BTN3A2 protects against Escherichia coli infection: Insights from genome-wide association and Mendelian randomization

Background Escherichia coli (E. coli) urinary tract infections (UTIs) pose a significant and growing global health burden, yet host genetic susceptibility remains poorly understood. Methods In this genome-wide association study (GWAS), we analysed 56,044 Swedish individuals drawn from three national cohorts (Epihealth, Simpler and the Swedish Twin Registry) with linked microbiological culture records from 2005 to 2024. A total of 12,962 individuals with at least one culture-verified E. coli -positive urine sample or UTI diagnosis were defined as cases, while 43,082 controls had no recorded culture-positive results. In addition, we analysed 6,807 patients who were admitted to hospital with an E. coli infection as a severe phenotype against 43,842 controls. Genome-wide analysis identified one genome-wide significant locus on chromosome 4 for E. coli susceptibility and one locus on chromosome 6 for severe infection. Results After linkage disequilibrium pruning, two independent signals remained. The top SNP for E. coli susceptibility, rs6819738 (4:36479467:T:C; p = 1.78×10⁻ ⁸ ) has no previously reported associations in the GWAS Catalog. The lead variant for severe E. coli (rs198820, 6:26472114:A:G; p = 2.18x10 ^− 08 ) infection acts as a significant expression quantitative trait locus (eQTL) for BTN3A2 (p = 1.7 × 10⁻ ³⁶ ), a gene involved in T-cell regulation and confirmed to have elevated protein expression in infectious disease states using data from the Human Protein Atlas. Mendelian randomisation analyses of independent cis-eQTL variants for BTN3A2 (rs9357006, rs71557332, rs1624064, rs3799378) suggested that higher expression of BTN3A2 is associated with a lower risk of E. coli infection and severe infection. Conclusion This is the first GWAS based on culture-verified E. coli infection at a single site, the urinary tract, and provides functionally supported evidence for novel host genetic determinants of infection risk. These findings underscore the value of precise microbial phenotyping and functional annotation in advancing infectious disease genomics.