Mitochondrial metabolism-related candidate genes in major depressive disorder: integrative bioinformatics analysis and RT-qPCR assessment

Major depressive disorder (MDD) is a prevalent and complex mental disorder, with mitochondrial metabolism implicated in its pathophysiology. However, specific candidate genes linking mitochondrial metabolism to MDD remain unidentified. This study aimed to discover and preliminarily assess candidate genes through integrated bioinformatics and experimental approaches. Differentially expressed genes (DEGs) were identified using the GEO database. An intersection was drawn between these DEGs and 1234 mitochondrial metabolism-related genes (MMRGs) acquired from public databases to identify candidate genes. Machine-learning-based expression-screening approaches were employed. A nomogram integrating these candidate genes was constructed and evaluated. Gene set enrichment analysis (GSEA), immune infiltration analysis, and drug prediction were conducted. Finally, the differential expression of these candidate genes was assessed using RT-qPCR. Stepwise bioinformatic screening based on public GEO datasets revealed ALDH3B1, ALOX15B and UPB1 as candidate genes. The three-gene nomogram exhibited moderate discriminatory performance in the training dataset. Correlation-based GSEA linked these genes to transcriptional programs involving lysosomal function, protein turnover, ribosomal processes, cellular signaling, and energy metabolism. Immune-cell enrichment analysis detected significant differences in six immune-cell signatures between the MDD and control groups, with several myeloid-cell signatures correlated to the candidate genes. Exploratory drug-gene interaction analysis identified four, eighteen, and two approved drugs associated with ALDH3B1, ALOX15B, and UPB1, respectively. RT-qPCR analysis revealed significantly elevated expression of ALDH3B1 ( p = 0.0138) and ALOX15B ( p < 0.0001) in MDD patients relative to healthy controls, while UPB1 showed a non-significant upregulation trend (fold change: 1.3245, p = 0.4842). This study nominates ALDH3B1 and ALOX15B as preliminary candidate genes associated with transcriptional perturbations of mitochondrial metabolism in MDD based on preliminary experimental observations from a small-cohort RT-qPCR assay. These candidates require further investigation in larger independent cohorts. In contrast, UPB1 serves as a bioinformatics-derived candidate gene necessitating further independent experimental validation.