Abstract
BACKGROUND: Osteoporosis (OP) is a common metabolic bone disease characterized by decreased bone mass and increased fracture risk. Recent studies suggest that oxidative phosphorylation (OXPHOS) plays a crucial role in the pathogenesis of OP. This study aims to investigate the differential expression and potential functional roles of OXPHOS-related genes in OP. METHODS: We downloaded gene expression data from two OP-related datasets, GSE56815 and GSE7429, using the GEOquery package. We also collected OXPHOS-related genes from the GeneCards and MsigDB databases. The limma package was used for differential expression analysis of GSE56815, and differentially expressed genes (DEGs) were identified. We intersected these DEGs with OXPHOS-related genes to identify OXPHOS-related differentially expressed genes. Functional enrichment analyses, including Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG), were conducted using the clusterProfiler package. Additionally, we performed gene set enrichment analysis (GSEA). The Mann-Whitney U test analyzed differences in the expression of OXPHOSRDEGs, and their diagnostic potential was assessed by Receiver Operating Characteristic (ROC) curves. Correlation analysis, Protein-Protein Interaction (PPI) network construction, mRNA-miRNA, mRNA-TF interaction network construction, and immune infiltration analysis using CIBERSORT were also conducted. RAW264.7 cells were induced in vitro for 3 days to differentiate towards osteoblasts, and RT-PCR assay was used to verify the differentiation and detect the differential expression of target genes. RESULTS: Our results identified 31 DEGs in GSE56815, with 26 upregulated and 5 downregulated genes. Among these, we identified 10 OXPHOSRDEGs: VPS35, TBC1D2, UBQLN2, SH3GLB2, WWP1, NFKBIA, MFSD10, SLC2 A3, RP2, and ZNF91. GO and KEGG enrichment analyses revealed significant involvement of these genes in mechanisms such as the positive regulation of the protein catabolic process and the endocytosis pathway. ROC analysis demonstrated high diagnostic accuracy for VPS35 (AUC = 0.832) and TBC1D2 (AUC = 0.751). Correlation analysis indicated strong relationships between certain OXPHOSRDEGs. The PPI network highlighted 8 hub genes with significant functional similarity among them. CONCLUSION: This study systematically elucidates the differential expression and potential mechanisms of OXPHOS-related genes in OP through comprehensive bioinformatics analyses. The identified key genes offer valuable insights into the molecular underpinnings of OP and present potential diagnostic biomarkers and therapeutic targets for further investigation.