Abstract:Objective: To mine key genes and pathways of osteoarthritis (OA) based on bioinformatics and validate them using an in vitro chondrocyte model.Methods: This study was based on the GSE269735 dataset from the GEO database, using Weighted Gene Co-expression Network Analysis (WGCNA) to identify key gene modules, combined with differential expression gene (DEG) analysis to screen candidate genes. GO and KEGG pathway analyses were performed on the common genes between DEGs and WGCNA key gene modules, with the STRING database and Cytoscape software used to identify key genes among the DEGs. To validate the robustness of the screening results, the GSE2204879 dataset was used for cross-validation, and a plasmid was constructed with CSF-1 as the candidate gene for transfection interference experiments in OA chondrocytes. Finally, the expression changes of CSF-1 at the cellular level were detected by Western blot, and its effects on cellular function were evaluated to verify its potential role in OA.Results: WGCNA analysis identified multiple modules closely related to OA phenotypes, and key genes were co-screened with DEGs. GO and KEGG analysis showed that the main biological processes involved cytokine activity, chemokine activity, collagen binding, etc. Key genes included CCL20, CSF-1, CXCL11, CXCL3, CXCL5, and CXCR4. In the GSE2204879 dataset, CSF-1 and CXCR4 still showed a consistent expression trend. At the cellular level, CSF-1 and CXCR4 were upregulated in OA model cells. Furthermore, a CSF-1 knockdown plasmid was constructed and transfected into OA chondrocytes; the results showed that CSF-1 knockdown suppressed CXCR4 expression, reduced the expression levels of MMP3 and MMP13 in the cells, and increased the protein expression level of Col2a1, suggesting that CSF-1 may participate in the OA inflammatory response and matrix degradation process by regulating CXCR4. Conclusion: This study, combining bioinformatics and cellular experiments, has screened out multiple key genes closely related to OA. Among them, CSF-1 and CXCR4 show high expression in OA. The experiments indicate that CSF-1 may participate in the OA inflammatory response and matrix degradation process by regulating CXCR4 expression. In summary, CSF-1 may be a key pathogenic factor and potential therapeutic target for OA.