Background
Tuberculosis (TB), caused by infection with
Mycobacterium tuberculosis (MTB), remains one of the world’s deadliest communicable diseases. According to the World Health Organization, an estimated 10.4 million people developed TB and 1.7 million died of the disease in 2016 [
1]. However, most individuals exposed to MTB experience latent MTB infection (LTBI) and do not develop active disease. Considerable evidence suggests that host genetic factors play a key role in determining an individual’s susceptibility to TB [
2‐
5].
Toll-like receptors (
TLRs) are a class of pattern recognition molecules, which are known to play important roles in the innate and adaptive immune system [
6], by recognizing pathogen-associated molecular patterns. Most
TLRs are expressed on the cell surface, whereas other
TLRs (3, 7, 8, and 9) are expressed intracellularly [
7,
8].
TLR7, TLR8, and
TLR9 have been implicated in immune diseases due to their ability to recognize oligonucleotide-based (RNA-and DNA-based) molecular patterns as agonists [
9].
TLR8 is located in the membranes of the endosomal compartment and recognize single-stranded RNA, regulating in the induction of interferon (IFN) and inflammatory cytokines [
10‐
12]. Previous studies have shown that
TLR8 variants influence the expression of
TLR8 [
13‐
15]. The
TLR8 single nucleotide polymorphism (SNP) rs3764880 (Met1Val) regulates the translation of the two main
TLR8 isoforms, and plays a significant part in the immune response [
12,
15]. A study conducted by Davila et al. [
14], was the first to demonstrate that SNPs in
TLR8 were associated with TB in adults. Since
TLR8 is located on chromosome X (Xp22.3-p22.2), males carrying a single copy of the defective allele may have higher risk of TB. In addition, several studies have shown that the G allele of
TLR8 rs3764880 was associated with TB susceptibility in males [
14,
16‐
18]. These studies demonstrated that the rs3764880 SNP in
TLR8 play important roles in TB.
The ligands for
TLR9 are DNA-containing CpG motifs [
19].
TLR9 is located in the endosomal compartment of plasmacytoid dendritic cells and monocytes/macrophages [
19], and plays a vital role in autoimmune diseases and inflammatory diseases by the regulation of type I IFN and inflammatory cytokines [
19‐
21]. The rs187084 and rs5743836 SNPs located in the promoter are the most important and have been associated with various inflammatory diseases [
9,
21‐
24]. Previous functional analyses have shown that both rs187084 and rs5743836 SNPs influence the transcription of
TLR9 by regulation of promoter activity [
22,
25,
26]. Previous studies indicated that
TLR9 is one of the most important receptors in the control of infections with pathogens such as hepatitis C virus [
22], Brucella [
27], and MTB [
23,
28,
29]. Some studies found that the rs187084 in
TLR9 showed no association with TB in Vietnam and Iran [
28,
30]. However, no study has explored the association between rs187084 and MTB infection or the process from LTBI to TB. The rs5743836 in
TLR9 showed a strong association with tuberculosis in African-Americans and Caucasians [
31], while the association was not found in Vietnam [
28] or Mexico population [
29]. Wu L et al. [
32] also reported that rs5743836 was a risk factor for LTBI.
These findings demonstrated that TLR8 and TLR9 play important roles in infectious diseases, and also emphasized the role of the rs3764880 SNP in TLR8 and rs187084 and rs5743836 SNPs in TLR9. To date, the SNPs of TLR8 and TLR9 have been studied in association with susceptibility to TB, but such studies addressing host genetic susceptibility to TB was limited, whether an association implies susceptibility for developing active disease or just acquisition of MTB infection is unclear. In this study, we investigated the associations of SNPs of TLR8 (rs3764880) and TLR9 (rs187084, rs5743836) with TB in a Chinese Han sample. Then we explored the associations of these SNPs with LTBI or PTB in a second sample of Chinese Han individuals.
Discussion
Multiple groups have investigated associations between the three SNPs which we examined in the in current study and TB susceptibility in a variety of populations, but the results were inconsistent [
14,
17,
18,
28,
30,
32,
36‐
40]. Few studies have classified LTBI and HC without MTB infection. The current study first analyzed the association between
TLR8/9 SNPs and TB in the first sample, and then further explored the
TLR8/9 variants with LTBI and active PTB in the second sample. Our results suggest that genetic variants might have different roles in the development of active PTB and LTBI.
