Liver X Receptor Gene Polymorphisms in Tuberculosis: Effect on Susceptibility

Objectives The Liver X receptors (LXRs), Liver X receptor A (LXRA) and Liver X receptor B (LXRB), regulate lipid metabolism and antimicrobial response. LXRs have a crucial role in the control of Mycobacterium tuberculosis (M.tb). Lacking LXRs mice is more susceptibility to infection M.tb, developing higher bacterial burdens and an increase in the size and number of granulomatous lesions. We aimed to assess the associations between single nucleotide polymorphisms (SNPs) in LXRs and risk of tuberculosis. Methods We sequenced the LXRs genes to detect SNPs and to examine genotypic frequencies in 600 patients and 620 healthy controls to investigate for associations with tuberculosis (TB) in the Chinese Han population. DNA re-sequencing revealed eight common variants in the LXRs genes. Results The G allele of rs1449627 and the T allele of rs1405655 demonstrated an increased risk of developing TB (p<0.001, p = 0.002), and the T allele of rs3758673, the T allele of rs2279238, and the C allele of rs1449626 in LXRA and the C allele of rs17373080, the G allele of rs2248949, and the C allele of rs1052677 in LXRB were protective against TB patients compared to healthy controls (p = 0.0002, p = 0.006, p<0.001, p = 0.004, p = 0.008, p = 0.003, respectively). All SNP genotypes were significantly associated with TB. An estimation of the frequencies of haplotypes revealed two potential risk haplotypes,GGCG in LXRB (p = 0.004,) and TTCG in LXRA (p<0.001, p = 0.004). Moreover, three protective haplotypes, TTAT and CCAT in LXRA and CATC in LXRB, were significantly “protective” (p = 0.008, p<0.001, p = 0.031) for TB. Furthermore, we determined that the LXRs SNPs were nominally associated with the clinical pattern of disease. Conclusions Our study data supported that LXRs play a fundamental role in the genetic susceptibility to TB and to different clinical patterns of disease. Thus, further investigation is required in larger populations and in additional areas.


Introduction
Approximately one-third of the population is infected with Mycobacterium tuberculosis (M.tb) worldwide, which is the main etiological agent of tuberculosis (TB). M.tb is a prototypic intracellular macrophage pathogen, and resides for a long period of time to avoid recognition of the human immune system during clinical dormancy. The host protective immunity plays a pivotal role in the primary immune response and active immune response, which is central in inflammation processes.
Liver X receptors (LXRs), Liver X receptor a (LXRA, NR1H3) and Liver X receptor b (LXRB, NR1H2), are a subset of nuclear receptor transcription factors that have emerged as master regulators of lipid [1,2], inflammation [3,4] and glucose metabolism [5,6,7,8,9]. Recently, a large body of literature has also indicated a role of LXRs in inflammation. LXRs, which are expressed in many immune cells, such as the dendritic cell [10,11], macrophage [12], and Lymphocyte, regulate network of genes that affect many aspects of the immune system [13,14].
In a mice study, Hannelie et al. reported that mice deficient in LXRs were more susceptible to M.tb infection, and demonstrated a decrease clearance of bacterial load in the lungs. These data support the crucial role of LXRs in the control of mycobacterial infection [13]. So we hypothesized that LXRs might also influence immunity of human. The aim of this case-controlled association study was to evaluate if LXR gene single nucleotide polymorphisms (SNPs) contribute to the susceptibility of tuberculosis in a Chinese Han population. In addition, we also investigated the role of LXR gene variants in two groups of tuberculosis subjects with different clinical patterns; pulmonary tuberculosis patients and extra-pulmonary tuberculosis patients.

Subjects
As a control population, we included 620 individuals (252 females, 368 males) who were older than 18 years of age and did not demonstrated inflammatory, autoimmune disease or a history of chronic infectious disease. These patients had no history of tuberculosis and no evidence of prior tuberculosis noted using chest radiography. Individuals of the control group were matched with the case group according to age, sex and ethnic.
A total of 600 patients (278 females, 322 males) were recruited from Shanghai Pulmonary Hospital in China. All patients were diagnosed on the basis of radiographic and clinical presentation, and demonstrated a positive smear and culture for M.tb. According to the different location of infection, the case subjects were then further divided into two subgroups: (1) pulmonary tuberculosis patients (PTB, n = 424) and (2) extra-pulmonary patients (EPTB, n = 176), as indicated in Table 1.

Ethics Statement
All controls and patients were Chinese, and all subjects were HIV-negative. Written consent was obtained from all subjects, and the Ethics Committee of the Shanghai Pulmonary Hospital Affiliated to Tongji University School of Medicine, China, provided authorization for the study.

