Association between cytokine gene polymorphisms and tuberculosis in a Chinese population in Shanghai: a case–control study

Tuberculosis (TB) remains to be a leading cause of morbidity and mortality in developing countries, and the incidence of the disease is increasing in developed countries [1]. One third of the world’s population is infected by Mycobacterium tuberculosis (Mtb), but only about 5% of infected individuals develop the disease within the first year of infection and another 5% develop the disease later in life. Candidate gene and association studies have identified various host genetic factors that play significant roles in susceptibility to TB [2,3]. Identifying the host genes responsible for susceptibility and resistance to TB may lead towards better understanding of the pathogenesis of TB and development of prophylactic or treatment strategies.

The cytokines produced at the site of disease after the interaction between T lymphocytes and infected macrophages are essential for the pathogenesis of TB [4]. The course of Mtb infection is regulated by two distinct T-cell cytokine patterns. T-helper 1 (Th1) cytokines, interleukin(IL)-2 and interferon γ (IFN-γ), are associated with resistance to infection, whereas T-helper 2 (Th2) cytokines, IL-4 and IL-13, are associated with progressive disease [5]. Polymorphisms in several cytokine genes have been described and shown to influence gene transcription, leading to inter individual variations in cytokine production [6]. Cytokine gene polymorphisms have been shown to be involved in the susceptibility, the severity and clinical outcomes of several diseases, including infectious diseases [7]. A single nucleotide polymorphism (SNP) located in the first intron of human the IFN-γ gene (+874 T⁄ A) putatively influences the secretion of the cytokine, with an impact on the outcomes of infections. This polymorphism displays variable associations with TB disease susceptibility and severity [8] in some ethnic populations but not in others [9]. A specific nucleotide polymorphism within the promoter region of IL-10 (1082A⁄G) is also highly variable. This polymorphism alters the level of this cytokine and possibly the Th1/Th2 balance, with major implications for TB infection [6]. Data on the association between the 1082G/A SNP of IL-10 and TB are also highly variable. Among the proinflammatory cytokines, tumor necrosis factor α (TNF-α) SNP −308 is associated with either TB susceptibility or severity [10]. Furthermore, because the immune interactions that determine TB disease susceptibility and severity are complex, multigene combinations may give more meaningful information than single SNPs in determining disease outcomes [11].

Most studies have not differentiated TB disease according to its natural history. Therefore, it is difficult to evaluate the differences in their results. Our study groups were stratified into latent TB infection (LTBI) and active TB disease. We investigated the influence of IFN-γ, TNF-α, and interleukin cytokine gene polymorphisms on the susceptibility of individuals to LTBI and TB disease in terms of the cytokine levels expressed and their associations with ex vivo stimulate cytokine and pulmonary TB in a population from Shanghai. We also analyzed combinations of cytokine gene polymorphisms.

SNPs in several candidate genes have been linked to an increased risk of TB [12]. In this polymorphism association study, we investigated the significance of the relationship between several cytokine gene polymorphisms, their plasma products, and susceptibility to TB in a Chinese population.

The IFN-γ polymorphism is the best-studied polymorphism in terms of its association with TB disease sites and severity. However, previous reports have been contradictory insofar as the −874A allele was more common in patients with TB and the T allele was more common in the controls in Italian [13], South African [14], and Spanish populations [15], whereas in China, one study reported no association [16] while a later study of the same population showed an association between TB and the A allele [17]. Our results are consistent with the latter study. In the present report, we show that the A allele at IFN-γ −874A/T was significantly overrepresented in the populations of LTBI and PTB patients, with A allele frequencies 1.30-fold and 1.76 fold higher, respectively, than those in the HC. Significantly reduced IFN-γ production by stimulated PBMCs was observed in individuals with the A allele at IFN-γ −874. Similar genotype distributions and associations were observed for the t875 CA repeats (non-12/12) as at −874A/T in our study, and the non-12 CA repeats allele was absolutely linked to the −874 A allele in our population, as previously reported in a UK population. Because the −874 A allele corresponds to lower IFN-γ expression, the lower IFN-γ levels may impair the activation of macrophages, resulting in the development of TB. These observations indicate that the A allele at IFN-γ −874, or another nearby genetic variation linked to it, increases the host’s susceptibility to TB.

