Key nodes governing cell death fate in osteoarthritis: GPX4, caspase-8 and ESCRT-III

Osteoarthritis (OA) is increasingly supported by evidence as a whole-joint disorder, in which chondrocyte injury drives whole-joint degeneration rather than merely accompanying cartilage breakdown. Under sustained mechanical, metabolic, and inflammatory stress, chondrocyte cytoprotective systems become progressively exhausted. These protective systems include the system Xc − /GSH/GPX4 axis and parallel anti-ferroptotic defenses, iron storage and export, mitochondrial quality control, and autophagy. This review focuses on the staged failure of these buffering systems during OA progression, from adaptation and compensation to threshold crossing, rather than presenting a catalogue of independent cell death pathways. Beyond this threshold, the predominant mode of chondrocyte death shifts from immunologically silent apoptosis to membrane-lytic inflammatory forms including ferroptosis, pyroptosis, and necroptosis, which are frequently co-activated within the same tissue. Two interrelated propositions are put forward. First, these cell death modalities share a limited set of upstream perturbations: accumulation of reactive oxygen species, NF-κB activation, dysregulated iron homeostasis, and mitochondrial dysfunction. Second, downstream cell death execution is determined by a small set of molecular switches. These switches include caspase-8 catalytic activity, ESCRT-III-dependent membrane repair capacity, and GPX4 abundance. CHMP5, a key component of membrane repair, is consistently depleted in OA cartilage. OA-specific evidence for these switches is now emerging, including ZBP1-dependent PANoptotic signalling in cartilage. Approximately one hundred regulators associated with regulated cell death (RCD) reported in OA joint tissue are catalogued and annotated by direction of change, tissue compartment, OA-specific experimental support, and druggability. A metabolic–epigenetic circuit is proposed to reinforce death-prone chondrocyte states. This circuit involves histone lactylation, m 6 A modification of GPX4 and ATG7, and α-ketoglutarate depletion. From this perspective, the limited efficacy of single-target therapies reflects the architecture of the shared network. Stage- and compartment-matched interventions represent a more tractable therapeutic strategy.