Mechanistic Biomarker Discovery of Perfluorooctanoic Acid Exposure in Parkinson's Disease Through Integrated Toxicogenomic Analysis

Perfluorooctanoic acid (PFOA) is a common polyfluoroalkyl substance (PFAS) with possible neurotoxic effects, but its role in Parkinson's disease (PD) is not clear. We aimed to identify shared genes and pathways that link PFOA to PD, to test if the link is causal and to develop blood-based biomarkers for case identification. Using the Comparative Toxicogenomics Database (CTD), we found 62 genes common to both PFOA and PD, which were enriched in pathways related to oxidative stress, xenobiotic metabolism and neurodegeneration. Mendelian randomization and colocalization analyses identified BST1 as a key gene, where higher BST1 expression was causally linked to increased PD risk (OR = 1.41, p = 0.001). The eQTL and PD signals shared a lead variant (rs4389574; PP.H4 = 0.980), confirming a common genetic origin. We built a cross-sectional case identification model from Parkinson's Progression Markers Initiative (PPMI) whole-blood RNA-seq using PFOA-linked genes. It achieved moderate discrimination (OOF AUC 0.625) and surpassed a demographic baseline (AUC 0.498); external validation without refitting produced AUCs of 0.642 (GSE6613) and 0.632 (GSE99039). Finally, we explored potential protein-chemical interactions in silico. Molecular docking and dynamics simulations demonstrated that PFOA can maintain stable binding within the BST1 pocket, providing a structural hypothesis for future experimental validation.