Rivaroxaban suppresses factor Xa–Driven PAR-2–ERK inflammatory, catabolic, osteoclastogenic, and mitochondrial dysfunction signaling in osteoarthritis-relevant chondrocytes: a drug-repurposing strategy for joint inflammation

Background Osteoarthritis (OA) is increasingly recognised as an inflammation-driven whole-joint disease in which cartilage catabolism, osteochondral remodelling, and mitochondrial dysfunction drive structural progression. Factor Xa (FXa), beyond haemostasis, is a signalling protease that activates protease-activated receptor-2 (PAR-2) and downstream MAPK-dependent inflammation; whether it directly drives this programme in OA-relevant chondrocytes, and whether rivaroxaban can intercept it, is undefined. Objective To determine whether rivaroxaban attenuates FXa-induced PAR-2/ERK inflammatory signalling and downstream cytokine, catabolic, osteoclastogenic, and mitochondrial stress responses in bone marrow-derived mesenchymal stem cell (BMSC)-derived chondrocytes. Methods Human BMSC-derived chondrocytes, phenotypically validated by collagen type II immunostaining and Alcian blue and toluidine blue staining, were exposed to non-cytotoxic FXa and rivaroxaban concentrations (MTT), with rivaroxaban-mediated FXa inhibition confirmed by an S-2765 chromogenic amidolytic assay. FXa-stimulated chondrocytes ± rivaroxaban were then profiled for PAR-2, ERK1/2, p-ERK1/2, DUSP6, TNF-α, IL-1β, MCP-1, SOX4, ADAMTS5, RANK/RANKL, and mitochondrial function by immunoblotting, RT-qPCR, ELISA, flow cytometry, and Rhodamine 123/Janus Green B staining; PAR-2 dependency was interrogated by siRNA-mediated F2RL1 knockdown. Results FXa elicited a coordinated phenotype—PAR-2 upregulation, ERK1/2 activation, increased TNF-α, IL-1β, and MCP-1, induction of SOX4 and ADAMTS5, enhanced RANK/RANKL, and mitochondrial dysfunction. Without compromising viability, rivaroxaban suppressed FXa amidolytic activity and reduced PAR-2 (transcript, total and cell-surface protein), ERK1/2 phosphorylation, p-ERK/ERK output, and DUSP6, denoting attenuated sustained ERK flux; this translated into lower cytokine secretion, diminished catabolic and osteoclastogenic effector expression, and preserved mitochondrial membrane potential and redox-associated staining. PAR-2 knockdown reduced, but did not abolish, the FXa response, indicating PAR-2-dominant but not exclusively PAR-2-dependent signalling. Conclusion Rivaroxaban suppresses a coherent FXa-responsive inflammatory network in OA-relevant chondrocytes via the PAR-2–ERK axis and its downstream cytokine, catabolic, osteoclastogenic, and mitochondrial sequelae. These findings extend direct FXa inhibition beyond anticoagulation and support rivaroxaban as a mechanistically plausible repurposing candidate for OA-associated inflammatory pathology, pending validation in primary OA chondrocytes, osteochondral explants, and in vivo models.