SFRP1 variations influence susceptibility and immune response to Mycobacterium tuberculosis in a Chinese Han population

Infection, Genetics and Evolution 37 (2016) 259–265

Contents lists available at ScienceDirect

Infection, Genetics and Evolution journal homepage: www.elsevier.com/locate/meegid

Research paper

SFRP1 variations influence susceptibility and immune response to Mycobacterium tuberculosis in a Chinese Han population Zhenzhen Zhao a,1, Wu Peng a,1, Xuejiao Hu a, Jingya Zhang a, Mengqiao Shang a, Juan Zhou a, Yi Zhou a, Xingbo Song a, Xiaojun Lu a, Binwu Ying a,⁎,1, Xuerong Chen b,⁎ a b

Department of Laboratory Medicine, West China Hospital, Sichuan University, Chengdu 610041, PR China Division of Pulmonary Disease, Department of Respiratory Medicine, West China Hospital, Sichuan University, Chengdu 610041, PR China

a r t i c l e

i n f o

Article history: Received 4 September 2015 Received in revised form 15 November 2015 Accepted 27 November 2015 Available online 28 November 2015 Keywords: Tuberculosis Secreted frizzled-related protein 1 (SFRP1) Single nucleotide polymorphisms (SNPs) Inflammatory markers

a b s t r a c t Objectives: SFRP1 acts as a well-established inhibitory regulator of the Wnt signaling pathway, whose polymorphisms have been demonstrated to be associated with the risk of inflammation, infection as well as cancer. We verified the hypothesis that single nucleotide polymorphisms (SNPs) within SFRP1 gene are associated with susceptibility and clinical characteristics of tuberculosis disease in a Chinese Han population. Methods: Six candidate SNPs were genotyped using MassARRAY method in a case–control design (260 tuberculosis patients and 252 healthy controls). A comprehensive analysis of single locus including the genotypic, allelic frequencies and the genetic models, haplotypic construction as well as gene–gene interaction was conducted to investigate the relationships between SNPs and TB. Significant SNPs were further interrogated in relation to TB clinical features and host inflammatory status. Results: Genotype frequencies of rs4736958 and rs7832767 within SFRP1 gene were significantly different (p = 0.011, p = 0.008, respectively) between tuberculosis group and control group. Subjects carrying C allele for rs4736958 showed a decreased tuberculosis risk (OR = 0.66, 95% CI = 0.51–0.87, p = 0.003), whereas individuals carrying rs7832767 T allele had a significant increased risk in tuberculosis susceptibility (OR = 1.32, 95% CI = 1.01–1.74, p = 0.046). Genetic model analysis revealed that dominant, co-dominant and recessive models of rs4736958 were associated with decreased susceptibility to tuberculosis (p all b0.05), while the recessive and co-dominant models of rs7832767 were related to significantly increased risk for tuberculosis (p all b0.05). There was a reduced tuberculosis risk in association with the haplotype CC (representing rs3242 and rs4736958) of SFRP1 (OR = 0.73, 95% CI = 0.56–0.96, p = 0.026). Further stratification analysis indicated that TB patients with genotype CT for rs4736958 were associated with higher CRP concentrations, and heterozygous patients (CT genotype) of rs7832767 trended towards greater ESR levels. Conclusion: SNPs rs4736958 and rs7832767 of SFRP1 gene were significantly associated with tuberculosis susceptibility and might influence the expression levels of inflammatory markers of tuberculosis patients in a Chinese Han population. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Tuberculosis (TB) caused by infection of Mycobacterium tuberculosis (MTB), remains an enormous global public health threat nowadays (Lönnroth et al., 2010), with 9.0 million incident cases and 1.5 million TB-related deaths annually (WHO, 2014). The epidemic of TB has exacerbated due to co-infection with HIV as well as the emergency of multidrug-resistant TB. It is estimated that one-third of the world's ⁎ Corresponding authors. E-mail addresses: [email protected] (Z. Zhao), [email protected] (W. Peng), [email protected] (X. Hu), [email protected] (J. Zhang), [email protected] (M. Shang), [email protected] (J. Zhou), [email protected] (Y. Zhou), [email protected] (X. Song), [email protected] (X. Lu), [email protected] (B. Ying), [email protected] (X. Chen). 1 These authors contributed equally to this work.

