1 INTRODUCTION

Tuberculosis, caused by Mycobacterium tuberculosis, is a global public health problem that seriously threatens human health. There are 10 million new tuberculosis cases worldwide each year, and this number has remained relatively stable in recent years. The 2019 World Health Organization (WHO) report on tuberculosis pointed out that 1.2 million people worldwide died of tuberculosis, one of the top 10 causes of death, and the number of deaths among human immunodeficiency virus HIV)-negative people was as high as 250,000 (WHO, 2019). The 2020 WHO report on tuberculosis indicated that in 2019, tuberculosis was still the most common cause of death caused by a single infectious pathogen. An estimated 10 million people worldwide had tuberculosis, and China has become the world's third-largest country with tuberculosis (8.4%; WHO, 2020). The incidence of tuberculosis in the Xinjiang Uygur Autonomous Region (China) increased from 180.8 cases per 100,000 in 2011 to 195.8 cases per 100,000 in 2015, with the highest burden in the southern region and with 76% of cases living in rural areas (X. He et al., 2017). The fourth epidemiological survey on tuberculosis in Xinjiang found that the epidemic of tuberculosis in southern Xinjiang was higher than that in northern Xinjiang, and the prevalence of tuberculosis in ethnic minority populations was much higher than that of the Han population (Cheng et al., 2018). Tuberculosis is highly contagious. Poverty, overcrowding, malnutrition, social and economic backwardness and HIV infection are the reasons for the high incidence of tuberculosis (Cui & Yu, 2020).

Inflammatory factors, including tumour necrosis factor α (TNF-α), interleukin (IL)-6, IL-17A, IL-17F and interferon γ (IFN-γ), play an important role in the control of M. tuberculosis infection (Domingo-Gonzalez et al., 2016). It is shown that the serum TNF-α level of pulmonary tuberculosis patients is higher than that of non-pulmonary tuberculosis patients (Mirzaei & Mahmoudi, 2018) and that TNFα can maintain the structural integrity of granuloma and promote the anti-mycobacterial effect in macrophages, thus playing a key role in the control of M. tuberculosis infection (Nie et al., 2020). Mycobacterium tuberculosis can interact with host cell receptors, including fragment crystallizable (FC) receptors, complement receptors and Toll-like receptors (TLRs), especially TLR2/4 (Jee, 2020). TLRs are pathogen recognition receptors and play a vital role in the recognition of M. tuberculosis. At present, 10 functional human TLRs (TLR1-TLR10) have been identified (Lee et al., 2012), and D299G (rs4986790) and T399I (rs4986791) in TLR4 are the most studied genetic variants of all TLRs. The methylation level of the Vitamin D receptor (VDR) gene in the Vitamin D metabolic pathway is related to the risk and prognosis of tuberculosis (Wang et al., 2018). IL-17A, a pleiotropic cytokine that can act on a variety of cells, is the earliest identified member of the IL-17 family (Xu et al., 2016). IL-17 plays an important role in the initial immune response after M. tuberculosis infection. Plasma level of IL-17 is high in sputum smear-positive pulmonary tuberculosis patients, but it is reduced after treatment and sputum smears become negative (Xu et al., 2016). Additionally, it has been reported that the serum IFN-γ level of pulmonary tuberculosis patients is significantly higher than that of healthy people, and it is closely related to changes in the patient's condition (Martínez-Morales et al., 2017).

Many studies have focused on the relationship between gene polymorphisms of inflammatory factors and susceptibility to tuberculosis. For example, the IL-17A rs3748067 allele C is reported to be related to the susceptibility to tuberculosis in Asian populations (J. Zhao et al., 2016). IL17F rs763780 can be used as a biomarker of susceptibility to tuberculosis in Argentina (Rolandelli et al., 2019). VDR rs731236, rs2228570 and rs1544410 are not related to tuberculosis susceptibility (Junaid et al., 2016). Similarly, there is no significant correlation between IFNGR1 rs1327474 and the risk/protection of pulmonary tuberculosis (Naderi et al., 2015). The research results of the relationship between the rs4986790 gene polymorphism in TLR4 and tuberculosis are inconsistent in different populations (Ortega et al., 2020; Thada et al., 2013; Zhou & Zhang, 2020; L. Zhao et al., 2015). However, no study on the association between gene polymorphisms of inflammatory factors and susceptibility to tuberculosis has been carried out in the Xinjiang Uyghur population.

