Systematic druggable genome-wide Mendelian randomization identifies therapeutic targets for major depressive disorder
Authors: Zhou Menghan, Liu Shuning, Jiang Tao, Sun Zhuangzhuang, Cao Lingling, Su Xin, Yu Cheng, Guo Junpeng
BACKGROUND: The occurrence of major depressive disorder is typically associated with genetic and environmental factors. Currently, the diagnosis of major depressive disorder mainly relies on clinical interviews and symptom assessments, lacking clear and reproducible biological markers. This can lead to misdiagnosis and missed diagnoses, delaying the timing of treatment. OBJECTIVE: To identify druggable genes that may act as potential therapeutic targets for major depressive disorder by conducting comprehensive genome-wide Mendelian randomization analysis. METHODS: By integrating expression quantitative trait locus (eQTL) data and protein quantitative trait locus (pQTL) data from pharmacologically actionable genes with genome-wide association study (GWAS) data on major depressive disorder (including 177 377 cases and 445 321 controls), Mendelian randomization analysis was conducted to identify druggable genes that have a causal relationship with major depressive disorder. Additionally, enrichment analysis, protein-protein interaction network construction, drug target identification, and molecular docking simulations were performed to further explore potential therapeutic strategies. RESULTS AND CONCLUSION: A total of 4 394 druggable genes were analyzed, and 21 druggable genes considerably associated with major depressive disorder were identified. Bayesian colocalization analysis indicated that BTN3A3, CISD1, and PSMB4 had posterior probabilities of hypothesis 4 (H4.abf) > 0.5, supporting the possibility of shared causal variants. GO enrichment analysis mainly involved 'antigen processing and presentation', 'protein degradation and processing', 'mitochondrial outer membrane', and 'immune receptor activity' pathways related to major depression. Protein-protein interaction network analysis showed moderate connectivity among the identified genes (21 nodes, 14 edges). Drug target identification determined gemcitabine (CID 60750), fucose (CID 17106), and isococculidine (CID 2826) as main candidate compounds, which had strong associations with several key genes. Molecular docking analysis revealed stable drug-protein interactions, with isococculidine showing the most stable binding energy (-52.74 kJ/mol) with BTN3A3. In conclusion, Mendelian randomization combined with genomics and structural biology analysis provides valuable decision-making basis for target prioritization and drug repurposing, offering new ideas and directions for efficient utilization of basic research resources and drug development for major depressive disorder.