Assessing the causal relationship between the plasma proteome and epilepsy: A Mendelian randomization study

Understanding the role of plasma proteins in the pathophysiology of epilepsy is crucial for uncovering novel biological mechanisms and therapeutic targets. Mendelian randomization (MR) provides a valuable tool for dissecting potentially causal associations between circulating proteins and disease risk. This study aimed to systematically assess potential causal relationships between the plasma proteome and epilepsy.

We implemented a multistage MR framework using genetic instruments for 4907 plasma proteins. Following rigorous quality control and instrument selection, a proteome-wide, two-sample MR analysis was performed against three epilepsy phenotypes (epilepsy, generalized epilepsy (GE), and focal epilepsy (FE)). Proteins showing nominal significance were further evaluated using Summary-data-based MR (SMR) and colocalization analysis to probe for a shared genetic basis. We additionally incorporated analyses using alternative pQTL resources to evaluate the robustness of our findings.

Our proteome-wide screen identified several candidate proteins associated with epilepsy risk at a nominal significance threshold (P-IVW < 0.01). SMR analysis provided support for candidates across the epilepsy subtypes, including CRABP2 and CD300C for epilepsy, GM2A for FE, and PCBD1 for GE. However, subsequent colocalization analysis did not yield strong evidence for a shared causal variant for any candidate, with some showing only moderate evidence (e.g., ACVRL1), underscoring a complex genetic architecture and the need for cautious interpretation. Supplementary analyses based on UKB-PPP pQTLs paired with FinnGen genome-wide association studies (GWAS) showed several nominal associations but limited overlap with the deCODE-FinnGen results.

Our exploratory assessment of the plasma proteome did not identify definitively causal proteins but did generate a prioritized list of candidates potentially linked to epilepsy. Proteins such as CRABP2, CD300C, GM2A, and PCBD1, which are involved in relevant neurodevelopmental, immune, and metabolic pathways, represent valuable leads for future experimental validation. By comparing results across different pQTL datasets, we also highlight the importance of replication and cross-resource validation in future epilepsy research.

Epilepsy is a common brain disorder, but its exact causes are often unknown. Our study used genetic data to investigate if thousands of proteins circulating in the blood have a causal effect on epilepsy risk. While we did not pinpoint a single definitive protein, we identified several promising candidates, such as CRABP2, CD300C, GM2A, and PCBD1, that may be linked to epilepsy risk. We also repeated the analysis using an additional protein dataset, which yielded some different results, showing that findings can vary depending on the data source. These findings provide valuable new leads for understanding epilepsy and developing future therapies.