Integrating population epidemiology and bioinformatics to decipher the hypertensive effects of per- and polyfluoroalkyl substances mixtures

• Among the 26 candidate pollutants identified through six machine learning algorithms, the top 10 PFAS compounds were 6:2 Cl-PFESA, PFHpS, PFHxS, PFHpA, 8:2 Cl-PFESA, PFOS, PFNA, PFBS, PFTrDA, and PFOA. • Exposure to these PFAS compounds, both individually and combined, increased hypertension risk. • The combined effect of pollutants on hypertension is stronger in individuals aged ≤ 65 years and women. • Toxicological evidence revealed that PFAS-induced hypertension involved multiple pathways, such as the MAPK and the PI3K-Akt pathways, as well as genes including CYP3A4 and CYP1A1. Exposure to per- and polyfluoroalkyl substances (PFAS) and their alternatives has been linked to risk factors for hypertension. However, a comprehensive understanding of the interactions within complex PFAS mixtures and their biological mechanisms remains limited. As a result, this study aimed to examine the combined effects and underlying mechanisms of PFAS exposure on hypertension by integrating epidemiological data with bioinformatics analyses. Serum levels of 26 PFAS and their alternatives were measured in 1312 Chinese adults. Multiple machine learning models filtered the variables, and associations were examined using Bayesian kernel machine regression (BKMR) and weighted quantile sum (WQS) regression. Bioinformatics resources (CTD, GEO, GeneCards) were utilized to identify PFAS-modulated target genes. Machine learning models identified 10 pollutants. Exposure to a mixture of PFAS and their alternatives was positively associated with hypertension in the BKMR model, with PFBS identified as the primary contributor in the WQS model (WQS weight = 0.38). The combined effect of pollutants was stronger in participants aged 65 or younger and in women. Bioinformatics analysis identified 141 potential hypertension-related target genes modulated by PFAS exposure. Enrichment analyses suggested that PFAS-associated hypertension involves multiple pathways, including the mitogen-activated protein kinase (MAPK) and the mitogen-activated protein kinase (PI3K-Akt) pathways. Maximal clique centrality pinpointed 10 key genes, such as CYP3A4, CYP1A1, CYP2E1, CYP1A2, EPHX1, MAPK3, MAPK1, CREB1, AKT1, and TP53. Our findings demonstrate that PFAS mixtures are positively associated with hypertension risk, potentially involving mechanisms through the MAPK and PI3K-Akt pathways.