Identification and Experimental Validation of Mitochondria-Related Biomarkers for Depression and Neuropathic Pain

Background Depression and neuropathic pain (NP) often coexist, significantly affecting patients' quality of life. The present investigation sought to explore the significance of genes associated with mitochondrial dysfunction (MDRGs) in the context of depression comorbid with NP. Methods The current study sourced gene expression datasets GSE92718 and GSE89224. Subsequent analysis involved the functional enrichment techniques and the generation of protein interaction networks (PPI) to elucidate relevant biological mechanisms. To formulate a predictive model for depression diagnosis, diverse analytical approaches were employed, encompassing logistic regression analysis, least absolute shrinkage and selection operator (LASSO) regression, random forest (RF), development of nomograms, calibration assessments, DCA, and ROC curves. Based on median mitochondrial dysfunction scores, depression samples underwent stratification into high- and low-score categories, and consensus clustering was subsequently performed to delineate unique depression subtypes. Furthermore, we conducted a study involving six patients with depression and NP and six healthy controls who met the specific criteria, assessing the expression levels of hub genes. Results Logistic regression, LASSO regression, and RF analyses identified Mcl1 and Cdkn1a as significant predictors of depression. Nomogram results emphasized the diagnostic potential of Mcl1 for identifying depression. Calibration curves, DCA, and ROC curves demonstrated its good diagnostic performance. Single-sample Gene Set Enrichment Analysis (ssGSEA) assessed immune infiltration across 28 immune cell subsets, uncovering significant differences between mitochondrial dysfunction score-defined groups. Principal component analysis (PCA) visually depicted clear segregation between the two depression subtype groups. Quantitative reverse transcription-PCR (RT-qPCR) analysis showed significant changes in the expression of hub genes in patients with depression and neuropathic pain (NP) compared to healthy controls. Conclusions Our study demonstrated that the diagnostic model incorporating Mcl1 and Cdkn1a shows robust predictive capability, highlighting these genes as promising biomarkers for early clinical detection.