Large-scale plasma proteomics identifies protein signatures linking air pollution to epilepsy

Air pollution has been increasingly associated with epilepsy risk, but the underlying biological mechanisms remain unclear.

We conducted a large-scale proteome-wide analysis in a population-based cohort including 51,428 participants and 2254 plasma proteins. Concentrations of particulate matter (PM_2.5, PM_2.5-10) and gaseous pollutants (NO₂ and NO) were estimated using land-use regression models. Principal component analysis was performed to derive an air pollution index (API). Logistic regression and Cox proportional hazards models were applied to examine interactions between plasma proteins and air pollution exposures in relation to epilepsy prevalence and incidence, respectively. Functional enrichment, protein-protein interaction (PPI) network, neuroimaging association, and tissue- and cell-type transcriptomic enrichment analyses were conducted to explore the biological relevance of proteins exhibiting interaction effects.

We identified protein signatures exhibiting consistent interactions with multiple air pollution exposures in relation to epilepsy risk, including CIT-PM₂.₅, PTPRR-NO₂, and BCL2L15-API. Functional enrichment analyses indicated that these proteins were enriched in pathways related to immune regulation, inflammation, and cell death-related processes, which was consistent with PPI network analysis showing a highly interconnected module centered on immune-related proteins. Neuroimaging analyses suggested that these protein-air pollution interactions were associated with structural brain measures, including hippocampal volume and white matter hyperintensities. Additionally, tissue and single-nucleus RNA sequencing analyses showed that the corresponding genes were enriched in immune-related tissues and specific brain cell types, particularly neurons and endothelial cells.

This study identifies protein signatures that may modify the association between air pollution exposure and epilepsy risk, providing new insights into potential biological links and highlighting candidate molecular signatures for future mechanistic and translational investigations.