Because the modulation of the T-cell response by IL-10 seems to influence the susceptibility of the host to TB infection [18], identifying the polymorphisms in the IL-10 gene may be useful in predicting disease susceptibility. The frequency of the IL-10 − 1082G allele in our population (4%) was similar to that in Korean (7.4%) [19] and Japanese (6.5%) populations [20], but was significantly different from that reported in Caucasians (48%) [21]. The present results also show that the genotype frequency of the IL-10 − 1082A allele was significantly increased in TB patients compared with the LTBI subjects, especially when it was associated with the IFN-γ +874A allele (OR = 3.59; p = 0.045). This is consistent with the low levels of mycobacterial-stimulated IFN-γ and IL-10 in PTB patients. In fact, there remains a discrepancy with the −1082 polymorphism in the IL10 promoter, for which no association with TB was reported in Gambian [22] and Korean populations [23]. However, an association between the IL-10 − 1082A allele and TB was observed in an Italian (Sicilian) [24] population, and GA heterozygosity was associated with PTB in Cambodia [25]. These observations may reflect ethnic-specific genetic variations or suggest that other more-distal promoter elements are involved.

A number of studies have investigated the association between the TNF-α 308 (rs1800629 G/A) polymorphism and susceptibility to TB in different populations. However, the association between TB with TNF-α differs with ethnicity, and many of these studies have produced inconsistent results [26]. In the present study, a significant negative association between TB and the TNF-α 308 GA genotype and a significant positive association with the GG genotype were found, and the G allele was carried by 90% of patients and by 85.8% of controls. The A allele is associated with high TNF-α production and the G allele is associated with low TNF-α production. The high frequency of the genotype associated with increased protein production may be attributable to selection during the course of evolution, increasing the frequency of the most advantageous genotype. Our data indicate that TNF-α production was reduced in our PTB patients. Although the study by Scola et al. demonstrated a reduction in the low-producing −308 G/G TNF-α homozygous individuals in the affected group [24], and the study of Bikmaeva et al. revealed an association between the high-producing A allele and a higher risk of PTB [27], although several studies have failed to confirm the association between TNF-α and TB [28].

In the present study, we also analyzed several other candidate gene polymorphisms and their association with the susceptibility to PTB or LTBI. We found a significant negative association between the −330 TT genotype of IL-2 and PTB; the T allele was significantly overrepresented in the LTBI group and the HC relative to the PTB group. This suggests that the GG haplotype is associated with a susceptibility to PTB. Although translation studies have demonstrated the increased activity of the G allele, allele expression studies have shown higher expression of IL-2 in the PBMCs of individuals with the IL-2 − 330 TT genotype than in those of individuals with the GG genotypes, suggesting that the IL-2 promoter activity is also influenced by other polymorphisms, probably +160 T/G [29]. No significant association between TB and IL-4 levels was observed for the VNTR polymorphism. The results of the present study do not preclude the possible influence of the VNTR polymorphism on the IL-4 response because IL-4 was undetectable in most of the subjects tested, which could be attributable to the downregulation of the Th2 responses by the IFN-γ secreted in response to the antigens used. In our study, the IL-6 − 174 CC genotype was significantly negatively associated with the LTBI group and with the reduced production of IL-6. In previous studies, significant levels of IL-6 were shown to be produced in response to Mtb infection. Human and murine macrophages secrete IL-6 in response to Mtb in vitro [30], and elevated concentrations of IL-6 are present in the plasma of patients with TB [31]. It has also been reported that the induction of IL-6 activity by Mtb inhibits IFN-γ signaling, so the bacterium evades eradication by the cellular immune response [32]. We have reported that the IL-12 AA genotype is associated with LTBI susceptibility. This is consistent with reports from Japan and France, where polymorphisms in the IL-12 receptor β1 gene (IL-12RB1) are known to influence the susceptibility to and severity of TB [33].

A consensus seems to be emerging that the combined effects of several cytokine SNPs may play a crucial role in susceptibility to LTBI and TB diseases [29]. The combination of the IFN-γ polymorphism with other cytokine polymorphisms was most effective in predicting LTBI. This observation is consistent with the recently described functions of IFN-γ in a mouse model, in which mice with a disrupted IFN-γ gene were unable to control Mtb infection [34]. The combination of IL-10 polymorphisms with other cytokine polymorphisms may predict a susceptibility to TB. These results also substantiate an earlier report [35] that implicated IL-10 in the reactivation of TB in humans and in TB disease in a mouse model. A recently published meta-analyses reinforced the critical importance of IFN-γ +874 T/A as a genetic marker of TB resistance [8], and stressed that although IL-10 has a specific effect on the form and severity TB, it does not affect susceptibility per se. Our results are consistent with these findings. BCG vaccination cannot be ignored [36]. However, we found no evidence that BCG vaccination interacts with gene polymorphisms in affecting susceptibility to TB.

In this study, all cases of TB disease were included, but the majority of patients had PTB. Therefore, it was not possible to demonstrate an association between the system involved and cytokine polymorphisms. A larger series is required to test the possible associations between cytokine gene polymorphisms and organ preference in TB, and the clinical severity of the disease.