Causal Effect of Multicohort Circulating Proteome on the Risk of Aortic Aneurysm: A Mendelian Randomization Study

The pathogenesis of aortic aneurysm (AA) remains unclear, and there are no effective therapeutic drugs or targets. Circulating plasma proteins are considered biomarkers of AA and potential therapeutic targets for AA. This study aimed to systematically evaluate the causal effects of plasma proteins on AA using a multicohort Mendelian randomization (MR) approach.

Protein quantitative trait loci (pQTLs) was obtained from 9 published proteome genome-wide association studies (GWASs) and AA GWAS data from the FinnGen cohort. Independent pQTLs were selected as instrumental variables (IVs). Two-sample MR analysis was performed using inverse-variance weighted, MR-Egger regression, weighted median, weighted mode, and simple mode methods. Heterogeneity and pleiotropy were assessed using Cochran's Q test, I² statistic, MR-Egger intercept, MR-PRESSO, and Leave-one-out analysis. Steiger filtering was used to test the causal direction. Colocalization analysis and pQTL-expression quantitative trait loci overlap assessment were conducted to validate the findings. Pathway enrichment and drug target analyses were performed to explore the biological and clinical implications of the MR results.

A total of 8,285 pQTLs for 4,421 proteins were retained as IVs. Using cis-pQTLs for IVs, MR analysis identified 154 proteins associated with thoracic aortic aneurysm (TAA; 76 protective and 78 risk factors) and 211 proteins with abdominal aortic aneurysm (AAA; 112 protective and 99 risk factors) Using cis-pQTLs combined with trans-pQTLs as IVs, MR analysis identified 236 proteins associated with TAA and 309 proteins with AAA. A subset of these associations survived false discovery rate (FDR) correction (FDR <0.05), representing the most robust findings. Comparison of the TAA and AAA proteomic profiles revealed both shared proteins (e.g., AHSG, MMP7, RARRES2, THBS2, CCL25) and condition-specific proteins (e.g., OVCA2, STAT3, and HPSE for TAA; PLAU, LPA, SERPING1, and SMPDL3A for AAA), reflecting the distinct embryonic origins and pathological drivers of these 2 conditions. Steiger filtering confirmed the expected direction of effect from circulating proteins to AA. Colocalization analysis found evidence of shared causal variants between multiple proteins and AA. Pathway enrichment analysis revealed involvement in stress response, immune regulation, cytokine-cytokine receptor interaction, and metabolic processes. Nearly two-thirds of the associated proteins were classified as druggable or potentially druggable targets.

This study identified a large number of potentially novel pathogenic proteins and therapeutic targets for AA, providing important references for elucidating the molecular pathogenesis of AA and advancing drug development. These findings warrant further validation through experimental studies and prospective clinical investigations.