Polymorphism in the EREG gene confers susceptibility to tuberculosis

In this case-control study, 1224 subjects were recruited: 600 healthy controls (HC), 424 pulmonary TB patients (PTB) and 200 extra-pulmonary TB patients (EPTB). All volunteers were enlisted from the Shanghai Pulmonary Hospital. Members of the control population were > 18 years of age and attested to no history of TB; their PPD tests and QFT tests were negative, and no evidence of prior TB presented in the chest radiographies. There were 340 males and 260 females, and the mean age was 34.66 ± 9.70. Mtb infections were confirmed in the TB patients included according to evidence of positive sputum smears and cultures, as well as clinical and radiography features. In the PTB groups, there were 250 males and 174 females, and the mean age was 35.44 ± 13.65. In the EPTB groups, there were 121 males and 79 females, and the mean age was 35.63 ± 17.22; there were 13 patients with intestinal tuberculosis, 10 patients with bone tuberculosis, 16 patients with lymph node tuberculosis, 60 patients with meningeal tuberculosis, 26 patients with genital tuberculosis, 64 patients with pleurisy tuberculosis, and 11 patients with renal tuberculosis, as shown in Table 1.

All subjects volunteered for the study and signed informed consent forms. In addition, the Ethics Committee of Shanghai Pulmonary Hospital affiliated with Tongji University School of Medicine approved the study.

Thirty healthy controls and 50 TB patients, who were not selected from the genotyped individuals, were picked randomly to have their plasma separated from 200 μl EDTA anticoagulated blood by centrifugation. Following the manufacturer’s protocol, ELISA kits (R&D systems, MN, USA) were used for detecting EREG levels.

We selected 5 SNPs from EREG (rs10518126, rs2367707, rs3806794, rs6446993, rs6836436), and the tag SNPs were chosen from the 1000 Genomes Project Phage3. The general rule for selecting tagged SNPs were an R² linkage disequilibrium of > 0.8 and a minor allelic frequency of > 0.1. PCR primers were designed with Primer 3 software (http://​bioinfo.​ut.​ee/​primer3-0.​4.​0/​). The genetic information and the primers are shown in Table 2.

We amplified all of the fragments by a Multiplex PCR reaction. The amplification system included 1 × HotStar Taq buffer, 0.3 mM dNTP, 3.0 mM Mg²⁺, 1 U HotStar Taq polymerase, 0.1 μM primer and 1 μl of DNA template. The reaction program carried out was as follows: 95 °C for 2 min; 11 cycles of 94 °C for 20 s, 65 °C ± 0.5 °C /cycle for 40 s, and 72 °C for 1 min 30 s; 26 cycles of 94 °C for 20 s, 59 °C for 30 s, and 72 °C for 1 min 30 s; and finally 72 °C for 2 min. Then, we added 5 U of shrimp alkaline phosphataseand 2 U of Exonuclease I to 15 μl of PCR product for purification. Next, we incubated the mixture at 37 °C for 60 min and then at 75 °C for 15 min. We performed the extension reaction with a SNaPshot® Kit, and a SNaPshot genotyping system was optimized sufficiently and validated by sequencing. The extension primers used to detect polymorphisms in the EREG genes were rs2367707SF (TTTTTTTTCCGTCCACCAACCTTTAAGCAAAGA), rs10518126SR (TTTTTTTTTTTGGGCCACTTTATAGAATTTGGAAATA), rs6836436SF (TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTCCCTTCTAGGCTGACAGCCGC), rs3806794SF (TTTTTTTTTTTTTTTTTTTTTTTAATAACAGGAATTTTCTTCACAATGACT), and rs6446993SR (TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTAACAGAAAGGCTATTTAGAAAA). The extension reaction mixture was as follows: 5 μl of SNaPshot mix, 2 μl of the extension primer mix, 2 μl of purified PCR product and 2 μl of ultrapure water. The extension procedure was as follows: 96 °C for 1 min and 28 cycles of 96 °C for 10 s, 52 °C for 5 s, and 60 °C for 30 s. Then, we added 1 U of SAP to the extension products and incubated the mixture at 37 °C for 60 min followed by 75 °C for 15 min to purify the extension products. Then, we denatured the mixture using 1 μl of purified extension product or 0.5 μl of purified linkage product, 9 μl of HiDi formamide and 0.5 μl of the Liz120 size standard (Applied Biosystems, Foster City, USA) at 95 °C for 5 min. Finally, the mixture was loaded on an ABI 3730XL DNA Analyser (Applied Biosystems, Foster City, USA), and the results were analyzed using GeneMapper.