2 MATERIALS AND METHODS 2.1 Study subjects The sample size was determined by using power and sample size. The formula is as follows: where nA is the sample size of case group; nB is the sample size of the control group; pA is the exposure rate of the case group; pB is the exposure rate of the control group; α is Type I error; β is Type II error, meaning 1−β is power; κ = nA/nBκ = nA/nB is the matching ratio; Φ is the standard normal distribution function and Φ−1 is the standard normal quantile function. Through pilot experiments, it was obtained that α = .05, β = .1, pA = 0.071 and pB = 0.024. Based on this, the sample size for this study should be at least 850 cases (nA = 425 and nB = 425).

Finally, 961 subjects were enrolled, including 507 cases in the case group and 454 cases in the control group. In detail, the patients with pulmonary tuberculosis treated at the First People's Hospital of Kashgar Prefecture and Kashgar Pulmonary Hospital in Xinjiang from June 2015 to June 2019 were selected as the case group (n = 507). The clinical data, including sex, age, region of residence, sputum smear, X-ray examination results and so forth were collected. Inclusion criteria: (1) All patients were diagnosed in accordance with the guidelines for the diagnosis and treatment of tuberculosis by the Chinese Medical Association Tuberculosis Branch; (2) patients did not receive anti-tuberculosis treatment before enrollment; (3) patients were all local Uyghur residents who had lived in the selected region for a long time. All patients were further grouped into those with active pulmonary tuberculosis and those with inactive pulmonary tuberculosis. Active tuberculosis was diagnosed based on (1) positive sputum smear or positive sputum culture or (2) exudative lesions, cavities or proliferative lesions in the lungs on chest X-ray (if smear or culture is negative). Exclusion criteria: Patients suffering from other lung diseases, malignant tumours, immunodeficiency diseases and genetic diseases, or taking immunosuppressive drugs were excluded. Healthy volunteers who underwent physical examinations in the same period were selected as the healthy control group (n = 454). They were all Uyghur residents who had lived locally for a long time. They had no pulmonary tuberculosis-related symptoms, pneumonia, asthma, bronchitis, chronic necrotic lung disease, lung cancer and other lung diseases, other malignant tumours or other chronic diseases.

This research was approved by the Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University. All subjects signed the written informed consent.

2.2 Single nucleotide polymorphism (SNP) genotyping

Peripheral venous blood (2∼4 ml each) was collected from all subjects, and DNA was extracted from blood using QIAamp DNA Blood Mini Kit (Qiagen Inc.). The improved multiplex ligase detection reaction (iMLDR) method was used for SNP genotyping. Briefly, the iMLDRTM multiple SNP typing kit (Center for Genetic & Genomic Analysis, Genesky Biotechnologies Inc.) was used for iMLDR. The multiplex polymerase chain reaction (PCR) system included DNA, 1x GC-I buffer (Takara), 3.0 mM Mg2+, 0.3 mM dNTP and 1 U HotStarTaq polymerase (Qiagen Inc.). The PCR reaction procedure was 95°C for 2 min; 11 cycles of 94°C for 20 s, 65°C for 40 s and 72°C for 15 min; 24 cycles of 94°C for 20 s, 59°C for 30 s and 72°C for 1.5 min and finally 72°C for 2 min. The multiplex PCR products were purified, ligated and sequenced on ABI3730X. The raw data were analysed with GeneMapper 4.1 (Applied Biosystems, Foster City, CA, USA). The SNP genotyping and data analysis were all performed by Genesky Biotechnologies Inc.

2.3 Screening of SNPs

According to the literature, the immune and inflammation-related genes that may play a role in the pathogenesis of tuberculosis were selected as candidate genes, including IL6, IL17, TNF, VDR, TLR4, haptoglobin-related protein, Par-3 family cell polarity regulator beta, purinergic receptor P2X, ligand-gated ion channel, 7 (P2RX7), IFN-gamma receptor 1 (IFNGR1) and cutaneous T-cell lymphoma-associated antigen 1 (CTAGE1). The SNPs of these candidate genes were screened with the HapMap database (http://www.hapmap.org) and the dbSNP database (http://www.ncbi.nlm.nih.gov/projects/SNP/). The polymorphic sites included promoter region, exon region and 3′ untranslated region. The screening criterion was minor allele frequency (MAF) ≥ 0.05.

