Neurotoxicity of PM2.5-Bound Multi-Pollutant Mixtures in Parkinson’s Disease: Exploring the INPP5D/SHIP1-PI3K-AKT Axis via an Integrative Multi-Omics Study

Abstract

Accumulating epidemiological evidence links ambient PM2.5 exposure to elevated Parkinson’s disease (PD) risk, yet the neurotoxicity triggered by PM2.5 multi-pollutant mixtures under real co-exposure scenarios remains insufficiently elucidated. This study integrated two-sample Mendelian randomization (MR), mixture-oriented network toxicology, multi-omics bioinformatics, machine learning and in vitro cell validation to systematically uncover the causal molecular network of PD pathogenesis induced by mixed PM2.5 pollutants. Two-sample MR based on public GWAS datasets suggested a potential positive genetic association between PM2.5 exposure and increased PD susceptibility. We intersected PM2.5 component targets, PD risk genes and differential transcripts to obtain 195 shared genes, among which INPP5D (encoding SHIP1) was identified as the hub gene via LASSO and random forest algorithms. Molecular docking simulations further predicted that metal–organic pollutant complexes may exhibit stronger in-silico binding affinity toward SHIP1 than organic pollutants alone. Pathway enrichment screened the PI3K-AKT cascade as the core regulatory pathway. In SH-SY5Y cells, NIST SRM 1648a (urban particulate matter) treatment upregulated SHIP1 expression and phosphorylation, suppressed PI3K-AKT activity, and boosted neuroinflammation and apoptosis, whereas INPP5D knockdown rescued these pathological changes. Overall, our multi-omics and in vitro findings raise a plausible mechanistic hypothesis that NIST SRM 1648a-mimicked PM2.5-related pollutant mixtures may exacerbate PD-like neuronal injury through the SHIP1/PI3K-AKT signaling axis. This multi-dimensional evidence provides an analytical paradigm for mixed pollutant-induced neurodegeneration and highlights INPP5D as a promising intervention target for PM2.5-associated PD.