http://dx.doi.org/10.1016/j.meegid.2015.11.031 1567-1348/© 2015 Elsevier B.V. All rights reserved.

population are infected with MTB, however, less than 10% will finally develop into clinically active TB disease (Lönnroth and Raviglione, 2008). The progression of TB disease is influenced not only by environment elements and the pathogenic features of MTB, but also by the host genetic factors (Torres-García et al., 2013). Among them, the role of host susceptibility to MTB infection has recently attracted increased attention of scholars. The Wnt signaling pathway is initially known as an evolutionarily highly conserved signal transduction system, which governs embryonic development as well as maintains homeostasis in the adult (Clevers and Nusse, 2012). Recent studies have implied that the Wnt signaling pathway is involved in the development and progression of TB (Schaale et al., 2011; Wu et al., 2014; Wu et al., 2015; Li et al., 2014; Neumann et al., 2010; Blumenthal et al., 2006). Blumenthal et al. (2006) found Wnt5a and its receptor Fzd5 to be induced in human macrophages in

260

Z. Zhao et al. / Infection, Genetics and Evolution 37 (2016) 259–265

response to MTB. It has been reported that the activated Wnt/β-catenin signaling could promote apoptosis that in part through the caspasedependent pathway as well as reduce necrosis partly by ROS-mediated PARP-1/AIF signaling pathway in BCG-infected RAW264.7 macrophage cells in Wu et al. (2014, 2015) research. The secreted frizzled-related protein (SFRP) family was first identified and well-characterized as antagonists of the Wnt/β-catenin signaling. SFRP1 protein contains a frizzledlike cysteine-rich domain (CRD) and a netrin-related motif (Bovolenta et al., 2008). Furthermore, both subunits are capable of binding to Wnt proteins, and subsequently block the pathway downstream cascades (Lin et al., 1997; Lopez-Rios et al., 2008) responsible for the antagonistic effects of Wnt signaling pathway. Abnormal regulation of SFRP1, especially the loss of expression caused by epigenetic silencing, is a well-known phenomenon for various carcinomas, including colon, breast and kidney cancers (Caldwell et al., 2004; Veeck et al., 2006; Dahl et al., 2007). Also, genetic variants in SFRP1 have been demonstrated to be associated with the risk of advanced hepatic inflammation and bladder cancer (Liu et al., 2013; Rogler et al., 2013). Barandon et al. (2011) recently demonstrated that SFRP1 might serve as a novel anti-inflammatory factor by switching the balance between pro-inflammatory and anti-inflammatory cytokines. There are very limited available data on the role of SFRP1 in the infectious diseases such as tuberculosis, whether and how the SFRP1 polymorphisms influence the development and progression of TB remain undetermined. Candidate genetic association study will be a remarkably helpful approach to elucidate the roles of SFRP1 in the host immune response to MTB. We have found that two single nucleotide polymorphisms (SNPs) in CTNNB1 and SFRP1 genes were correlated with the tuberculosis risk in the prior study (Hu et al., 2014), suggesting polymorphisms within SFRP1 may impact the susceptibility to TB. Till now, no more data are available enough to support the effect of SFRP1 on predisposition to TB. Considering the nominally significant but limited information about polymorphisms of SFRP1 in the former work, it is significantly important to explore more genetic determinants of SFRP1 gene and further functional relevance with TB disease. We therefore genotyped SNPs located in SFRP1 gene (rs3242, rs4736958, rs72643819, rs7832767, rs921142, and rs72643820) in a Chinese Han population to investigate their relations to TB risk, as well as clinical phenotype and inflammatory status of TB patients.

