Immune Gene INHBA is Associated With Osteoarthritic Cartilage Damage and May Mediate the Temporal Activation of the...

This study aimed to screen and identify osteoarthritis (OA)-related core immune genes, elucidate the mechanism underlying the temporal activation of key signaling pathways mediated by these genes, and screen potential therapeutic compounds, thereby providing a theoretical basis for the precision diagnosis and targeted intervention of OA.

Four sets of OA cartilage transcriptomic datasets with normal controls (NCs) were downloaded from the Gene Expression Omnibus (GEO) database. After batch correction, a training set and two independent validation sets were constructed to screen immune-related differentially expressed genes (Immune-DEGs). A total of 113 combined models were built based on 12 basic machine learning algorithms; the models were optimized via 10-fold cross-validation and evaluated in the validation sets to identify important immune genes. Subsequent functional enrichment analysis was performed to clarify the enriched signaling pathways. Core immune genes were identified by integrating Shapley additive explanations (SHAP) analysis and weighted gene coexpression network analysis (WGCNA), and single-cell pseudotime analysis was used to decipher their dynamic expression patterns during OA progression. Finally, real-time quantitative PCR (RT-qPCR) and western blot (WB) were employed to verify the expression changes of core genes and key pathway proteins in cell models and OA animal cartilage tissues with different lesion severities. Additionally, drug repurposing was conducted using the DSigDB database to screen potential therapeutic compounds, and molecular docking and molecular dynamics simulations were performed to validate the binding ability of clinically commonly used drugs to core targets.

A total of 12 immune-DEGs significantly associated with OA were screened. Through systematic evaluation of multiple models, an exploratory OA classification model based on 10 important immune genes was constructed, showing good classification ability in the limited samples of this study. Functional enrichment analysis revealed that these important immune genes were significantly enriched in the TGF-β signaling pathway. Integration of SHAP and WGCNA analyses identified INHBA as a key hub gene associated with OA, and single-cell pseudotime analysis indicated that INHBA expression was sustainably upregulated with OA progression. In vitro and in vivo experimental validation showed that INHBA mRNA levels in human chondrocytes and INHBA mRNA levels in rat cartilage tissues continuously increased with the aggravation of OA lesions; WB results demonstrated that total TGF-β1 protein exhibited a temporal change of first increasing and then decreasing, while INHBA and downstream phosphorylated p38 MAPK protein levels continuously rose during disease progression, suggesting a temporal activation switch between the TGF-β pathway and the p38 MAPK pathway. Drug repurposing yielded 10 potential therapeutic compounds including progesterone; molecular docking indicated that diclofenac, celecoxib, glucosamine, and chondroitin sulfate could stably bind to core targets, and molecular dynamics simulations further confirmed the stable binding conformation between celecoxib and INHBA protein.

This study successfully constructed an exploratory OA classification model based on 10 important immune genes, demonstrating that the core immune gene INHBA is strongly associated with OA progression and may contribute to cartilage degeneration in OA by mediating the temporal activation of the TGF-β/p38 MAPK pathway.