Screening for DNA Sequence Variants in LXR Genotyping
PCR primers were designed using Primer 3 software (http:// frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi). The primers are shown in Table 2. To determine the frequency of known SNPs and potential functional SNPs in LXRs, the coding exons and sequence promoter regions were amplified and sequenced in 24 healthy controls and 24 patients, using previously described primers. Sequencing reactions were performed using Big-Dye v1.1 (Applied Biosystems, Foster City, USA). The sequencing data were aligned using Sequencher 4.2 (Gene Codes Corporation, Michigan, USA).
Genomic DNA was extracted from 0.2 ml EDTA-anticoagulated blood samples using the QIAamp DNA Blood Mini Kit (Qiagen) according to the manufacturer's protocol. Genotyping of single nucleotide polymorphisms (SNPs) was performed using the SNaPshot kit purchased from Applied Biosystems. This is a rapid and robust assay used to simultaneously genotype SNPs using single nucleotide primer extension (mini-sequencing) and can obtain population genetic data.
A 20 ml mixture was prepared to amplify all fragments in a Multiple PCR reaction and included 1x HotStarTaq buffer, 2.8 mM Mg 2+ , 0.3 mM dNTP, 0.1 mM of each primer, 1U HotStarTaq polymerase (Qiagen Inc.) and 1 ml template DNA. The cycling program was the following: 95uC for 15 min; 11 cycles of 94uC for 20 s, (62uC 20.5uC/cycle) for 40 s, 72uC for 1.5 min; 26 cycles of 94uC for 20 s, 56uC for 30 s, 72uC for 1.5 min; and 72uC for 5 min.
After the completion of multiple amplification, we used shrimp alkaline phosphatase (SAP) and ExonucleaseI (ExoI) to purify the PCR product. Next, 2U SAP and 1U ExoI were added to 5 ml of PCR product. The mixture was incubated at 37uC for 80 min, followed by incubation at 75uC for 15 min. Next, the purified PCR products were used as templates for the mini-sequencing reaction using the commercially available SNaPshot Kit. To detect polymorphisms, we used the following mini-sequencing extension primers rs2248949SF: TTTTCAGGGGCTTTGGGAAGAGG, rs3758673SF: TTTTTTTTCCCTATCCCAGGCCTGCCA, rs2279238SR: TTTAGGCCTTGTCCCCACACAC, rs1405655SF: TTTTTTTTTTTTCTGAGAAAATTGAGGAT-CAGGC, rs1449626SF: TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTGATC-CGCCTCCAGGCCCAG, rs1449627SR: TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-TCAGATGCCAAAGGAAAA, rs17373080SF: TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-TTTTTTTCGCATGAATGAGCTAAAGCCA, rs1052677SF: TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT-TTTTTTTTTTTTTTTTTATGGCTCTCCCCCCTAGCC. The SNaPshot Extension Reaction mixture included 5 ml SNaPshot mix, 2 ml purified PCR product, 2 ml ultrapure water and 2 ml extension primer mix (0.8 mM for each extension primer). The cycling protocol was the following: 96uC for 1 min; 28 cycles of 96uC for 10 s, 52uC for 5 s, and 60uC for 30 s. To purify the extension products, we added 1 U SAP to the extension products, incubated at 37uC for 60 min and then incubated at 75uC for 15 min. One microlitre of purified extension product was then mixed with 9 ml HiDi Formamide and 0.5 ml Liz120 (Applied Biosystems) size standard and denatured at 95uC for 5 min. Next, we loaded the mixture on the ABI 37306l DNA Analyser (Applied Biosystems).

Statistical Analysis
The allele and genotypic frequencies were determined by direct quantification. The distribution of the LXR SNPs was compared between TB patients and healthy controls using the X2 or Fisher's exact test. The data were analyzed for appropriateness between the observed and expected genotypic values, and then evaluated for fit in Hardy-Weinberg equilibrium (HWE). HWE was calculated to ensure that all of the data were within population equilibrium. The SHEsis system (http://analysis.Bio-x.cn/ SHEsisMain.htm) was used to analyze the haplotypes and determine linkage disequilibrium (LD) in the patient cases and controls. This system used a Full-Precise-Iteration (FPI) algorithm in haplotype reconstruction and frequency estimation between the patient and control [15]. The haplotypes were automatically selected. P values ,0.05 were considered statistically significant.  To account for multiple comparisons, a Bonferroni correction was applied in this study so that the corrected P-values were seven times the observed P-values.