reference number is NO. 198 (2014), and signed informed consent forms were obtained from all the participants. 2.2. Target SNPs selection We retrieved dbSNP database (http://www.ncbi.nlm.nih.gov/ projects/SNP/), Hapmap database (http://hapmap.ncbi.nlm.nih.gov/) as well as 1000 Genomes database (http://www.1000genomes.org/) and combined the reported previously literature (Sims et al., 2008; Feng et al., 2011; Rogler et al., 2013), six candidate SNPs were identified for genotyping, mainly according to the principles as follows: (1) A reported minor allele frequency (MAF) of more than 0.10 in the Chinese Han Beijing population; (2) SNPs located in potentially functional regions mainly including exons, UTRs, band promoters (within 2 kb of the genes) were preferentially selected; (3) For tag SNPs, only those with correlation coefficient greater than 0.80 were selected; and (4) The amplification primers and single base extended primers should be successful for candidate SNPs. The selected target SNPs set listed in Supplementary Table 1. 2.3. Genotyping analysis Four milliliter (ml) EDTA-anticoagulated whole blood samples were collected from all participants. Genomic DNA was extracted using QIAamp® DNA blood mini kit (Qiagen, Germany) according to the manufacturer's protocol, and finally the genomic DNA diluted to 10 ng/ml for experiment. Target SNP genotyping was carried out using MassARRAY mass spectrometry (Sequenom, USA). Locus specific primers and probes were designed using MassARRAY Assay Design 3.0 software (Supplementary Table 1). Briefly, this assay comprised an initial locusspecific PCR reaction, followed by single base extension using massmodified dideoxy-nucleotide terminators of an oligonucleotide primer. The SNP allele with the distinct mass of the extended primer was distinguished by the application of MALDI-TOF mass spectrometry, as described in detail elsewhere (Gabriel et al., 2009). Positive and negative controls Table 1 Demographic and clinical characteristics of the study population. Characteristics

2. Materials and methods

Age (year) Gender (male/female)

2.1. Study population ALT (IU/L)

A total of 260 tuberculosis patients and 252 healthy controls were recruited in the present study. TB patients were consecutively recruited from the West China Hospital, Sichuan University during November 2012 to April 2014. All patients were newly diagnosed TB, having the typical signs and symptoms of TB and satisfying any of the inclusion criteria as follows: (1) smear positive for at least two separate clinical specimens and/or culture positive for MTB and/or examination positive for TB-DNA; (2) clinical and radiological findings consistent with TB disease; (3) pathological evidence of TB disease; and (4) tuberculosis interferon-gamma release assay (TB-IGRA) positive for at least two separate blood samples. These patients never received anti-tuberculosis treatments, or medication durations less than a week. Patients with HIV-seropositive results, other infectious diseases, cancers or immunosuppression therapy were excluded. The 252 healthy controls were randomly selected from the cohort of physical examination population, tuberculosis-free individuals on the basis of normal radiographic findings and physical examinations, no history of TB, and frequency matched to the case on gender and sex. All subjects were from the same district and were unrelated ethnic Han Chinese. This study was approved by the ethical committee of West China Hospital, Sichuan University, and the relevant Judgement's

AST (IU/L) Alb (g/L) Cys-C (mg/L) Urea (mmol/L) Crea (μmol/L) RBC (×1012/L) Hb (g/L) Hct (L/L) 9

PLT (×10 /L) WBC (×109/L) Neutrophil (×109/L) 9

Lymphocyte (×10 /L) Monocyte (109/L)

Group

Number

X ± SD or M(IQR)

Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case Control Case

258 250 151/107 153/98 257 211 257 211 257 211 257 244 257 250 257 250 258 249 258 249 258 249 258 249 258 249 258 249 258 249 257

41.00 ± 19.25 44.22 ± 11.92 – – 36.79 ± 10.51 28.84 ± 9.61 37.18 ± 11.32 26.68 ± 10.07 36.23 ± 6.37 46.56 ± 2.63 1.10 ± 0.41 0.85 ± 0.13 4.73 ± 1.56 5.57 ± 1.86 62.80(51.00–75.00) 75.90(64.67–87.62) 4.90 ± 0.80 4.82 ± 0.47 120.00(101.00–134.00) 148.00(135.00–159.50) 0.36 ± 0.08 0.45 ± 0.07 236.00(164.75–337.00) 164.00(135.00–194.50) 7.39 ± 3.46 6.20 ± 1.34 5.61 ± 1.80 3.70 ± 1.01 1.88 ± 0.48 2.00 ± 0.52 0.53 ± 0.16

p 0.107 0.576 0.244 0.015 b0.001 0.106 0.011 0.920 0.879 b0.001 b0.001 b0.001 0.067 b0.001 0.833 b0.001

