Glioblastoma (GBM) is a highly aggressive central nervous system malignancy with a dismal 5-year survival rate of less than 5%, and poorly understood environmental factors complicate its treatment. One such factor is bis(2-ethylhexyl) phthalate (DEHP, also known as di-2-ethylhexyl phthalate), a common plasticizer with documented neurotoxicity, yet its potential role in GBM pathogenesis remains elusive. In this study, we employed an integrative computational framework that combined network toxicology, single-cell transcriptomics, proteome-wide and metabolome-wide Mendelian randomization (MR), and molecular dynamics (MD) simulations to systematically investigate the interplay between DEHP and GBM risk. Cross-dataset targets were identified by mining toxicogenomic and disease databases, followed by the construction of a protein-protein interaction (PPI) network. This approach identified 76 overlapping targets between DEHP and GBM, which were refined to 24 hub genes through topological analysis. Notably, MR analyses revealed putative causal associations between higher genetically predicted plasma levels of three hub proteins-CD63, CTSS, and S100A4-and an increased risk of GBM, with S100A4 showing the strongest effect (odds ratio OR = 2.03, 95% confidence interval CI 1.15-3.58, p = 0.0149). This association was consistently validated across 11 independent cohorts, including TCGA, GTEx, and GEO datasets. Molecular docking and dynamics simulations identified S100A4 as the predominant binding target of DEHP, revealing a high-affinity interaction that may stabilize a metastasis-associated conformation. A two-step MR mediation analysis further indicated that S100A4 partially influences GBM risk by altering plasma lipid metabolites, with erucic acid mediating ~17% of the total effect. In conclusion, our analysis provides converging computational and genetic epidemiological evidence for a novel DEHP-S100A4-lipid metabolic axis that may contribute to GBM development. This pathway conceptually bridges environmental toxicology and neuro-oncology and highlights S100A4 and associated lipid disturbances as potential targets for preventive intervention. However, the proposed mechanistic links remain inferential, and definitive confirmation will require future in vitro and in vivo experiments to directly test the impact of DEHP on S100A4 expression, function, and downstream metabolic reprogramming in GBM models.
Tan et al. (Thu,) studied this question.