Identification of SNPs in LXR Genes
The tagging SNPs were selected on the basis of information provided by public data-bases (http://www.ncbi.nlm.nih.gov/ entrez/query.fcgi?db = snp) and re-sequencing results, using reasonable inclusion criteria. These SNPs occurred at allele frequencies of $0.1, according to the international Hap Map project data for Chinese populations. Three SNPS were chosen using the binning algorithm of LDSelect (http://droog.gs. washington.edu/IdSelect.pI) at c 2 thresholds of 0.5 and minor allele frequencies of $0.1.
All non-synonymous exonic variants and any variants that suggested differences in frequencies between patients and controls were included. After considering the LD structure based on Hap Map Han Chinese and pairwise LD D' and r 2 values, we selected the total of eight representative SNPs across the genes; four genes in LXRA and four genes in LXRB ( Table 2).

Analysis of Association between the LXR Allele and the Genotype and TB in Patients and Controls
The allele and genotypic frequencies of the eight LXR SNPs in TB patients and control subjects are shown in Table 3 and Table 4. Both populations were consistent with HWE for all polymorphisms tested (p.0.05).
Analyses of the contribution of LXR gene allele frequencies revealed that the G allele of rs1449627 in LXRA and the T allele of rs1405655 in LXRB were more common in TB patients compared healthy controls (p,0.001, OR (95%CI) = 1.62(1.38-1.92); p = 0.002, OR (95%CI) = 1.46 (1.14-1.87), respectively); the T allele of rs3758673, T allele of rs2279238, and C allele of rs1449626 in LXRA and the C allele of rs17373080, G allele of rs2248949, C allele of rs1052677 in LXRB were less in TB patients compared to healthy controls (p = 0. Analysis of LXRs genotype frequencies demonstrated that the rs3758673 T/T, rs2279238 T/T, rs1449627 G/G, and rs1449626 C/C in LXRA were significantly more common in TB patients than healthy controls (p,0.001, p = 0.014, p,0.001 and p = , 0.001, respectively). In the case of LXRB gene, rs17373080 C/C, rs1052677 C/C, rs2248949 G/G and rs1405655 T/T genotype variants were associated with TB (p = 0.008, p = 0.005, p = 0.008 and p = 0.01, respectively). Statistical significant differences in the allele (G allele of rs2248949), and genotypes (rs2279238 of LXRA, and rs17373080, rs1405655, rs1052677 of LXRB), however, were lost after Bonferroni correction of the p values (Pc).

Linkage Disequilibrium (LD) and Haplotype Analyses
The LD pattern across SNPs was assessed using both D' and r 2 , and the LD structure (LXRA and LXRB) was shown in Fig. 1 and Fig. 2. Thus, the haplotypes were constructed with all eight variants and the frequencies of the haplotypes were analyzed by excluding the rare haplotypes (those with a frequency ,0.03 in cases or controls). This revealed a significant difference in the distribution of the global haplotypes between the TB group and healthy controls (p = 0.001). SHEsis identified four haplotypes for markers genotyped in LXRA and two haplotypes for makers in LXRB ( Table 5). The degree of LD over the genomic region containing LXRs was high. Moreover, the respective gene generated one haploblock ( Fig. 1 and Fig. 2).

Comparison of LXR Data between Pulmonary TB Patients and Extra-pulmonary TB Patients
To further understand the association between LXR SNPs and the risk of different clinical patterns, we evaluated the SNP contribution on clinical variables related to tuberculosis. No significant association was detected between the allele of LXRs and the risk of different clinical forms of tuberculosis (date not shown). However, the genotype at the LXRA SNP rs1449627 and LXRB SNPs rs17373080, rs1405655 and rs1052677 was associated nominally associated to PTB, p = 0.006, p = 0.013, p = 0.004 and p = 0.0007, respectively. We also performed haplotype analyses using the LXR SNPs in two clinical groups. Seven haplotypes were observed in samples of two subgroups, but only the CATC haplotype in LXRB demonstrated a significant association with an increase in the risk of PTB (p = 0.045 OR (95%CI) = 1.511 (1.007-2.269)). After Bonferroni correction, CATC haplotype and LXRB SNPs rs17373080 were lost significant associated with the different clinical forms of the disease.

