Abstract
N-(1,3-dimethylbutyl)-N'-phenyl-p-phenylenediamine (6PPD) and its oxidation product 6PPD-Q are emerging tire-derived environmental contaminants that have been detected in human biological samples; however, their neurotoxic mechanisms remain unclear. This study employed network toxicology, transcriptomic analysis, and machine learning to explore the potential neurotoxic associations between 6PPD/6PPD-Q and Parkinson's disease (PD). By integrating four target prediction platforms and three disease databases, a total of 260 potential toxic targets associated with 6PPD/6PPD-Q-related PD were identified. Eighty-two differentially expressed genes were identified from GEO datasets, and five candidate genes were obtained through intersection analysis. Among the five candidate genes, KCND3, PSEN2, and SLC18A2 were consistently selected by three machine-learning algorithms (LASSO, SVM-RFE, and Boruta). Validation analysis showed that all core genes were significantly downregulated in the substantia nigra tissues of PD patients, with area under the curve (AUC) values of 0.917 for KCND3, 0.757 for PSEN2, and 0.819 for SLC18A2. Mendelian randomization analysis provided preliminary genetic evidence consistent with a causal association between elevated SLC18A2 expression levels and reduced PD risk (OR = 0.488, 95% CI: 0.273-0.870, P = 0.015). Molecular docking analysis showed favorable predicted binding scores between 6PPD/6PPD-Q and the three target proteins. Exploratory immune analysis suggested alterations in activated dendritic cells and γδ T cells in PD. Collectively, these findings suggest potential neurotoxic relevance of 6PPD/6PPD-Q to PD and provide a basis for further experimental investigation.