. Thereafter cells were exposed to TGF-β1 (1 ng/ml) for 30 mins. The results indicate that the effects of ATO on TGF-β1 induced Smad2 and Smad3 phosphorylation are not due to disruption of the of the TGF-β1 ligand. (D) NHLFs were pre-treated with ATO for 24 hrs., and then treated with TGF-β1 (1 ng/ml) for 12 hrs. Akt phosphorylation was induced by TGF-β1 and this induction was diminished by ATO. (E) NHLFs were pre-treated with ATO for 24 hrs, then treated with TGF-β1 (1 ng/ml) for 30 mins. ATO increased TGF-β1 induced p38 phosphorylation. (F) TGF-β1 induces H₂O₂ in NHLFs. ATO at the indicated concentrations did not induce H₂O₂, but did diminish TGF-β1-induced H₂O₂ production. (G) NHLFs were pre-treated with ATO for 24 hrs, then treated with TGF-β1 (1 ng/ml) with or without H₂O₂ for 30 mins. Smad3 phosphorylation was up-regulated by H₂O_2.(H) TGF-β1 induced NOX-4 mRNA expression in NHLFs, whereas ATO at the indicated concentrations inhibited TGF-β1-induced NOX-4 mRNA expression. Western blot data represents consistent trend in three independent repeats with the best image quality. RT-PCR data represent results from three independent experiments with duplicate repeats. *P value < 0.05, **P value <0.01, ***P value < 0.001.

Figure 4

TGF-β1 induced PAI-1, fibronectin and Smad3 phosphorylation are impaired in PML -/- MEFs. (A) NHLFs were treated with ATO (20nM) for 24 hrs. Cytoplasmic and nuclear compartments were separated and extracted. PML protein expression was decreased in both the cytoplasm and nucleus in response to ATO treatment. (B) Immunofluorescent staining shows that PML bodies in NHLFs were decreased in response to ATO treatment. (C) Wild type and PML-/- mouse embryonic fibroblasts (MEFs) were treated with TGF-β1 (1 ng/ml) for 24 hrs. Fibronectin-EDA and PAI-1 protein expression was induced by TGF-β1 in WT-MEFs but not PML -/- MEFs. (D) Wild-type and PML -/- MEFs were treated with TGF-β1 (1 ng/ml) and cells were harvested at the indicated time points. Smad3 phosphorylation was induced in WT-MEFs and the level of phosphorylation was markedly decreased in PML -/- MEFs. Western blot data represents consistent trend in three independent repeats with the best image quality.

Figure 5

Daily injection of ATO reduces PML body in C57BL/6 mouse lungs. (A) Bleomycin (2units/kg) and ATO (1 mg/kg) were administered to C57BL/6 mice (n = 7) as indicated. Bleomycin was administered by oropharyngeal aspiration. ATO was administered daily by intra-peritoneal injection. (B) Mice were sacrificed 14 days after bleomycin administration. Bleomycin induced PML body formation, while ATO disrupted their formation, in mouse lungs. (C) Densitometric quantification of PML body intensity for Figure 5B. *P value < 0.05, **P value <0.01, ***P value < 0.001.

Figure 6

ATO inhibits bleomycin induced lung fibrosis phenotypes in C57BL/6 mice. (A, B) C57BL/6 mice were exposed to bleomycin and ATO as described in Figure 5. Type-1 collagen and α-SMA mRNA expression were induced by bleomycin and this effect was diminished by ATO. (C, D) Increased Type-I collagen protein expression in response to bleomycin treatment was blocked by ATO, and was statistically significant. (E) Representative histology showing that bleomycin induced inflammation and collagen deposition in mouse lungs, and these features were reduced by ATO treatment. (F) Total collagen expression on lung sections was evaluated using the Aperio software scanning system. Bleomycin induced total collagen expression and ATO treatment decreased this effect.

Figure 7

Depiction of in vitro experiments summarizing how ATO regulates TGF-β1 signaling. ATO blocks TGF-β1 induced fibroblast to myofibroblast differentiation by suppressing Smad2, Smad3 and Akt phosphorylation. ATO also blocks TGF-β1 induced NOX-4 mRNA expression and therefore abrogates TGF-β1-induced H₂O₂. Furthermore, ATO degrades PML bodies and suppresses PML protein expression, which regulates Smad3 phosphorylation and may play an important role in the pathogenesis of pulmonary fibrosis.