Z. Zhao et al. / Infection, Genetics and Evolution 37 (2016) 259–265

261

Table 2 Association between genetic polymorphisms and TB in SFRP1. SNP

Case na(%)

Control na(%)

OR (95% CI)

P

rs3242

T C

28(5.5) 482(94.5)

27(5.4) 469(94.6)

1.01(0.59–1.74)

0.974

rs4736958

C T

139(27.4) 369(72.6)

179(36.2) 315(63.8)

0.66(0.51–0.88)

0.003

rs72643819

T G

213(41.8) 297(58.2)

206(41.2) 294(58.8)

1.02(0.80–1.32)

0.855

rs7832767

T C

160(31.1) 354(68.9)

128(25.5) 374(74.5)

1.32(1.01–1.74)

0.046

a

TT TC CC CC CT TT TT TG GG TT TC CC

Case na(%)

Control na(%)

P

5(2.0) 18(7.0) 232(91.0) 19(7.5) 101(39.8) 134(52.7) 44(17.3) 125(49.0) 86(33.7) 35(13.6) 90(35.0) 132(51.4)

2(0.8) 23(9.3) 223(89.9) 32(13.0) 115(46.6) 100(40.4) 36(14.4) 134(53.6) 80(32.0) 14(5.6) 100(39.9) 137(54.5)

0.372

0.011

0.527

0.008

Some subjects did not successfully genotype.

(without DNA) were included on each reaction, and five percent of all samples were randomly repeatedly tested. Experimenter had no idea about the case or control status of all specimens. Detected loci with a call rate b90% and an equivocal demarcation of genotyping clustering in all subjects were excluded from subsequent analysis. 2.4. Statistical analysis All genotype frequencies were analyzed for Hardy–Weinberg equilibrium (HWE) in the control group using the goodness-of-fit χ2 test. HWE was calculated to ensure that all of the data were within population equilibrium. The control population must satisfy the genetic balance (p N 0.05), whereas due to genetic drift induced by diseases, the case population may be accepted genetic imbalance (p b 0.05) (Salanti et al., 2005). The normal distributed data was presented as mean ± standard deviation (SD), the continuous non-normal distributed variables were recorded as median (interquartile range, IQR), and the categorical variables were summarized as number and percentage. Demographic and clinical features between TB patients and healthy controls were analyzed using SPSS version 16.0 software. The differences of the distributions of the allele and genotype frequencies of SFRP1 SNPs between groups were evaluated using χ2 test, and odds ratio (OR) and 95% confidence intervals (95% CI) were further calculated. The associations of inheritance models (including co-dominant model, dominant model, and recessive model) with TB risk were

analyzed through χ2 test. The linkage disequilibrium (LD) measurement, haplotype analysis were conducted using Haplotype version 4.2 software. Gene–gene interaction analysis was carried out by the logistic regression analysis and general multiple dimensionality reduction (GMDR) software. The statistically significant result was estimated at p-value b 0.05, and all statistical analyses were two-tailed. To avoid missing any possible significant associations in this exploratory study, we did not consider the adjustment for multiple comparisons.

3. Results 3.1. Demographic data and clinical characteristics of study subjects The demographic and clinical features of TB patients and healthy controls were summarized in Table 1. In regards to age and gender, no statistical differences were observed (p = 0.107, p = 0.576, respectively), and most subjects were middle-aged with mean age of 41.00 years for cases and 44.22 years for controls. There were a few significant differences in the routine laboratory results between populations from patient and control groups. The statistical significances were found in AST, Alb, Urea, Hb, Hct, PLT, neutrophil number and monocyte number (p all b0.05), while ALT, Cys-C, Crea, RBC, WBC, and lymphocyte number were not statistically significant (p all N0.05).

