Integrated population-based evidence, network toxicology, single-cell profiling, and in vitro studies elucidate...

Environmental exposure to polycyclic aromatic hydrocarbons (PAHs) has been associated with cardiovascular disease, yet the specific metabolites and cell-type-resolved mechanisms underlying stroke risk remain poorly defined. Here, we applied an integrated multilevel framework combining population-based exposure analysis with network toxicology, transcriptomics, and single-cell validation to investigate how PAH exposure, using 2-naphthol (2-NAP) as a representative metabolite, may contribute to stroke. PAH metabolites were analyzed in relation to stroke using logistic regression and restricted cubic splines, while mixture effects were evaluated using weighted quantile sum regression and machine-learning models. Network toxicology combined with transcriptomic and machine-learning analyses was used to identify candidate genes and prioritize four putative core genes (DPP4, IFIH1, KAT6A, and LAMP2). Single-cell RNA sequencing further resolved the cell-type-specific expression of these genes, highlighting endothelial cell and macrophage subsets and implicating TGF-β-related signaling pathways. Molecular docking suggested a direct interaction between the key metabolite 2-NAP and LAMP2, which was further examined in vitro. Functional experiments demonstrated that 2-NAP exposure increased LAMP2 expression, induced endothelial DNA damage, impaired microglial migration, and enhanced the production of pro-inflammatory cytokines. Together, these findings suggest that 2-NAP may represent an important contributor linking PAH exposure to stroke and that LAMP2-associated pathways may participate in PAH-related cerebrovascular injury. This study provides mechanistic insight by bridging epidemiological associations with defined molecular pathways and cerebrovascular cell populations, strengthening the biological plausibility of PAH-related stroke risk.