Benzo[a]pyrene promotes gastric cancer progression via activation of the Correa cascade through modulation of the STAT3-TP53-MMP9 molecular axis

To investigate the role of Benzo[ a ]pyrene (BaP) in driving the Correa cascade during gastric cancer development, we employed an integrated strategy combining network toxicology, machine learning, and molecular dynamics (MD) simulations. We identified 301 co-expressed genes spanning the Correa sequence, from chronic inflammation to invasive carcinoma. A protein-protein interaction network was constructed using STRING, and CytoHubba analysis highlighted five hub genes: TNF, IL6, IFNG, IL1B, and STAT3. Using CHEMBL and SUPER-PRED, we predicted 846 potential BaP targets. Intersection with disease-related genes revealed 62 common targets. Among eight candidate hub genes, an integrated Stepglm[both] and Random Forest model identified STAT3, TP53, and MMP9 as core targets. Receiver operating characteristic analysis confirmed their strong diagnostic potential (AUC > 0.78), while SHAP analysis ranked STAT3 as the most influential factor (SHAP = 0.241). Notably, these genes exhibited synergistic expression patterns in tumors (STAT3-TP53: ρ = 0.175; STAT3-MMP9: ρ = 0.261; TP53-MMP9: ρ = 0.216; all P < 0.01) and showed a dose-dependent association with disease progression. Genomic profiling revealed frequent mutations and amplifications in STAT3, TP53, and MMP9, with TP53 exhibiting the highest mutation rate. Analysis using UALCAN demonstrated significant upregulation of their mRNA levels in tumor tissues compared to normal tissues (P < 0.05). Clinically, high STAT3 and TP53 expression correlated with poorer survival, whereas elevated MMP9 levels were associated with improved outcomes. Mechanistic studies, including molecular docking and dynamics simulations, confirmed stable BaP-target interactions (e.g., STAT3 binding energy = −8.285 kcal/mol) mediated by non-covalent interactions, which disrupt the bidirectional STAT3-TP53 regulatory axis (STAT3 → MDM2 ⊣ TP53; TP53 → PIAS3 ⊣ STAT3). In summary, this study identifies STAT3, TP53, and MMP9 as central mediators of BaP-induced progression along the Correa cascade via a synergistic regulatory network. These findings provide new insights into environmental gastric carcinogenesis and highlight potential therapeutic strategies, including dual STAT3/MDM2 inhibition or MMP9 blockade. • Integrated Computational Toxicology: This study is the first to employ a comprehensive strategy combining network toxicology, multi-algorithm machine learning, and molecular dynamics simulations to systematically decipher the molecular pathogenesis of Benzo[ a ]pyrene (BaP)-induced gastric carcinogenesis along the Correa cascade. • Identification of a Central Molecular Axis: We identified and validated STAT3, TP53, and MMP9 as the core hub genes that mediate BaP toxicity, serving as robust diagnostic and prognostic biomarkers with synergistic expression patterns in gastric tumors. • Mechanistic Elucidation of a Perturbed Regulatory Network: We elucidated a novel bidirectional STAT3-TP53 regulatory axis (STAT3→MDM2⊣TP53 / TP53→PIAS3⊣STAT3), conceptualized as a “regulatory balance system,” which is directly disrupted by stable BaP binding, thereby promoting inflammation, genomic instability, and invasion. • Translation to Therapeutic Strategies: The deciphered molecular network provides a rational framework for precision intervention, proposing actionable strategies such as dual STAT3/MDM2 inhibition and MMP9 blockade for the management of BaP-associated gastric cancer.