Table 3 Association between genetic models and TB risk in SFRP1. SNPs

Model Cod

rs3242

Dom Rec Cod

rs4736958

Dom Rec Cod

rs72643819

Dom Rec Cod

rs7832767

Dom Rec

Genotype

Case n(%)

Control n(%)

OR (95% CI)

CC TC TT TC + TT CC + TC TT TT TC CC TC + CC TT + TC CC GG TG TT TG + TT GG + TG TT CC TC TT TC + TT CC + TC TT

232(90.98) 18(7.06) 5(1.96) 23(9.02) 250(98.04) 5(1.96) 134(52.76) 101(39.76) 19(7.48) 120(47.24) 235(92.52) 19(7.48) 86(33.73) 125(49.02) 44(17.25) 169(66.27) 211(82.75) 44(17.25) 132(51.36) 90(35.02) 35(13.62) 125(48.64) 222(86.38) 35(13.62)

223(89.92) 23(9.27) 2(0.81) 25(10.08) 246(99.19) 2(0.81) 100(40.48) 115(46.56) 32(12.96) 147(59.52) 215(87.04) 32(12.96) 80(32.00) 134(53.60) 36(14.40) 170(68.00) 214(85.60) 36(14.40) 137(54.58) 100(39.84) 14(5.58) 114(45.42) 237(94.42) 14(5.58)

1.00 0.75(0.40–1.43) 2.40(0.46–12.51) 0.88(0.49–1.60) 1.00 2.46(0.47–12.80) 1.00 0.65(0.45–0.95) 0.44(0.23–0.82)) 0.60(0.42–0.86) 1.00 0.54(0.29–0.98) 1.00 0.87(0.59–1.28) 1.14(0.67–1.94) 0.93(0.64–1.34) 1.00 1.24(0.77–2.00) 1.00 0.93(0.64–1.35) 2.59(1.33–5.04) 1.13(0.80–1.61) 1.00 2.66(1.39–5.09)

p 0.385 0.451 0.686 0.285 0.026 0.009 0.006 0.043 0.476 0.638 0.680 0.380 0.719 0.004 0.467 0.002

262

Z. Zhao et al. / Infection, Genetics and Evolution 37 (2016) 259–265

3.2. Genotyping results The genotype agreement rate between duplicate samples was 100%, and the call rate for all SNPs N 90%. All participants were CT genotype for rs72643820 of SFRP1 gene, indicating that no polymorphism existed in this locus within our cohort. Genotype distributions of SNP rs921142 in both cases and controls deviated from Hardy–Weinberg equilibrium (p all b0.05). Therefore, both rs72643820 and rs921142 were precluded from further analysis. Genotype distributions of SNP rs3242 were in accordance with Hardy–Weinberg equilibrium (p N 0.05) in the control group, which indicated a p-value of less than 0.05 for the case group might attribute to the TB infection induced gene drift. We included SNP rs3242 in the following analysis as previously described (Salanti et al., 2005). Genotype frequencies of the remaining three SNPs in both groups were in agreement with the Hardy–Weinberg equilibrium (p all N 0.05), suggesting the genetic equilibrium of population enrolled and representative of samples. The genotype plots of target SNPs were shown in Supplementary Figure 1, which were implemented by MassARRAY mass spectrometry. 3.3. Association of SNPs with TB in SFRP1 gene The distributions of genotype and allele frequencies of 4 SNPs were depicted in Table 2. The frequency distributions of allele and genotype of rs4736958 showed significantly different between cases and controls (p = 0.003, p = 0.011, respectively), and the occurrence of C allele was more common in the control group (36.2%) compared with that in cases (27.4%), suggesting C allele of rs4736958 might be associated with significantly decreased susceptibility to TB (OR = 0.66, 95% CI = 0.51– 0.87, p = 0.003). There were significant distinctions in both allele and genotype frequencies of rs7832767 between two groups (p = 0.046, p = 0.008, respectively). Subjects carrying T allele for rs7832767 were correlated with a significant increase in TB predisposition in comparison with C allele carriers (OR = 1.32, 95% CI = 1.01–1.74, p = 0.046). Three genetic models were compared to determine the significance of each SNP: codominant pattern, dominant pattern, and recessive pattern. For example, if A was the ancestral allele and G was the mutant allele in a SNP locus, the codominant mode referred to GG/AG versus AA, the dominant model was identified as AG + GG versus AA, as well as the recessive model was defined as GG versus AG + AA. In rs4736958 locus, under the codominant model, the TC and CC genotypes were associated with the decreased risk for TB, estimating ORs of 0.65 and 0.44 (p = 0.026 and 0.009, separately), when we analyzed the dominant model (TC + CC versus TT), the association was at p-value of 0.006, with OR = 0.60 (95% CI = 0.42–0.82), as well as the association of the recessive model (CC versus TT + TC) was at p-value of 0.043, with OR = 0.54 (95% CI = 0.29–0.98). Rs7832767 showed an association with TB under a recessive model (p = 0.002), with an OR of developing tuberculosis for TT subjects versus those with CC/TC genotype estimated at 2.66. For the rs7832767 locus, given the codominant model, TT genotype carriers exhibited a 2.59-fold higher susceptibility to TB with p-value of 0.004 as compared to CC genotype (Table 3). The other two SNPs (rs3242 and rs72643819) showed no significant differences between patients and controls in either genotype (p = 0.337 and 0.527, respectively) or allele (p = 0.974 and 0.855, severally) frequencies comparisons. Additionally, the same statistical significances can be seen in their co-dominant, dominant and recessive models.

