cis-eQTL mapping of TB-T2D comorbidity elucidates the involvement of African ancestry in TB susceptibility

Abstract

The validation of genome-wide association signals for tuberculosis (TB) susceptibility and the development of type 2 diabetes (T2D) across diverse populations remain problematic. The ancestry-specific variants (coding and non-coding) that contribute to previously identified differentially expressed genes (DEG) in patients with TB, T2D and comorbid TB-T2D, remain unknown. Identifying ancestry-specific expression quantitative trait loci (eQTLs) can aid in distinguishing the most probable disease-causing variants for population-specific therapeutic interventions. Therefore, this study conducted cis-eQTL mapping in TB, T2D and TB-T2D patients to identify variants associated with DEG. Both genotyping (Infinium H3A array with [~]2.3 M markers) and RNA sequencing data of 96 complex multi-way admixed South Africans were used for this purpose. Importantly, both global-and local ancestry adjustment were included in statistical analysis to account for complex admixture. Unique gene-variant pairs were associated with TB-T2D on chromosome 7p22 whilst adjusting for Bantu-speaking African ancestry (PRKAR1B:rs4464850; P=7.68e-07) and Khoe-San ancestry (PRKAR1B:rs117842122; P=3.66e-07). In addition, IFITM3 (a biomarker for the development of TB) was associated with three SNPs (rs11025530, rs3808990, and rs10896664) on chromosome 11p15 while adjusting for Khoe-San ancestry. Our results also indicated that the upregulation of the NLRP6 inflammasome is strongly associated with people with TB-T2D while adjusting for Khoe-San ancestry. Three African-specific eGenes (NLRP6, IFITM3 and PRKAR1B) would have been missed if local ancestry adjustment was not conducted. This study determined a list of ancestry-specific eQTLs in TB-T2D patients that could potentially guide the search for new therapeutic targets for TB-T2D in African populations. Author SummaryThe limitation of genome-wide association study (GWAS) is that the particular biological pathway impacted by a variant might not be evident. eQTL mapping can be conducted to determine the impact that a genetic variant might have on the expression of a specific gene in a biological pathway. In this study the use of cis-eQTL mapping was explored to elucidate the underlying genetic variants that regulate gene expression between TB-T2D and T2D patients, and between TB patients and healthy controls with multi-way genetic admixture from South Africa. Using RNA sequencing data and newly genotyped dataset of 96 individuals (Illumina Infinium H3Africa array with [~]2.5 M markers), we were able to identify ancestry-specific eQTLs. eQTLs of indigenous Khoe-San ancestral origin were identified in genetic regions previously implicated in TB and T2D in African populations. If local ancestry was not incorporated in the cis-eQTL mapping analysis these important African-specific eQTLs would have been missed. Our results provide a list of possible ancestry-specific causal variants associated with TB-T2 comorbidity that could guide the search for new therapeutic targets for African-specific populations. Including populations with complex ancestry and admixture in genetic studies is necessary to improve the quality of genetic research in sub-Saharan African groups.