Discussion
Tuberculosis, a chronic inflammatory disease that occurs within the entire body, is a complex process, and can cause multi-system complications, such as pulmonary, osteoarticular, and renal, central nerve systems (CNS). The initiation and progression of tuberculosis involves a dynamic interplay between numerous signaling pathways and cell types, where macrophages, dendritic cells (DCs), T cells, and other cellular elements present in the infection sites contribute to its pathogenesis. Many studies have investigated the potential role of LXRs in various bacterial  infections, such as Listeria monocytogenes [16], Leishmania [17], Mycobacterium tuberculosis [13]. However, in different infections, the phenotypic effect is more discriminatory, in which a multiplicity host-association is present between the pathogen and host immune system.
We have analyzed eight common genetic variations in LXRs, LXRA and LXRB, for an association with tuberculosis in a Chinese Han population and for an association with different disease clinical patterns in two subgroups of tuberculosis patients. We assessed the hypothesis that sequence variations in the LXR gene may contribute to TB susceptibility. We screened LXR genes for sequence variations in 24 healthy controls and 24 TB patients, resulting in eight representative SNPs across genes, four SNPs for each respective gene.
In this study, we observed an association of LXR SNPs with TB. Frequencies of the allele and genotype in the LXR genes were different between the healthy controls and TB patients. The G allele of rs1449627 in LXRA and the T allele of rs1405655 in LXRB were more common in TB patients compare to healthy controls. The T allele of rs3758673, the T allele of rs2279238, and the C allele of rs1449626 in LXRA and the C allele of rs17373080, the G allele of rs2248949, the C allele of rs1052677 in LXRB exhibited a significant protective effect. Moreover, all genotypes of LXRs were significantly associated with TB. Statistical significant differences in the allele (G allele of rs2248949), and genotypes (rs2279238 of LXRA, and rs17373080, rs1405655, rs1052677 of LXRB), however, were lost after Bonferroni correction of the p values (Pc).
To investigate a potential role of LXR SNPs in two subgroups of TB patients, we evaluated the SNP contribution on the clinical variables related to tuberculosis. No LXR allele associated to the investigated phenotypes in two patient subgroups was observed. In addition, the genotype at the LXRA SNP rs1449627 and LXRB SNPs s17373080, rs1405655, and rs1052677 was nominally associated to pulmonary TB, and the haplotype CATC in LXRB was significantly associated with an increased risk of PTB. After Bonferroni correction, CATC haplotype and LXRB SNPs rs17373080 were lost significant associated with disease risk.
The LXRB SNP rs17373080 in the promoter region was nominally associated with TB, particularly with PTB. A previously study revealed that the LXRB promoter SNP rs17373080, resides proximally to the transcription start site and could potentially contribute to the transcription potential of the LXRB gene promoter [18]. The strong LD over the genomic region containing LXRB may underlie the intronic SNP association between rs2248949 and rs1405655 with disease. Thus, further studies investigating the effect of various alleles at this position in relationship to the LXRB promoter activity and transcription factor binding are needed. The LXRA SNP rs2279238 is located at an exon splicing enhancer (ESE) where splicing factor SRp55 binds but the nucleotide variation is synonymous, resulting in a protein sequence identical to that of the wild type [19].
The LXRs, LXRA and LXRB, are ligand-activated transcription factors that belong to the nuclear hormone receptor superfamily, and have emerged as important regulators of lipid metabolism and inflammation. LXRA is highly expressed in liver, and at lower levels in the intestine, macrophages, adipose tissue, lung, adrenal glands and kidney, all of which play important roles in lipid homeostasis. In contract, LXRB is ubiquitously expressed [20]. Activation of the LXRs inhibited inflammatory gene expression of inflammatory cytokines, including TLR4, IL-1b, and TNF-a [21,22]. In recent years, there has been growing evidence that LXRA and LXRB exert key inhibitory roles in the control of inflammation [4]. Moreover, many studies have reported a crucial role for LXR in inhibiting inflammation in models of atherosclerosis [23], colorectal cancer [24], lung inflammation [25,26], autoimmune encephalomyelitis [27], spinal cord trauma [28], and sepsis [29]. Modulation of inflammatory signaling by LXRs substantiates its use as a therapeutic target for inflammatory diseases. With respect to pathological bone disease, activation of LXRs has been demonstrated to play a protective role in the prevention of rheumatoid arthritis [30], osteoarthritis [31,32], and postmenopausal osteoporosis [33]. In the past years, there have been several experiments focusing on the association between LXRA and/or LXRB gene polymorphisms with different diseases. Dahlman et al. reported that LXRA SNP rs2279238 CC carriers had lower body mass index and rs2279238 CT carriers were associated with obesity phenotypes [34,35]. Legry et al. documented that rs11039155 AA carriers of LXRA had higher HDL cholesterol and a 30% decrease in risk of having the metabolic syndrome in two cohorts of French subjects [36]. In contract, Rooki et al. have documented that two common SNPs in LXRA (rs11039155 and rs2279238) and in LXRB (rs17373080 and rs2695121) are not potential contributors to the risk of metabolic syndrome and related traits in an Iranian population [37].
In conclusion, our results suggested associations between common genetic variants of LXRs and the risks of pulmonary tuberculosis, confirming the recent study by Korf H et al. This investigation suggested that gene polymorphisms in LXRs are likely to affect the risk of developing tuberculosis, and demonstrated the potential of LXR modulators as therapeutic targets for tuberculosis. However, these findings must be interpreted in other large population samples and performing meta-analyses are necessary before a link between LXRs gene variability and tuberculosis can definitely be established. Moreover, understanding the molecular mechanisms behind the LXRs SNP-disorder association will require further experimental research.