Toxic Mechanisms of PM2.5 Constituents (LPS and BPDE) in Cardiometabolic Disease: Insights from Integrated Machine Learning and Molecular Dynamic Simulations

Exposure to specific ambient pollutants, including certain PM2.5-bound constituents like benzo[a]pyrene-7,8-dihydrodiol-9,10-epoxide (BPDE) and lipopolysaccharide (LPS), has been increasingly implicated as a significant risk factor in the global burden of cardiometabolic diseases (CMD). However, the precise toxicological mechanisms through which these pollutants adversely affect cardiometabolic health remain poorly understood. Accordingly, this study aimed to delineate the effects of BPDE and LPS, both individually and in combination, on CMD, and to investigate the underlying molecular mechanisms driving its pathogenesis. We identified 366 and 287 potential targets for BPDE and LPS, respectively, in CMD through a multi-database screening (SwissTargetPrediction, ChEMBL, PharmMapper, CTD, GEO). Rigorous bioinformatic screening—integrating the STRING platform, Cytoscape (v3.10.0), and three machine learning (ML) methods—identified nine core targets: EGFR, ESR1, JAK2, PIK3R1, HDAC3, SORD, ADK, HDAC2, and SOD2. GO and KEGG analyses demonstrated their significant enrichment in key signaling (e.g., FoxO) and metabolic pathways (e.g., lipid metabolism and atherosclerosis), suggesting a mechanistic link to pollutant-induced CMD. Molecular docking and dynamics simulations demonstrated robust binding of BPDE and LPS to EGFR and MMP9, respectively, identifying these complexes as promising therapeutic targets for pollutant-associated CMD.