Identifying Feature Biomarkers Related to Disulfidptosis and Immune Cell Infiltration in Osteoarthritis through Bioinformatics Analysis

Abstract

Abstract Objective: This study aims to identify feature genes, pathways, and infiltrating immune cells related to the metabolic mechanisms of cellular disulfidptosis in osteoarthritis (OA) through bioinformatics analysis. Method: Expression profiles from two Gene Expression Omnibus datasets (GSE207881 and GSE98918) were analyzed to study OA. The datasets included 63 and 12 OA patients, respectively, alongside control subjects. Differential expression analysis was performed after data preprocessing using the ‘limma’ package in R. A co-expression network was constructed using weighted gene co-expression network analysis (WGCNA), and modules highly correlated with disulfidptosis were identified. Gene functionality was explored through Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA). Additionally, the protein–protein interactions (PPI) of the key genes were analyzed using GeneMANIA. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted on the network genes. Furthermore, the diagnostic potential of the selected genes was evaluated, and immune infiltration analysis was performed. Result: A total of 522 differentially expressed genes with statistical significance were identified. GSEA and GSVA analyses revealed multiple significantly enriched signaling pathways, such as ribosome, melanogenesis, and leukocyte transendothelial migration. Nine co-expression modules related to disulfidptosis were screened by WGCNA, of which the blue module (n = 353) showed the strongest positive correlation (r = 0.78, p < 0.05). Intersection analysis further identified 13 hub genes. Through PPI networks and GO and KEGG analyses, these hub genes were found to be significantly enriched in the Notch signaling pathway, and the expression of genes in this pathway was validated. The area under the receiver operating characteristic curve of these hub genes was greater than 0.6, suggesting their potential as biomarkers for OA. Immune cell analysis showed that the genes TUSC3 and SOX5 have a significant relationship with type 17 T helper cells (p < 0.001). An RNA-binding protein (RBP)–mRNA interaction network comprising 68 nodes, 61 RBPs, 7 mRNAs, and 271 edges was constructed using the StarBase online database. Conclusion: This study used bioinformatics techniques to reveal 13 hub genes, complex co-expression networks, and unique immune cell interactions, thereby providing insights into the cellular mechanisms of disulfidptosis in OA. These findings lay the groundwork for future approaches to diagnosis and therapeutic intervention.