Exploration of key genes associated with oxidative stress in polycystic ovary syndrome and experimental validation

Abstract

Introduction: The current study demonstrated that oxidative stress (OS) is closely related to the pathogenesis of polycystic ovary syndrome (PCOS). However, there are numerous factors that lead to OS, therefore, identifying the key genes associated with PCOS that contribute to OS is crucial for elucidating the pathogenesis of PCOS and selecting appropriate treatment strategies. Methods: Four datasets (GSE95728, GSE106724, GSE138572, and GSE145296) were downloaded from the gene expression omnibus (GEO) database. GSE95728 and GSE106724 were combined to identify differentially expressed genes (DEGs) in PCOS. weighted gene correlation network analysis (WGCNA) was used to screen key module genes associated with PCOS. Differentially expressed OS related genes (DE-OSRGs) associated with PCOS were obtained by overlapping DEGs, key module genes, and OSRGs. Subsequently, the optimal machine model was obtained to identify key genes by comparing the performance of the random forest model (RF), support vector machine model (SVM), and generalized linear model (GLM). The molecular networks were constructed to reveal the non-coding regulatory mechanisms of key genes based on GSE138572 and GSE145296. The Drug-Gene Interaction Database (DGIdb) was used to predict the potential therapeutic agents of key genes for PCOS. Finally, the expression of key OSRGs was validated by RT-PCR. Results: In this study, 8 DE-OSRGs were identified. Based on the residuals and root mean square error of the three models, the best performance of RF was derived and 7 key genes (TNFSF10, CBL, IFNG, CP, CASP8, APOA1, and DDIT3) were identified. The GSEA enrichment analysis revealed that TNFSF10, CP, DDIT3, and INFG are all enriched in the NOD-like receptor signaling pathway and natural killer cell-mediated cytotoxicity pathways. The molecular regulatory network uncovered that both TNFSF10 and CBL are regulated by non-coding RNAs. Additionally, 70 potential therapeutic drugs for PCOS were predicted, with ibuprofen associated with DDIT3 and IFNG. RT-qPCR validation confirmed the expression trends of key genes IFNG, DDIT3, and APOA1 were consistent with the dataset, and the observed differences were statistically significant (P < 0.05). Conclusion: The identification of seven key genes and molecular regulatory networks through bioinformatics analysis is of great significance for exploring the pathogenesis and therapeutic strategies of PCOS.