Ecotoxicol Environ Saf. 2026 Jan 12. pii: S0147-6513(26)00045-X. [Epub ahead of print]309
119716
Bisphenol compounds are pervasive environmental contaminants linked to neuropsychiatric disorders, yet their molecular interactions with major depressive disorder (MDD) pathogenesis remain unclear. This study employed an integrative approach combining network toxicology, machine learning, single-cell analysis, molecular docking, molecular dynamics simulations, and animal experiments to systematically identify key targets and pathways through which bisphenols may contribute to MDD. Network analysis of 123 shared targets between bisphenols and MDD revealed enrichment in corticosteroid responses, nuclear receptor activity, and oxidative stress pathways. Machine learning analysis prioritized five high-confidence biomarkers: PTGS1, MMP8, MAPK14, DAO, and BCHE, all of which exhibited significant differential expression in MDD patients (p < 0.05). Single-cell RNA sequencing revealed cell-type-resolved expression patterns, with BCHE enriched in oligodendrocyte precursor cells and reduced in MDD, while MAPK14 was broadly expressed across neuronal and glial populations, and PTGS1 showed relatively higher signal in microglia. Molecular docking suggests that the bisphenol compound exhibits stable binding affinities toward these five targets. Molecular docking and molecular dynamics simulations demonstrated strong binding affinities between bisphenols and DAO/BCHE, with stability confirmed over a 100 ns simulation. Animal experiments supported these findings, showing that bisphenol-exposed mice exhibited exacerbated depressive-like behavioural phenotypes. Consistently, qPCR and Western blot analyses of mouse brain tissue homogenates revealed that DAO expression was significantly downregulated, whereas PTGS1, MMP8, MAPK14, and BCHE were significantly upregulated, consistent with computational predictions. These findings provide a novel molecular framework for understanding the link between environmental pollutants and mental disorders, confirming the neurotoxic effects of bisphenol compounds.
Keywords: Bisphenol compounds; Machine learning algorithms; Major depressive disorder; Molecular docking; Molecular dynamics simulations; Network toxicology