2.4 Statistical analysis

(1) Hardy–Weinberg equilibrium analysis: Chi-square test was used to detect whether the genotype frequency of the case group and the control group conforms to the Hardy–Weinberg equilibrium. p > .05 was considered as Hardy–Weinberg equilibrium. (2) The Chi-square test was used to compare the allele frequency and genotype frequency, and the difference was statistically significant with p < .05. (3) The unconditional logistic regression was used to analyse the relationship between each SNP and tuberculosis under five different genetic models (dominant, recessive, codominant, additive and allele). The OR value and 95% CI were calculated. The Benjamini and Hochberg method was used for the correction of multiple comparisons and the false discovery rate was calculated. (4) Haplotype analysis: Through linkage disequilibrium analysis, the blocks with a strong correlation with tuberculosis were identified. Logistic regression analysis was performed on these blocks, and the association of SNPs with tuberculosis was obtained based on case/control information.

3 RESULTS 3.1 Basic information of subjects

A total of 961 subjects were enrolled, including 507 cases in the case group (171 cases of patients with active tuberculosis and 336 cases of patients with inactive tuberculosis) and 454 cases in the control group. There were no statistically significant differences in age, sex, region of residence and smoking history between the case group and the control group and between the subgroups (p > .05; Table 1).

TABLE 1. Basic data of enrolled subjects Tuberculosis cases (n = 507) Characteristics Active tuberculosis (n = 171) Inactive tuberculosis (n = 336) Controls (454) Age (years) 57.63 ± 17.90 55.84 ± 16.86 55.14 ± 18.03 Sex Male 79 143 194 Female 92 193 260 Region of residence Rural 137 251 340 urban 34 85 114 Smoking history Yes 35 62 89 No 136 274 365 3.2 SNPs screening and Hardy–Weinberg equilibrium analysis results

By searching the HapMap database and dbSNP database, we identified 12 SNPs of nine immune and inflammation-related genes related to tuberculosis, namely: rs361525 of TNF, rs2066992 and rs1524107 of IL6, rs3748067 of IL17A, rs763780 of IL17F, rs731236, rs2228570 and rs1544410 of VDR, rs1327474 of IFNGR1, rs3751143 of P2RX7, rs4331426 of CTAGE1 and rs4986790 of TLR4. Among them, the MAF of rs361525 and rs4331426 was less than 0.05 and thus they were not further analysed in the subsequent experiments. Finally, 10 SNPs were included. The polymorphic sites of SNPs on genes and the results of Hardy–Weinberg equilibrium are shown in Table 2.

TABLE 2. The position of SNP in the gene and the results of Hardy–Weinberg equilibrium SNP Gene name Region CHR Position (hg19) Allele MAF HWE HWE_Case HWE_Control rs361525 TNF Upstream 6 31543101 G/A 0.04 0.72 0.62 1 rs2066992 IL6 Intronic 7 22768249 G/T 0.32 0.37 0.34 0.75 rs3748067 IL17A UTR3 6 52055339 C/T 0.15 1 0.85 0.72 rs731236 VDR Exonic 12 48238757 A/G 0.21 0.55 0.48 0.89 rs1327474 IFNGR1 Upstream 6 137541075 T/C 0.25 0.73 0.90 0.81 rs763780 IL17F Exonic 6 52101739 T/C 0.07 1 0.74 0.71 rs1524107 IL6 Intronic 7 22768219 C/T 0.32 0.37 0.34 0.75 rs3751143 P2RX7 Exonic 12 121622304 A/G 0.20 0.11 0.21 0.32 rs2228570 VDR Exonic 12 48272895 G/A 0.34 0.23 0.49 0.35 rs4331426 CTAGE1 Intergenic 18 20190795 A/G 0.04 0.00 0.00 0.57 rs1544410 VDR Intronic 12 48239835 C/T 0.23 0.79 0.89 0.54 rs4986790 TLR4 Exonic 9 120475302 A/G 0.06 0.72 1 1 Abbreviations: CHR, the chromosome number; CTAGE1, cutaneous t cell lymphoma-associated antigen 1; HWE, Hardy–Weinberg equilibrium; IFNGR1, interferon-gamma receptor 1; IL6, interleukin 6; IL17A, interleukin 17A; IL17F, interleukin 17F; MAF, minimum allele frequency; P2RX7, purinergic receptor P2 × 7; Position (hg19), the location of SNP on the chromosome in version hg19; SNP, single nucleotide polymorphism; TLR4, Toll-like receptor 4; TNF, tumour necrosis factor; VDR, vitamin D receptor. p < .05 is indicated in bold. 3.3 Allele frequency distribution of SNPs and relationship of SNPs with tuberculosis in case and control groups

