Genome-Wide Association Study Identifies Novel Variant Associated with Susceptibility to Pediatric Tuberculosis

Introduction RNA signatures and T cell activation responses to Mycobacterium tuberculosis (Mtb) have demonstrated good diagnostic and prognostic performance in adults. To further evaluate the implementation of these novel assays in children, we hypothesized that host genetic variation is associated with susceptibility to pediatric tuberculosis disease and blood biomarker expression. Methods We used a case-control study with a 12–36-mo prospective observation period to examine susceptibility to tuberculosis in children aged 2 mo to 5 yr with household Mtb exposure in Worcester, South Africa. Low-pass whole-genome sequencing followed by imputation was completed for 219 cases and 207 controls. A GWAS was performed to assess for association between genetic variants and susceptibility to tuberculosis disease. Lead genetic variants were subsequently evaluated for their effect on BCG-induced innate and adaptive immune responses. Whole blood collected from an independent cohort of 189 BCG-vaccinated 10-week-old South African infants was stimulated with BCG or media and examined with flow cytometry to measure cell surface-marker and cytokine expression. Results We identified a genome-wide significant variant rs4600676 (p = 4.5e-08) associated with increased susceptibility to pediatric tuberculosis. rs4600676 maps to a noncoding region of DNA and is positionally nearest genes GPR45 and TGFBAP1, suggesting it may regulate cytokine signaling. We identified an additional 25 lead genetic variants across 24 risk loci mapping to 55 genes at a prespecified suggestive significance level (p < 1e-05) associated with susceptibility to pediatric tuberculosis. MAGMA gene-set analysis of these suggestive variants revealed enrichment for chemokine receptor binding (p = 6.2e-06) and IRAK4-TLR2/4 deficiency (p = 2.2e-05). Six lead variants were associated with BCG-induced immune responses (p < 0.05). Specifically, rs34925982 maps to the T cell–inhibiting gene ILDR2, is associated with decreased BCG-induced T cell responses (p = 0.007), and increased susceptibility to childhood tuberculosis (p = 4.1e-06). Conclusion We identified a genome-wide significant variant associated with increased susceptibility to pediatric tuberculosis as well as an additional 26 variants suggestive of an association with tuberculosis susceptibility. These findings identify potential immunoregulatory mechanisms involved in host responses to tuberculosis that can be leveraged for therapeutic intervention. Future work will associate genetic variants with Mtb-induced RNA signatures and T cell activation responses to understand how performance of these assays may vary across populations.