Identification of circulating inflammatory proteins as potential drug targets for sepsis through Mendelian randomization and colocalization analyses.

Sepsis is a complex disease with limited therapeutics. We conducted a 2-sample Mendelian randomization (MR) analysis to identify circulating inflammatory proteins as potential drug targets for sepsis. Protein quantitative trait loci (pQTLs) at cis-region for 448 circulating inflammatory proteins were obtained from the UK biobank pharma proteomics project (UKB-PPP) (N = 54219), and summary-level genetic associations with sepsis were extracted from the FinnGen cohort (12,301 cases and 332,343 controls) for the primary MR analysis. The findings were further replicated using alternative cis-pQTL (Fenland, N = 10708) and sepsis genome-wide association study data (UK Biobank, 2811 cases and 406,150 controls). Multiple sensitivity analyses including summary-data-based MR (SMR), the HEIDI test, colocalization, and phenome-wide scanning were performed to verify the causality of the identified proteins. Moreover, protein-protein interaction analysis was performed to examine the interactions between identified proteins and current drug targets in clinical trials. With the adjusted P-value <.05, 1-SD increase in genetically determined APOE (apolipoprotein E) was associated with increased risk of sepsis (OR = 1.07, 95%CI = 1.04-1.11, P = 5.36 × 10-5). The findings were consistent in the validation analyses. Multi-SNPs SMR analysis (P = 1.42 × 10-3), HEIDI test (P = .53), and colocalization analysis (PP4 = 0.95) further supported the causality of APOE. Protein-protein interaction analysis suggested that APOE was interacted with multiple drug targets tested in the clinical trials for sepsis. This integrative analysis combing the pQTL and genome-wide association study data identified APOE as a promising drug target for sepsis. Further experimental studies were warranted to explore the details of the underlying biological mechanisms and feasibility of drug development.