Abstract
BACKGROUND: Major depressive disorder (MDD) is a prevalent mental health condition in which mitochondrial dysfunction and cellular senescence contribute to its pathogenesis. This study aims to identify biomarkers related to mitochondria-associated genes (MRGs) and aging-related genes (ARGs) in MDD using bioinformatics. METHODS: This study utilized data from GSE201332 and GSE52790, including 1,136 MRGs and 866 ARGs. Initially, candidate genes were selected by intersecting MRGs, ARGs, and differentially expressed genes (DEGs) derived from differential expression analysis in GSE201332. Biomarkers were identified through LASSO regression analysis of the candidate genes. The biomarkers were then evaluated using ROC curves, and artificial neural network (ANN) models were constructed. Subsequently, functional enrichment, immune-related analyses, drug predictions, and molecular docking were performed. Finally, the expression of biomarkers was validated using reverse transcription-quantitative polymerase chain reaction (RT-qPCR). RESULTS: Seven candidate genes were identified from the intersection of 4,041 DEGs, 1,136 MRGs, and 866 ARGs, with SLC25A5, ALDH2, CPT1C, and IMMT identified as potential biomarkers for MDD through LASSO regression analysis. ROC curve analysis in both GSE201332 and GSE52790 showed that these biomarkers effectively distinguished between MDD and control samples, with AUC values exceeding 0.7. ANN models further confirmed the diagnostic potential of these biomarkers. Gene set enrichment analysis (GSEA) revealed significant enrichment of SLC25A5, CPT1C, and IMMT in pathways related to cellular protein complex assembly and chromatin organization. Immune infiltration analysis demonstrated significant positive correlations between SLC25A5, ALDH2, and IMMT and most of the 18 immune cell types. Molecular docking predictions identified ALDH2 and SLC25A5 as potential targets for specific drugs, with NITROGLYCERIN showing the best binding affinity to ALDH2 (-6.4 kcal/mol). RT-qPCR validation showed significantly lower expression of SLC25A5 and IMMT, and higher expression of CPT1C, in patients with MDD compared to controls (p < 0.05), consistent with bioinformatics predictions. CONCLUSION: This study identified SLC25A5, ALDH2, CPT1C, and IMMT as biomarkers associated with MDD, offering insights into its molecular mechanisms.