Therapeutic estrogens address multiple facets of non-alcoholic steatohepatitis (NASH) pathophysiology. Their potential, however, is limited by undesirable side effects caused by persistent estrogen receptor a (ERa) activation. Estrogen receptor b (ERb) is functionally unique, capable of inducing estrogen’s benefits while limiting ERa-mediated side effects. We hypothesized that OSU-ERb-12, a novel carborane selective estrogen receptor modulator with affinity for ERb, would demonstrate therapeutic benefit in preclinical NASH models at doses absent of ERa-mediated effects (ERb-selective). Our objectives were first to describe OSU-ERb-12’s drug-like properties by characterizing its ERb-selectivity, pursuing early but critical preclinical safety studies, describing its pharmacokinetic properties, and finally by identifying ERb-selective doses to apply to future studies. We then assessed OSU-ERb-12’s therapeutic potential in two pre-clinical models of relevant NASH pathophysiology while simultaneously gaining mechanistic insight into drug efficacy.
We applied several in vitro techniques to demonstrate OSU-ERb-12’s drug-like properties. We established its ERb-selectivity first through a competitive radioligand-binding assay to demonstrate ligand affinity for ERb, then through a luciferase reporter assay to assess its preferential activation of ERb-mediated transcription. In both experiments, OSU-ERb-12 demonstrated selective properties comparable to LY500307, another ERb-selective agonist. We then put OSU-ERb-12 through nuclear hormone receptor and CYP-inhibition assays to identify potential off-target effects and drug-drug interactions, respectively; none were found suggesting compound safety. Multi-species microsomal stability assays and in vivo pharmacokinetic studies not only demonstrated remarkable OSU-ERb-12 stability in most species, but that the drug has superior pharmacokinetic properties to LY500307. Finally, to establish in vivo ERb-selective doses, we utilized uterotrophism in juvenile female mice to identify that ERβ-selective doses are ≤ 30 mg/kg.
We next tested in vivo efficacy in two murine models: a chronic hepatotoxin induced model of fibrosis, and a lipotoxicity driven streptozotocin and high-fat diet (STAM) model. To induce hepatic fibrosis, we administered CCl4 and intervened with ERb-selective agonists, OSU-ERb-12 and LY500307. Both compounds improved hepatic fibrosis, but only OSU-ERb-12 did so without inducing ERa-associated side effects. RNA microarray highlighted drug-mediated downregulation of genes critical to fibrosis induction including: TGFb, TGFbR1, Notch1, TIMP2, THBS2, Reln, and several collagen species. We then tested efficacy in the STAM NASH model, intervening with ERb-selective and non-selective OSU-ERb-12 doses. At both doses, OSU-ERb-12 improved NASH by reducing liver triglycerides and liver damage. Other outcomes demonstrated dose-dependent corrections with the non-selective dose generally more efficacious. Total RNA-Seq on livers collected at sacrifice identified drug-mediated improvements to STAM disruption of oxidative phosphorylation and lipid metabolism pathways. Of particular interest were statistically significant reductions in expression of ACOT1, ACOT2 and PDK4, genes critical to lipid and glucose metabolism and well documented targets of PPARa signaling, another promising therapeutic target in NASH. Another gene modulated by our drug was CCL2, an inflammatory gene also considered a promising therapeutic target. These studies highlighted OSU-ERb-12’s promise in NASH through its ability to ameliorate both causative lipotoxicity and subsequent fibrosis.
In a separate approach, we investigated an alternate hypothesis. While female sex-hormones, estrogens, demonstrate a protective effect against a number of chronic liver diseases including NASH, fibrosis, and hepatocellular carcinoma (HCC), there is considerable evidence implicating male sex-hormones, androgens, in disease promotion, especially HCC. There is consistent data in preclinical rodent models that early anti-androgen interventions (castration, genetic manipulation) delay hepatocarcinogenesis. The successes observed in the laboratory do not transition to the clinic, however, as pharmacologic anti-androgen therapies fail in late-stage HCC. We interpreted these discordant findings to reflect a mismatch in the timing of anti-androgen intervention. We therefore hypothesized that pharmacologic inhibition of the androgen signalling-axis may require early and prolonged intervention in hepatocarcinogenesis.
We tested this hypothesis in vivo in a rodent diethylnitrosamine (DEN) model of hepatocellular carcinogenesis. Our objectives were to evaluate the efficacy of prophylactic androgen receptor (AR) antagonism to prevent liver tumorigenesis, generate mechanistic hypotheses as to how early AR antagonism exerts efficacy, and characterize AR expression throughout liver carcinogenesis. We demonstrated that pharmacologic AR antagonism with enzalutamide (ENZ) indeed inhibits hepatocellular carcinogenesis. This efficacy, however, was in part mediated through a model-specific means as ENZ suppressed CYP2E1 and inhibited formation of carcinogenic DNA ethyl-adducts. We then characterized AR protein expression throughout carcinogenesis highlighting AR’s presence in portal fibroblasts and leukocytes, but absence from the hepatocytes. This suggests that that hepatocyte-autonomous AR-signalling is not essential for DEN-induced carcinogenesis and that the promotional action of AR may be through alternate routes, such as hepatic immune cells. We then found that ablating androgen signalling by surgical castration reduced pre-carcinogen Kupffer cell populations but did not alter DEN-mediated immune cell recruitment nor AR expression. In this study we describe a novel mechanism for how AR inhibition mitigates DEN carcinogenesis through CYP2E1 suppression, characterize where the hepatic AR is expressed throughout all stages of carcinogenesis, and provide novel insight into androgen’s modulation of hepatic immune cell populations in the carcinogen challenged liver.
In conclusion, we provided novel insight into how sex hormone signaling contributes to, and can be manipulated in NASH and HCC. We characterized and established a novel ERb-selective agonist, OSU-ERb-12, and demonstrated efficacy in two NASH models at a dose that avoids persistent ERa activity. We also provided promising insight into how ERb-selective agonists achieve efficacy through correcting metabolic and inflammatory pathways disrupted by NASH. In our analysis of AR in HCC, we determined how anti-androgens inhibit tumor production in the DEN model, and clarified its expression throughout DEN carcinogenesis, implicating a role for AR in carcinogen induced inflammation. In these studies, we gain insight into the contributions of androgens, estrogens, and their signaling mechanisms to liver diseases, and provide strong pre-clinical evidence that sex-hormone signaling is a promising and exploitable therapeutic target in NASH and HCC.