Fig. 1. Linkage disequilibrium (LD) structures of the SNPs in SFRP1 gene. LD between all pairs of SNPs was evaluated by D′-statistics. The D′-values (%) were presented in the squares. Pairwise D′-values were color coded: high D′-values were dark, low D′-values were light.

In line with individual SNP analysis, the haplotype CC was associated with reduced tuberculosis susceptibility at p-value of 0.026 (OR = 0.73, 95% CI = 0.56–0.96). 3.5. Gene–gene interactions analysis In order to evaluate whether there were gene–gene interactions between TB-associated genetic variations (rs4736958 and rs7832767), gene–gene interaction analysis was carried out by using the logistics regression analysis and GMDR software. Unfortunately, no potential gene–gene interactions between both SNPs were detected in the current study (Supplementary Table 2). 3.6. The relationships between tuberculosis clinical features and significant SNPs To further assess whether the genetic variances affect the tuberculosis clinical manifestation, we subsequently analyzed the demographic and clinical data among tuberculosis patients stratified according to three different genotypes (CC, CT and TT genotypes for both rs4736958 and rs7832767). Overall, the demographic and clinical characteristics of TB patients with different genotypes for both two discrepant SNPs loci were similar. Table 5 showed rs4736958 was significantly associated with CRP concentrations (p = 0.013), and patients with genotype CT presented higher levels of CRP as compared to the other two genotypes carriers. With respect to rs7832767, patients having CT genotype were related to greater levels of ESR (p = 0.043), while the association neared significance (p = 0.053) as regarding CRP level (Table 6). 4. Discussion Accumulating studies have indicated that host genetic factors contribute to the development, progression and outcome of TB, with an estimated heritability ranging from 36 to 80% (Möller et al., 2010), primarily through regulating the host immunity and inflammation in response to MTB infection. Several lines of evidence have shown roles for components related to Wnt signaling pathway in the pathogenesis of TB. As well-understood antagonist for canonical and non-canonical Wnt signaling pathway, SFRP1 may play a critical role in modulating

3.4. Haplotype analysis for SFRP1 We next constructed the haplotypes to analyze whether there were additive associations among SNPs tested. We performed the haplotype analyses for SFRP1 gene, as shown in Fig. 1, and found rs3242 and rs4736958 were in strong linkage disequilibrium (LD) with each other (D′ N 0.99). Table 4 summarized the haplotype frequencies of rs3242 and rs4736958 as well as their associations with tuberculosis disease.

Table 4 Association between haplotypes of rs3242 and rs4736958 of SFRP1 gene and TB. Haplotype

Control (freq)

Case (freq)

T-C C-C T-T

0.607 0.338 0.055

0.671 0.274 0.055

p-Value

OR(95% CI)

0.026 0.747

1.00 0.73(0.56–0.96) 0.91(0.53–1.58)

Z. Zhao et al. / Infection, Genetics and Evolution 37 (2016) 259–265

263

Table 5 Association of rs4736958 with clinical features of TB patients. Characteristics

