Identification of Potential Biomarkers for Major Depressive Disorder: Based on Integrated Bioinformatics and Clinical Validation.

Abstract

Major depressive disorder (MDD) is a severe mental illness characterized by a lack of objective biomarkers. Mounting evidence suggests there are extensive transcriptional molecular changes in the prefrontal cortex (PFC) of individuals with MDD. However, it remains unclear whether there are specific genes that are consistently altered and possess diagnostic power. In this study, we conducted a systematic search of PFC datasets of MDD patients from the Gene Expression Omnibus database. We calculated the differential expression of genes (DEGs) and identified robust DEGs using the RRA and MetaDE methods. Furthermore, we validated the consistently altered genes and assessed their diagnostic power through enzyme-linked immunosorbent assay experiments in our clinical blood cohort. Additionally, we evaluated the diagnostic power of hub DEGs in independent public blood datasets. We obtained eight PFC datasets, comprising 158 MDD patients and 263 healthy controls, and identified a total of 1468 unique DEGs. Through integrated analysis, we identified 290 robustly altered DEGs. Among these, seven hub DEGs (SLC1A3, PON2, AQP1, EFEMP1, GJA1, CENPD, HSD11B1) were significantly down-regulated at the protein level in our clinical blood cohort. Moreover, these hub DEGs exhibited a negative correlation with the Hamilton Depression Scale score (P < 0.05). Furthermore, these hub DEGs formed a panel with promising diagnostic power in three independent public blood datasets (average AUCs of 0.85) and our clinical blood cohort (AUC of 0.92). The biomarker panel composed of these genes demonstrated promising diagnostic efficacy for MDD and serves as a useful tool for its diagnosis.