Host genetic risk factors and susceptibility to tuberculosis in Hong Kong Chinese

Abstract

Tuberculosis (TB) is caused by infection with mycobacterium tuberculosis (MTB) and remains the leading cause of death due to infections in Hong Kong and worldwide. Host factors are important in TB development. In the present study, we investigated the host genetic risk factors for susceptibility to TB in two major areas. First, a replication study testing association between the SLC11A1 gene and TB susceptibility was performed in Hong Kong Chinese. The solute carrier family 11, member 1 (SLC11A1, also known as NRAMP1) gene is located at 2q35 and encodes an integral membrane protein expressed in macrophages/monocytes and polymorphonuclear leukocytes. SLC11A1 polymorphisms were found associated with TB in some populations, but not in others. The extent of linkage disequilibrium (LD) of genetic markers within and around the SLC11A1 locus was first examined in a group of 360 healthy Chinese blood donors. According to the LD pattern and the previous positive association findings, two genetic markers within SLC11A1 (SLC1 in the 5' untranslated region, and SLC6a/b in the 3' end of the gene), and one (IL8rb) within the nearby IL8RB locus were selected. SLC6a/b consisted of two markers (SLC6a and SLC6b) in perfect LD and thus was tested as a single marker. The SNP IL8rb was included in this replication study because it was in LD in the SNPs in the 5' end of the SLC11A1 locus and thus might account for previously reported association of SLC11Al with TB, unless proven otherwise. In total, 278 blood samples from pulmonary TB patients (TBP) and 282 blood samples from sex- and age-matched unrelated control individuals (TBC) proven with no active TB were collected for the association study. When TBP was compared with TBC, statistically significant differences in allelic (p = 0.0165) and genotypic (p = 0.0163) frequencies of the linked markers SLC6a/b (classically known as D543N and 3'UTR) were found, even after correction of multiple testing by false discovery rate at a level of 0.05 (adjusted cut-off p value = 0.01667). With stratification by sex or age, positive association results were observed in the female group and in the young age group (age <=65 years), and were also significant even after correction for multiple comparison. These results suggested that the contribution of the SLC11A1 gene to the susceptibility to TB was mainly restricted to these groups, a finding not previously reported. This also highlights the importance of controlling the confounding variables sex and age in genetic association studies for TB. Our study supports previous findings, adds a new dimension to the genetic association between SLC11A1 and TB susceptibility, and strengthens the knowledge of the presumptive immune function of the SLC11A1 gene against MTB. Second, a DNA pooling approach was adopted for exploring novel genetic factors involved in susceptibility to TB. A streamlined screening strategy was designed to screen a large number of potential TB susceptibility candidate genes. The primer extension reaction and denaturing high performance liquid chromatography system were coupled and used for estimating allele frequencies of genetic markers in case and control DNA pools. In total, 157 single nucleotide polymorphisms (SNPs) from 50 candidate genes were selected based on the immunology of TB, previous gene expression studies in MTB infection and mouse model studies of MTB infection. After the preliminary screening, nine SNPs from nine candidate genes showed very significant differences (p <0.005) in estimated allelic frequencies between the case and the control pools (n = 268 cases and 275 controls, respectively), and were then typed by individual genotyping. Positive association was identified in the markers CXCL2-3Ur (rs9131) and SELE-5U (rs5353) in genotype comparison. For CXCL2-3Ur, the less common homozygotes were more susceptible to TB (OR = 1.94, 95% CI = 1.01 - 3.74; p = 0.0439). Conversely, for SELE-5U, the less common homozygotes showed a protective role (OR = 0.57, 95% CI = 0.36 - 0.89; p = 0.0139). With stratification by sex or/and age, significant differences were also consistently identified in the male and the young age (age <=65 years) groups for both SNPs. In particular, the most significant result (p = 0.0009) was obtained for the young age group for the allele comparison of SNP CXCL2-3Ur. No attempt was made to correct for multiple comparisons in this screening study because of its exploratory nature. As a result, both the E-selectin gene (SELF) and the chemokine (C-X-C motif) ligand 2 gene (CXCL2) were implied as TB susceptibility genes for the first time, and should be followed up by studying more SNPs in these genes, preferably with independent sample sets. Both gene products are responsible for the recruitment of neutrophils, which play a key role in immune function against TB in the early stage of infection. In conclusion, the present study confirmed the association of SLC11A1 gene in susceptibility to TB. A streamlined screening strategy based on DNA pooling identified two new TB susceptibility genes, SELE and CXCL2. This screening approach is a rapid exploratory strategy of screening for novel disease susceptibility genes.