Identification of the Shared Gene Signatures and Molecular Mechanisms Between Polycystic Ovarian Syndrome and Major Depressive Disorder: Evidence From Transcriptome Data

Abstract

Abstract Background: Polycystic Ovarian Syndrome (PCOS) is the most common metabolic and endocrine disorder in reproductive-age women, while Major Depressive Disorder (MDD) is a relatively common psychiatric condition. Previous studies have suggested a potential link between PCOS and MDD, but the underlying pathophysiological mechanisms remain unclear. This study aims to identify differential expression genes (DEGs) between PCOS and MDD using bioinformatics methods, explore the associated molecular mechanisms, elucidate the TF-mRNA-miRNA regulatory network involved, predict potential drug molecules, and validate them through molecular docking. Methods: Microarray datasets GSE34526 and GSE125664 were downloaded from the Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) of PCOS and MDD were analyzed using the GEO2R online tool to obtain shared DEGs to both. Next, the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis for the shared DEGs were performed. Then, protein-protein interaction (PPI) network were constructed and the hub genes were identified using the STRING database and Cytoscape software. Next, NetworkAnalyst was used to construct network between target transcription factors (TFs), microRNAs (miRNAs), and hub genes. Finally, the DSigDB database was used to search for potential drug molecules for the treatment of PCOS combined with MDD, followed by molecular docking using the AutoDock Tools and visualization of the results using PyMol 2.4.0. Results: In the above two datasets, 158 shared DEGs were identified. GO and KEGG enrichment analyses showed that these shared DEGs were mainly enriched in pathways related to neural signaling, energy metabolism, and chronic inflammation with immune dysregulation. In addition, genes with greater than 2-fold median interaction number were further screened by Cytoscape's plugin, cytoNCA, and finally 6 hub genes were selected from the PPI network, ncluding GRIN1, CNR1, DNM1, SYNJ1, PLA2G4A and EPHB2. Then, through the construction of the TF-mRNA-miRNA regulatory network, it was concluded that hsa-miR-27a might be a strongly associated miRNA with the pathogenesis of PCOS and MDD, while TFAP2A might be a strongly associated TF. Finally, orlistat, docosahexaenoic acid (DHA), capsaicin, and myo-inositol were considered as potential drug molecules for the treatment of PCOS combined with MDD using the DSigDB database and related study finding, and then molecular docking was performed using AutoDock Tools. The drug-molecule combination with the lowest binding energy was visualized using PyMol software and it found to be well docked. Conclusions: In summary, we constructed a TF-mRNA-miRNA regulatory network for the first time to characterize the interactions among potential TFs, miRNAs, and hub genes associated with PCOS and MDD, and concluded that aberrant neuronal signaling, disturbed energy metabolism, and immune dysregulation with inflammatory response may be the common pathogenesis of PCOS and MDD. In addition, we identified potential drug molecules for the treatment of PCOS and MDD and performed molecular docking validation. This provides new insights to identify potential associations, potential biomarkers and therapeutic agents for PCOS and MDD.