X-linked multi-ancestry meta-analysis reveals tuberculosis susceptibility variants

Globally, tuberculosis (TB) presents with a clear male bias that cannot be completely accounted for by environment, behaviour, socioeconomic factors, or the impact of sex hormones on the immune system. This suggests that genetic and biological differences, which may be mediated by the X chromosome, further influence the observed male sex bias. The X chromosome is heavily implicated in immune function and yet has largely been ignored in previous association studies. Here we report the first multi-ancestry X chromosome specific meta-analysis on TB susceptibility. We identified X-linked TB susceptibility variants using seven genotyping data sets and 20,255 individuals from diverse genetic ancestries. Sex-specific effects were also identified in polygenic heritability between males and females along with enhanced concordance in direction of genetic effects for males but not females. These sex-specific genetic effects were supported by a sex-stratified and combined meta-analysis conducted using the X chromosome specific XWAS software and a multi-ancestry analysis using the MR-MEGA software. Seven significant associations were identified. Two in the overall analysis (rs6610096, rs7888114) and a second for the female specific analysis (rs4465088) including all data sets. For the ancestry specific meta-analysis three significant associations were identified for males in the Asian cohorts (rs1726176, rs5939510, rs1726203) and one in females for the African cohort (rs2428212). Several genomic regions previously associated with TB susceptibility were reproduced in this study, along with strong ancestry-specific effects. These results support the hypothesis that the X chromosome and sex-specific effects could significantly impact the observed male bias in TB incidence rates globally.   

Methods
This analysis includes 7 of the 17 published (and unpublished) GWAS studies of TB (with HIV-negative cohorts) prior to 2022. It excludes data from Iceland and Vietnam, as they declined to share data. It excludes data from China, Korea, Peru, and Japan, as data-sharing agreements could not be finalized in time for this analysis. The Indonesian data was not suitable for reliable imputation, and the Moroccan data was family-based and thus also not suitable for this meta-analysis. Data from Thailand, Japan, Estonia and Germany were excluded as they did not have X chromosome genotyping. Finally, genotyped TB cases and controls are also available in the UK Biobank, but this data was not included in this analysis as genetic association studies on such highly selected datasets need to be undertaken with caution, and to not bias results, were excluded for this analysis.

Included individuals were genotyped on a variety of genotyping arrays, and raw genotyping data were available for eight datasets, and for the remainder, association testing summary statistics were obtained to use in the meta-analysis. Quality control (QC) and imputation of the data with raw genotyping information available was done using Plink (v1.9), followed by pre-phasing using SHAPEIT and Impute2 with the 1000 genomes phase 3 reference panel. QC and imputation were done as described previously; briefly we used a minor allele filter of 0.025 and an individual and SNP missingness filter of 0.1. Hardy–- equilibrium threshold was set at a Bonferroni corrected p-value according to the number of SNPs testes (0.05/number of SNPs) and samples where sex could not be determined from genotyping were also removed. Imputed data were filtered at a quality score of 0.3, prior to individual and genotype filtration steps. Prior to QC and imputation, allele orientation was corrected using Genotype Harmoniser version 1.4.15, and the genome build of all datasets was checked for consistency (GRCh37) and updated if necessary, using the liftOver software from the UCSC genome browser. The final imputed datasets again went through a QC process, but this time in a sex-stratified manner using the XWAS software and XWAS pipeline. Sex-stratified GWAS analysis of the individual datasets was then done using the XWAS software. The results from the sex-stratified GWAS were then combined in a sex-stratified and combined meta-analysis using the XWAS software. The results of the sex-stratified and combined meta-analysis are uploaded here.