Bisphenol A (BPA) is a ubiquitous environmental endocrine disrupting chemical used in plastics and consumer goods. BPA can cross the blood–brain barrier and has been associated with neuroinflammation and affective symptoms in preclinical and epidemiologic work. Despite growing evidence linking BPA exposure to depressive phenotypes, the specific molecular targets and signaling pathways mediating BPA-induced depression remain incompletely defined. To address this gap, the authors integrated network toxicology, molecular docking, molecular dynamics (MD) simulation, transcriptomic dataset analysis, and in vivo experiments to systematically identify candidate toxicological targets and mechanistic pathways connecting BPA to depression pathogenesis.
Initial toxicity predictions for BPA were obtained using ProTox and ADMETlab tools to identify endpoints relevant to neurotoxicity (for example, blood–brain barrier permeability and estrogen receptor interaction). BPA chemical information including the SMILES string was retrieved from PubChem. Potential BPA targets were compiled from ChEMBL, STITCH, and SwissTargetPrediction, with STITCH and SwissTargetPrediction results filtered using a high-confidence threshold (prediction score ≥ 0.7). Depression-associated gene lists were obtained from GeneCards (relevance score > 10), OMIM, and the Therapeutic Target Database (TTD). Overlapping genes between BPA-related and depression-associated lists were identified using Venny and visualized in Cytoscape to generate an initial compound–disease network.
The overlapping gene set (29 targets) was submitted to STRING to construct a human PPI network with a high confidence threshold (minimum score > 0.7). Network topology metrics were calculated in Cytoscape to identify hub nodes. GO (BP, MF, CC) and KEGG enrichment analyses were performed using clusterProfiler, with terms retained when they included at least three overlapping genes and reached P < 0.05. Enrichment results emphasized processes related to neural synaptic function, neurotransmitter binding, and the neuroactive ligand–receptor interaction pathway, suggesting synaptic and receptor-mediated mechanisms by which BPA could influence mood regulation.
Molecular docking used AutoDock Vina to assess BPA binding affinity to candidate hub proteins. Protein 3D structures were sourced from RCSB PDB, prepared by removing waters and ligands, adding hydrogens, and assigning charges. Blind docking across full protein surfaces identified top pockets; focused docking was then centered on biologically relevant domains (for example, the ligand-binding domain of ESR1 and the substrate channel of SLC6A4). ESR1 (PDB: 1ERE) served as a positive control and lysozyme (PDB: 1AKI) as a negative control. Docking showed that five core hub proteins—INS, ESR1, SLC6A4, GRIA1, and NTRK2—had relatively favorable binding energies compared with the lysozyme control. Exact docking energies and grid parameters are reported in the original paper and supporting files.
To evaluate the dynamic stability of protein–BPA complexes, 100-ns MD simulations were performed for each identified complex using GROMACS with the amber99sb-ildb protein force field and GAFF parameters for BPA. Systems were solvated with TIP3P water and neutralized with ions. After energy minimization and NVT/NPT equilibration, production runs were carried out at 300 K and 1 atm. Convergence was assessed by RMSD plateaus in the final 20 ns of trajectories. Trajectory analyses included protein and ligand RMSD, ligand displacement relative to the protein, per-residue RMSF mapped to functional domains, radius of gyration (Rg), solvent-accessible surface area (SASA), hydrogen-bond counts over time, and construction of free-energy landscapes (RMSD versus Rg). For each complex, MM-GBSA binding free-energy calculations were performed on 200 evenly spaced snapshots from the 100-ns trajectories; energies were decomposed into van der Waals, electrostatic, polar-solvation, and non-polar-solvation contributions. These combined analyses supported stable, specific interactions between BPA and the five hub proteins with overall favorable binding free energies, as reported in the source.
Public transcriptome datasets from GEO (GSE26063, GSE102556, GSE38206, GSE32280) were analyzed to evaluate expression patterns of the identified hub genes in major depressive disorder (MDD) samples versus controls. Differential expression was performed using limma with adjusted P < 0.05. Across datasets, the five core genes (INS, ESR1, SLC6A4, GRIA1, NTRK2) were reported as markedly downregulated in MDD patients. Gene set enrichment analysis (GSEA) on GSE38206 highlighted suppression of neurotrophic and insulin-related pathways in MDD, aligning with the identification of INS and NTRK2 among the core targets. Volcano plots and expression comparisons per dataset are provided in the original article’s figures and supporting files.
Adult male C57BL/6J mice received daily oral administration of BPA (50 mg/kg in corn oil) or vehicle for 28 days. Behavioral assays were conducted 24 hours after the final dose. The sucrose preference test (SPT) assessed anhedonia, while the open field test (OFT) evaluated locomotor activity and anxiety-like behavior (total distance, average speed, central area distance) using video tracking. After behavioral testing, hippocampus and prefrontal cortex were dissected for molecular assays. Western blotting and quantitative PCR (qPCR) measured protein and mRNA levels of INS, ESR1, SLC6A4, GRIA1, and NTRK2, normalized to β-actin and GAPDH respectively. The study found that BPA-exposed mice exhibited aggravated depressive-like behaviors and significant downregulation of these five molecules at both mRNA and protein levels in brain tissues. Methodological details including antibody sources, qPCR cycling parameters, and primer listings are provided in the source.
Combining in silico network toxicology, structure-based binding analyses, transcriptomic validation, and in vivo experiments, the study presents convergent evidence that BPA interacts with multiple proteins implicated in synaptic transmission, neurotransmitter handling, neurotrophic signaling, estrogenic signaling, and insulin-related processes. The identification of INS, ESR1, SLC6A4, GRIA1, and NTRK2 as core BPA-associated targets links metabolic, hormonal, monoaminergic, glutamatergic, and neurotrophic pathways—pathways previously implicated in depression pathophysiology. These multi-target interactions offer a plausible mechanistic basis for how chronic BPA exposure might contribute to depressive phenotypes.
The work integrates multiple complementary approaches but, as with any multi-method study, limitations exist. Specific quantitative docking scores, full gene lists, and detailed behavioral effect sizes are contained in the original article and supporting files rather than reproduced in this summary. The study used male adult mice and a single BPA dosing regimen; sex differences, developmental exposure windows, dose–response relationships, and causal verification via genetic or pharmacologic manipulation of identified targets remain areas for future investigation. Overall, the study provides a mechanistic framework and experimental leads for further mechanistic and translational work on BPA-related mood disorder risk.