Age (year) Gender (male/female) Location (PTB/EPTB/P&EP)a TB-DNA (negative/positive) Smear (negative/positive) MTB culture (negative/positive) Smoke (no/yes) Drink (no/occasional/often)b ALT (IU/L) AST (IU/L) Alb (g/L) Cys-C (mg/L) BUN (mmol/L) RBC (×1012/L) WBC(×109/L) Neutrophil (×109/L) Lymphocyte (×109/L) Monocyte (×109/L) Crea (μmol/L) Hb (g/L) PLT (×109/L) CRP (mg/L) Hct (L/L) ESR (mm/h) a b

rs4736958

p

CC

CT

TT

43.95 ± 18.06 14/5 8/1/10 4/10 13/5 2/3 15/2 11/2/4 22.89 ± 7.90 36.22 ± 11.24 34.90 ± 9.36 1.12 ± 0.42 4.78 ± 1.15 4.10 ± 0.66 6.90 ± 2.32 5.16 ± 2.14 1.09 ± 0.46 0.44 ± 0.17 64.75(53.00–73.77) 116.00(102.00–132.00) 239.00(192.00–347.00) 40.52 ± 13.57 0.35 ± 0.06 56.00(21.00–89.00)

41.41 ± 20.22 54/45 46/13/40 29/29 70/25 22/6 82/10 73/8/11 32.02 ± 10.56 35.14 ± 12.29 36.26 ± 6.68 1.17 ± 0.61 5.67 ± 1.63 4.14 ± 0.85 7.40 ± 2.09 5.42 ± 2.98 1.35 ± 0.38 0.53 ± 0.13 65.00(56.55–79.87) 117.50(98.50–132.00) 234.00(164.25–307.00) 60.01 ± 18.32 0.36 ± 0.07 52.50(28.00–82.00)

38.55 ± 18.83 82/52 68/17/49 32/55 82/37 28/13 107/21 24/5/7 42.91 ± 13.73 39.37 ± 11.71 36.42 ± 8.36 1.04 ± 0.26 4.21 ± 1.25 4.28 ± 0.77 7.30 ± 1.77 5.30 ± 2.86 1.26 ± 0.67 0.53 ± 0.15 60.00(48.85–73.35) 121.00(101.00–136.00) 233.00(159.50–353.50) 41.28 ± 12.41 0.37 ± 0.06 42.50(23.00–78.00)

0.357 0.252 0.725 0.176 0.741 0.184 0.554 0.408 0.630 0.894 0.638 0.264 0.060 0.322 0.820 0.912 0.743 0.649 0.054 0.503 0.792 0.013 0.526 0.421

PTB—pulmonary tuberculosis; EPTB—extra pulmonary tuberculosis; P&EP—pulmonary combined extra-pulmonary tuberculosis. No—never drink; occasional—daily drinking about 1.3–20 g; often—daily drinking more than 20 g.

the host immune responses to MTB. Recently, Lee et al. (Lee et al., 2012) revealed a novel function for SFRP1 acting as a potent inducer of human Th17 cells differentiation, whose product interleukin-17 (IL-17) appears to be an activator of macrophages and its variances are associated with tuberculosis susceptibility in the Chinese Han population (Peng et al., 2013). In the present survey, we first systematically investigated the relationships between SNPs within SFRP1 gene and susceptibility and clinical characteristics of tuberculosis disease in Han population. T allele and recessive inheritance model of rs7832767 significantly increased risk for tuberculosis in the current results, which consist with the previous findings (Hu et al., 2014). In addition, our results extended former findings that TT genotype of the co-dominant model of

rs7832767 was also at significantly elevated susceptibility to TB. Despite different genotyping methods, we observed the similar statistical significances in rs7832767 locus, suggesting that SNP rs7832767 may be a hopeful susceptible locus for TB infection in Chinese Han ethnicity. Based on our observations, SNP rs4736958 C allele might be a protective marker against MTB infection. The dominant, recessive, and codominant models of rs4736958 exhibited decreased TB susceptibility, which seemed feasible that the occurrence of at least one C allele vice versa the absence of one T allele may decrease the risk for TB. Considerably less is known about the associations of rs4736958 with infectious diseases yet. To our knowledge, this is the first report of its association with susceptibility to tuberculosis infection in Chinese Han population.