The relationship between each SNP and tuberculosis under five different genetic models (dominant, recessive, codominant, additive and allele) was analysed with unconditional logistic regression. As shown in Table 3, the distribution frequency of SNP rs1327474 C allele of IFNGR1 gene in case group was lower than that in control group (OR = 0.80, 95%CI = 0.65–0.99, p = .04), suggesting that it is associated with a reduced risk of tuberculosis. After adjustment for sex and age, logistic regression analysis showed that under the additive model, the genotype distribution of SNP rs1327474 of IFNGR1 was statistically different between the case group and the control group (OR = 0.80, 95%CI = 0.65–0.99, p = .04).

TABLE 3. Comparisons of gene polymorphisms between case and control groups Gene/SNP Genetic model Genotype Case, N (%) Control, N (%) OR 95%CI p-value FDR_BH adjusted

IL6

rs2066992

Codominant G/G 251(49.51) 205(45.15) 1 G/T 205(40.43) 198(43.61) 0.94 0.71–1.25 .67 0.87 T/T 51(10.06) 51(11.24) 0.84 0.54–1.31 .45 0.87 Dominant G/G 251(49.51) 205(45.15) 1 G/T-T/T 256(50.49) 249(54.85) 0.92 0.70–1.20 .53 0.94 Recessive G/G-G/T 456(89.94) 403(88.77) 1 T/T 51(10.06) 51(11.23) 0.86 0.56–1.30 .50 0.93 Additive 0.92 0.76–1.12 .43 0.86 Allele G 707(69.72) 608(66.96) 1 T 307(30.28) 300(33.04) 0.88 0.73–1.07 .19 0.92

IL17A

rs3748067

Codominant C/C 373(73.57) 326(71.81) 1 C/T 125(24.65) 116(25.55) 0.94 0.69–1.27 .68 0.61 T/T 9(1.78) 12(2.64) 0.51 0.238–1.27 .15 0.61 Dominant C/C 373(73.57) 326(71.81) 1 C/T-T/T 134(26.43) 128(28.19) 0.89 0.66–1.20 .45 0.94 Recessive C/C-C/T 498(98.22) 442(97.36) 1 T/T 9(1.78) 12(2.64) 0.52 0.21–1.29 .16 0.83 Additive 0.86 0.66–1.13 .28 0.86 Allele C 871(85.90) 768(84.58) 1 T 143(14.10) 140(15.42) 0.90 0.70–1.16 .42 0.99

VDR

rs731236

Codominant A/A 328(64.69) 271(59.69) 1 A/G 163(32.15) 161(35.46) 0.81 0.61–1.07 .14 0.71 G/G 16(3.16) 22(4.85) 0.67 0.33–1.36 .27 0.71 Dominant A/A 328(64.69) 271(59.69) 1 A/G-G/G 179(35.31) 183(40.31) 0.79 0.60–1.04 .10 0.52 Recessive A/A-A/G 491(96.84) 432(95.15) 1 G/G 16(3.16) 22(4.85) 0.72 0.36–1.45 .36 0.93 Additive 0.81 0.64–1.03 .08 0.44 Allele A 819(80.77) 703(77.42) 1 G 195(19.23) 205(22.58) 0.82 0.66–1.02 .07 0.64

IFNGR1

rs1327474

Codominant T/T 299(58.97) 240(52.86) 1 T/C 180(35.50) 179(39.43) 0.79 0.60−1.04 .09 0.61 C/C 28(5.53) 35(7.71) 0.66 0.38–1.15 .14 0.61 Dominant T/T 299(58.97) 240(52.86) 1