The EREG levels were represented as the mean ± standard deviation, and the data between healthy controls and TB patients were analyzed with the t-test using SPSS 22.0. For each individual genotype, the OR and 95% confidence interval (CI) were estimated using logistic regression models adjusted for gender and age.

The Chi-squared test was used to evaluate whether the genotypic frequency in the population was consistent with the HWE. Data were confirmed to be within population equilibrium by calculating the HWE. We performed Chi-squared and Fisher’s exact tests to compare allelic and genotypic frequencies. The haplotypes and linkage disequilibrium in the HC and TB patients were analyzed by the SHEsis system. After Bonferroni correction, P*-values were five times the observed P-values, and P values < 0.05 were considered statistically significant.

The plasma EREG levels (mean ± SD) were 1014 ± 733.9 pg/ml in PTB group and 700.2 ± 676.6 pg/ml in EPTB group and were significantly different from those of the healthy control group (277 ± 105.4 pg/ml) (Fig. 1).

The allelic and genotypic frequencies of EREG polymorphisms in HC and PTB patients are shown in Table 3. We investigated the association of five EREG SNPs and PTB susceptibility using a case-control study. In the tested populations, all polymorphisms were consistent with Hardy-Weinberg equilibrium. Analyses of the distributions of PTB patients and healthy controls demonstrated that the EREG rs2367707 SNP contributed to susceptibility to PTB. The frequencies of the A allele of rs2367707 were significantly higher in HC than in PTB patients (P = 0.00019, OR (95% CI) =1.49 (1.23–1.81)).

Comparing the genotypic frequencies between healthy controls and PTB patients, we found that rs2367707 A/G was more common in PTB patients (P = 0.00051). For the distribution of the other four EREG SNPs between PTB patients and controls, no significant differences were observed.

The allelic and genotypic frequencies of EREG polymorphism in EPTB patients and healthy controls are indicated in Table 4. Analyses of the distributions of EPTB patients and healthy controls demonstrated that the EREG rs2367707 and rs6446993 SNPs contributed to susceptibility to EPTB. The frequencies of the A allele of rs2367707 and the T allele of rs6446993 were significantly higher in healthy controls than in EPTB patients (P = 0.00029, OR (95% CI) =1.64 (1.28–2.08); P = 0.00087, OR (95% CI) =1.54 (1.23–1.94)).

Comparing genotypic frequencies between healthy controls and EPTB patients, we found that rs2367707 A/G and rs6446993 A/T were more common in EPTB patients (P = 0.0012, P = 0.0069).

The haplotype data between the two disease forms are shown in Table 5 and Table 6. We identified 11 haplotypes with SHEsis, and seven were excluded with a frequency < 0.03. Haplotype analysis showed that the EREG CACAT and CGCTT haplotypes were significantly related to PTB and EPTB and that the CACAT haplotype was a significantly “beneficial” haplotype (P = 0.00031 for PTB, P = 0.000053 for EPTB); the other two haplotypes showed no significant association.

We assessed the LD among SNPs with both D’ and r². The LD structure constructed with five SNPs (rs10518126, rs237707, rs3806794, rs6446993, and rs6836436) is shown in Fig. 2. The degree of LD among the five SNPs was relatively high.