Bioinformatic identification of THY1 as a target in osteoarthritis and its effects on chondrocyte senescence and apoptosis.

Osteoarthritis (OA) is the most prevalent degenerative joint disease worldwide and imposes a substantial public health burden due to the lack of effective disease-modifying drugs. Exosome-related genes participate in exosome biogenesis, cargo sorting, and secretion, serving as critical regulatory elements in exosome-mediated intercellular communication. However, their systematic characterization in OA and their potential as diagnostic or therapeutic targets remain largely unexplored. We obtained OA and normal control samples from the GEO database. Using exosome-related gene signatures as a filtering strategy, we integrated differential expression analysis, protein-protein interaction network assessment, and machine learning algorithms to screen for key diagnostic genes, and constructed and externally validated diagnostic models and nomograms. Enrichment analysis and immune infiltration assessment were performed to explore biological functions and immune microenvironment features. Cellular experiments were conducted to validate the effects of the prioritized gene THY1 on chondrocyte apoptosis, senescence, mitochondrial function, and cell viability. We successfully identified THY1, FAP, and TIMP4 as independent diagnostic markers for OA and constructed a diagnostic model and nomogram with high applicability, whose reliability was validated in an external validation cohort. Enrichment and immune infiltration analyses preliminarily identified two OA subtypes with distinct immune infiltration profiles. Cellular experiments confirmed that THY1 overexpression promoted chondrocyte senescence and apoptosis under OA-like inflammatory conditions, while impairing mitochondrial function and cell viability. Through bioinformatic screening and in vitro experimental validation, this study reveals the pro-apoptotic and pro-senescent regulatory effects of THY1 on chondrocytes under inflammatory conditions, and establishes a diagnostic model and subtyping framework for OA based on three genes, offering new perspectives for stratified diagnosis and future translational research in OA.