Selective RAR gamma antagonist LY2813631 protects against retinoid induced cartilage degradation in preclinical models of arthritis

Abstract

Purpose: It has been shown that the natural ligand for the Retinoic Acid Receptors (RARs), All Trans Retinoic Acid (ATRA), is deleterious to articular cartilage health and is associated with the breakdown of cartilage in osteoarthritis. Additionally, it has been shown that retinoid levels are increased in the synovial fluid of OA patients. Both natural and synthetic retinoids (RAR agonists) are catabolic to cartilage, block early chondrogenesis and promote chondrocyte hypertrophy via RAR signaling through three RAR subtypes (RAR alpha, beta and gamma). It has been postulated that RAR antagonists may prevent or reverse retinoid-mediated cartilage destruction and a RAR pan antagonist was previously shown to improve clinical scores in the collagen-induced arthritis (CIA) model, albeit with unacceptable adverse effects on testes. We have postulated that the primary beneficial joint effects of RAR antagonists are associated with RARgamma, while the adverse effects on testes are associated with RARalpha. Thus, we have identified a highly selective RARgamma antagonist (LY2813631) to test this hypothesis. Methods: The RAR antagonist LY2813631 demonstrated in vitro selective affinity for RARgamma in a SPA-based binding assay using full length RAR alpha, beta and gamma protein and the synthetic pan agonist 3H-TTNPB. Functional antagonism and selectivity was demonstrated using a GAL4- RAR alpha, beta and gamma and Gal4 response element/Luciferase constructs, co-transfected into HEK 293 cells. Functional activity in chondrocytes was demonstrated using agonist TTNPB and primary bovine chondrocytes, looking at the ability of LY2813631 to regulate OA-relevant genes, such as MMP13, ADAMTS-5 and Type 2 Collagen. In vivo studies utilized Lewis rats to show reversal of RAR gamma agonist induced changes in OA relevant genes in the articular cartilage and reduction in OA-related neoepitopes in the synovial fluid. The collagen induced arthritis (CIA) model was used to determine joint efficacy in rats. Results: In vitro, LY2813631 binds RARgamma (Ki = 0.74 nM) with significantly higher affinity than RARalpha (Ki = 400 nM) and RAR beta (Ki = 25 nM) and shows selective functional antagonism at RAR gamma (Kb = 42 nM) in HEK 293 cells, compared to RARalpha (Kb = 2010 nM) and RAR beta (Kb = 359 nM). Additionally, LY2813631 normalized RAR agonist-induced increases in the catabolic enzyme ADAMTS-5 in primary bovine chondrocytes. In vivo, a selective RARgamma agonist was shown to increase mRNA levels of ADAMTS-5 and decrease mRNA levels of type 2 collagen in the articular cartilage of rats after 3 days of oral dosing. The animals also showed a 5-fold increase in the type 2 collagen neoepitopes 9A4 and CTX-II in synovial fluid. These effects could be blocked by co-dosing RARgamma antagonist LY2813631, demonstrating specific RARgamma-mediated target engagement in the joint space. LY2813631 also improved ankle and knee histopathology scores in the rat CIA model at doses that did not cause testicular degeneration. Conclusions: Our findings support a role for endogenous retinoids in the destruction of articular cartilage in OA and suggest that these effects are primarily mediated through RARgamma signaling. We have also shown that the beneficial effects of RARgamma antagonist LY2813631 can be demonstrated without concomitant adverse effects on the testes, supporting RARgamma as a potentially safe target for disease modification in arthritic diseases.