Davila et al. [
14] for the first time found that four SNPs (rs3764879G/C, rs3788935G/A, rs3761624G/A, rs3764880G/A) in
TLR8 showed evidence of association with TB susceptibility with minor alleles showing an increased susceptibility to PTB in males in Russian and Indonesian populations. Furthermore, associations have also been found in Turkish male children [
17], Pakistan population [
41], and South African population [
18]. However, neither Kobayashi et al. [
42] nor Chimusa et al. [
43] showed any association between rs3764880 and TB susceptibility. Our results showed that the rs3764880 A allele in
TLR8 greatly reduced TB risk in males. This result was consistent with that in a Pakistan population [
41], which showed the rs3764880 A allele had a protective role against TB,but different from that in Turkish [
17], South African [
18], Russian and Indonesian populations [
14], which showed the rs3764880 A allele increased susceptibility to TB in males. Subsequently, our data in the second sample provided new evidence for our understanding of the role of rs3764880 in the development of LTBI and PTB. The data suggested that the AA-genotype of rs3764880 increased susceptibility to PTB among females, however, might decrease susceptibility to LTBI. The difference between our results and that of other studies may be due to differences in race, or living environment. Therefore, the detection of this SNP among LTBI subjects may provide important information in the assessment of their risk profiles for susceptibility to development of PTB. Previous studies have found that the SNPs in
TLR8 might have gender effects across the genetic association studies on TB susceptibility [
14,
17,
18,
41]. Our results also found a gender difference: male carriers of rs3764880 allele A showed a decreased risk for TB, and females carrying the AA genotype increased the risk of PTB when compared with LTBI. What’s more, this gender effect was both demonstrated in the two independent study samples.
Several investigators have studied the roles of
TLR9 SNPs in TB. As reported previously in studies of Vietnam and Iran [
28,
30], we found the rs187084 in
TLR9 showed no association with TB in the first sample. In the second sample, the G-allele of
TLR9 rs187084 greatly increased PTB risk, and this association was also observed under a dominant genetic model, which suggested a risk role of the allele G. However, the GA genotype decreased the risk of MTB infection. A study by Digna Rosa Velez et al. [
31] showed a strong association between rs5743836 and tuberculosis in African-Americans and Caucasians population, while the association was not found in Vietnam [
28] or Mexico population [
29]. Another study conducted by Wu L et al. [
32] showed that rs5743836 was a risk factor for LTBI and its MAF was 0.27 in a Chinese population in Shanghai city. The MAF of this SNP is high among African-Americans, Africans and Caucasians [
31], while low in the Mexican population [
29] and Vietnamese people [
28]. These studies, thus indicated that the MAF of rs5743836 may vary between different ethnic groups. In this study, we found the MAF of rs5743836 was less than 1%. Since our sample size may not be large enough for a SNP with low MAF, and may potentially lead to false associations with the phenotype investigated, the rs5743836 was excluded in the data analyses. The MAF of rs5743836 in Beijing Han population from the International HapMap Project (
http://www.1000genomes.org/) is less than 0.01, which support our result and indicates a low mutation rate in Chinese Han population. Therefore, the higher rs5743836 MAF in the study of Wu L et al. may suggest that populations other than Chinese Han were included in the study [
32].
It has been discussed that some SNPs may actually be associated with MTB infection [
36,
37], some may be associated with active TB [
38,
39] and other studies suggested some SNPs may have different impacts on the susceptibility to LTBI and PTB [
32,
40,
44]. Lu et al. found that an immunity-related GTPase family M SNP was associated with active TB and LTBI, but also plays opposite roles in the development of active TB and LTBI [
44]. Our results provide strong evidence that the allele “A” of the polymorphism rs3764880 was associated with decreased risk of tuberculosis in the first sample. The genotype “AA” of rs3764880 was more commonly found in female PTB patients compared with LTBI, indicating this genotype increases the possibility of progression of tuberculosis infection to disease in the second sample. For the polymorphism of rs187084 in
TLR9, we found the genotype “GA” and allele “G” was more common in PTB patients compared with LTBI, which demonstrated that this allele increased the risk of progression from tuberculosis infection to active disease. However, the GA-genotype was found to reduce the risk of MTB infection, thereby indicating that the results of our two samples were inconsistent and may have been caused by a number of factors. It may be caused by the following reasons. First, we included all TB patients in our first sample without differentiating between the different types of TB, while we only included PTB patients in the second sample. Genetic polymorphisms may have different effects on PTB and extra-pulmonary TB (EPTB), which may be due to the fact that different immune mechanisms are involved in PTB and EPTB [
45]. Second, the results in the first sample may be explained, when the distinction between LTBI and HC without the infection of MTB is ignored. In other words, the results may be different because of the differences within the control group. Therefore, it is reasonable that these SNPs showed different results in the two samples.
Our study faced some intrinsic limitations. First, small sample size in the second sample may have limited our ability to detect potential influence of
TLR SNPs on the susceptibility to both LTBI and PTB. Further studies are therefore necessary to validate these associations in larger sample sizes and other populations. Second, TB case of the first sample included both clinical diagnosed and bacteria confirmed TB patients, regardless of their types, which could potentially reduce the validity of our conclusions. Third, LTBI and HC subjects included in the second sample were asymptomatic contacts of bacteria-confirmed active TB patients, the assessment of exposure to an active TB case relied on self-reported behavior, which may not be accurate among participants and may cause misclassification bias (once exposed, those controls could become cases). Forth, it is likely that linkage disequilibrium patterns differed among other populations and the Han Chinese population in this study, and thus, those associations may not replicate exactly [
14,
28,
31]. A plausible explanation is that the SNPs that we identified are in linkage disequilibrium with another mutation that either confers a functional or a regulatory change within those genes. Hence, to interpret results across studies, future experiments should cover all common variation in the associated
TLR8/9 gene region in order to pinpoint the causal polymorphism. Furthermore, one potential limitation in our study was a lack of correction for multiple comparisons. The reported statistically significant results would need to be regarded as nominally significant.