Activin B Promotes Hepatic Fibrogenesis

Abstract

Liver fibrosis is a common consequence of various chronic liver diseases. Although transforming growth factor β 1 (TGFβ1) expression is known to be associated with liver fibrosis, the reduced clinical efficacy of TGFβ1 inhibition or the inefficiency to completely prevent liver fibrosis in mice with liver-specific knockout of TGF receptor II suggests that other factors can mediate liver fibrogenesis. As a TGFβ superfamily ligand, activin A signaling modulates liver injury by prohibiting hepatocyte proliferation, mediating hepatocyte apoptosis, promoting Kupffer cell activation, and inducing hepatic stellate cell (HSC) activation in vitro. However, the mechanism of action and in vivo functional significance of activin A in liver fibrosis models remain uncertain. Moreover, whether activin B, another ligand structurally related to activin A, is involved in liver fibrogenesis is not yet known. This study aimed to investigate the role of activin A and B in liver fibrosis initiation and progression. The levels of hepatic and circulating activin B and A were analyzed in patients with various chronic liver diseases, including end-stage liver diseases (ESLD), non-alcoholic steatohepatitis (NASH), and alcoholic liver disease (ALD). In addition, their levels were measured in mouse carbon tetrachloride (CCl4), bile duct ligation (BDL), and ALD liver injury models. Mouse primary hepatocytes, RAW264.7 cells, and LX-2 cells were used as in vitro models of hepatocytes, macrophages, and HSCs, respectively. The specificity and potency of anti-activin B monoclonal antibody (mAb) and anti-activin A mAb were evaluated using Smad2/3 luciferase assay. Activin A, activin B, or their combination were immunologically inactivated by the neutralizing mAbs in mice with progressive or established liver fibrosis induced by CCl4 or with developing cholestatic liver fibrosis induced by BDL surgery. In patients with ESLD, NASH, and ALD, increases in hepatic and circulating activin B, but not activin A, were associated with liver fibrosis, irrespective of etiology. In mice with CCl4-, BDL-, or alcohol-induced liver injury, activin B was persistently elevated in the liver and circulation, whereas activin A showed only transient increases. Activin B was expressed and secreted mainly by the hepatocytes and other cells, including cholangiocytes, activated HSCs, and immune cells. Exogenous administration of activin B promoted hepatocyte injury, activated macrophages to release cytokines, and induced a pro-fibrotic expression profile and septa formation in HSCs. Co-treatment of activin A and B interdependently activated the chemokine (C-X-C motif) ligand 1 (CXCL1)/inducible nitric oxide synthase (iNOS) pathway in macrophages and additively upregulated connective tissue growth factor expression in HSCs. Activin B and A had redundant, unique, and interactive effects on the transcripts related to HSC activation. The neutralization of activin B attenuated the development of liver fibrosis and improved liver function in mice with CCl4- or BDL-induced liver fibrosis and largely reversed the already established liver fibrosis in the CCl4 mouse model. These effects were improved by the administration of additional anti-activin A antibody. Combination of both antibodies also inhibited hepatic and circulating inflammatory cytokine production in the BDL mouse model. In conclusion, activin B is a potential circulating biomarker and potent promotor of liver fibrosis. Its levels in the liver and circulation increase significantly in both acute and chronic states of liver injury. Activin B might additively or interdependently cooperate with activin A, which directly acts on multiple liver cell populations during liver injury and fibrosis, as the combination of both proteins increases pro-inflammatory and pro-fibrotic responses in vitro. In addition, the neutralization of both activin A and activin B in vivo enhances the preventive and reversible effects of liver injury and fibrosis compared to that when activin B alone is neutralized. Our data reveal a novel target of liver fibrosis and the mechanism of activin B-mediated initiation of this process by damaging hepatocytes and activating macrophages and HSCs. Our findings show that activin B promotes hepatic fibrogenesis, and that targeting of activin B has anti-inflammatory and anti-fibrotic effects, which ameliorate liver injury by preventing or regressing liver fibrosis. Antagonizing either activin B alone or in combination with activin A prevents and regresses liver fibrosis in multiple animal studies, paving way for future clinical studies.