Identification and Validation of Antidepressant Small Molecules Using Bioinformatics and Mouse Depression Models

Depression, a prevalent psychiatric disorder with limited effective treatments, can be addressed by repurposing existing small molecules via bioinformatics as a promising approach, though previous studies using tools like CMAP and GEO have successfully identified candidate drugs for neuropsychiatric disorders, few have combined in silico predictions with in vivo validation for it.

The aim of this study was to employ bioinformatics and in vivo experimental validation to mine potential antidepressant small molecule compounds.

This study utilized data from the GEO database, employing bioinformatics analysis methods to analyze the dataset. The CMAP platform was used to deeply explore potential antidepressant small-molecule compounds. In vivo experiments validated the antidepressant effects of the small-molecule compounds on a chronic restraint stress mouse model.

This study identified 311 differentially expressed genes (DEGs) from GSE182193-associated with the PI3K-Akt, MAPK, and neurotrophic factor signaling pathways, with key genes identified via Weighted Gene Co-expression Network Analysis (WGCNA) and immune correlation analysis-and screened 5 candidate compounds via CMAP, among which pyrimethamine, pifithrin-mu, and mibefradil significantly improved depressive behaviors in chronic restraint stress (CRS) model mice by regulating key protein expression in the PI3K-Akt and neurotrophic factor signaling pathways, as shown by a 33.44%-60.32% increase in movement distance in the open field test (P < 0.01 to P < 0.001) and a 20.25%-30.19% decrease in immobility time in the forced swim test (P < 0.01 to P < 0.001).

This study shows that pyrimethamine, pifithrin-mu, and mibefradil can regulate key proteins in the PI3K-Akt and neurotrophic factor pathways, improving depressive behaviors in mice and indicating their potential in alleviating depression; additionally, bioinformatics-driven repurposing of existing drugs for antidepressant discovery is more efficient than de novo development, and this study provides an exploratory demonstration of this.