Table 6 Association of rs7832767 with clinical features of TB patients. Characteristics

Age (year) Gender (male/female) Location (PTB/EPTB/P&EP)a TB-DNA (negative/positive) Smear (negative/positive) MTB culture (negative/positive) Smoke (no/yes) Drink (no/occasional/often)b ALT (IU/L) AST (IU/L) Alb (g/L) Cys-C (mg/L) BUN (mmol/L) RBC (×1012/L) WBC (×109/L) Neutrophil (×109/L) Lymphocyte (×109/L) Monocyte (×109/L) Crea (μmol/L) Hb (g/L) PLT (×109/L) CRP (mg/L) Hct(L/L) ESR (mm/h)

rs7832767

p

CC

CT

TT

39.30 ± 10.29 70/62 71/11/50 34/52 88/34 32/15 110/19 104/10/15 36.22 ± 11.72 39.38 ± 11.62 36.59 ± 6.37 1.00 ± 0.26 4.57 ± 1.59 4.29 ± 0.81 7.13 ± 2.59 5.10 ± 1.21 1.37 ± 0.68 0.54 ± 0.15 60.20(47.77–73.75) 121.00(102.25–134.75) 234.50(165.25–336.00) 44.02 ± 12.66 0.36 ± 0.06 41.00(21.50–74.50)

41.53 ± 17.03 58/31 37/15/37 25/31 59/24 16/6 70/12 60/8/14 34.61 ± 11.94 33.70 ± 11.84 35.43 ± 6.55 1.17 ± 0.45 5.00 ± 1.51 4.13 ± 0.76 7.64 ± 2.21 5.61 ± 1.66 1.26 ± 0.33 0.52 ± 0.13 64.00(55.00–80.10) 118.00(97.00–135.75) 258.50(171.50–316.50) 62.17 ± 17.83 0.35 ± 0.07 57.00(36.00–93.50)

39.47 ± 13.46 23/11 14/7/13 9/11 21/9 5/1 27/2 24/2/2 45.40 ± 14.05 37.94 ± 10.77 37.04 ± 5.85 1.20 ± 0.35 4.77 ± 1.14 4.19 ± 0.83 7.00 ± 1.84 5.41 ± 1.34 1.00 ± 0.31 0.47 ± 0.12 63.00(52.50–76.00) 122.00(102.00–132.00) 225.00(131.00–296.00) 36.75 ± 13.65 0.36 ± 0.06 39.00(19.00–84.00)

0.690 0.111 0.124 0.799 0.969 0.789 0.618 0.627 0.876 0.832 0.310 0.056 0.732 0.374 0.423 0.416 0.412 0.644 0.065 0.715 0.361 0.053 0.336 0.043

264

Z. Zhao et al. / Infection, Genetics and Evolution 37 (2016) 259–265

Rs4736958 is located in the 3′UTR of SFRP1 gene, which is an area for binding sites of microRNAs that affect stability and transport of the target mRNAs. MicroRNAs (miRNAs) are small, highly-conserved, noncoding RNA molecules that are thought to bear important roles in the pathogenesis of tuberculosis (Li et al., 2011; Iwai et al., 2015). According to bioinformatics software SNPinfo software (http://snpinfo.niehs.nih. gov/), rs4736958 is predicted to lie within the seed binding site for miR-9 modulating inflammatory responses of monocytes against invasion pathogens (Thulin et al., 2013). C-to-T mutation of rs4736958 may diminish the affinity of miR-9 with SFRP1 mRNA, as shown in Supplementary Figure 2, altering the regulatory ability of miR-9 and ultimately affecting host immune response and the development of inflammatory-related diseases such as TB. Further functional studies are urgently needed to validate this hypothesis and account for the biological mechanisms involving rs4736958. SNP rs3242 that has been extensively studied is also situated in 3′ UTR of SFRP1 gene and its TT genotype was found to be associated with the higher bone mineral density (BMD) values for femoral neck and total hip in Japanese women (Ohnaka et al., 2009). Rogler et al. demonstrated that patients harboring one T allele of rs3242 might have a higher bladder cancer risk in Caucasian population (Rogler et al., 2013). However, we found no significant association of rs3242 with the risk of tuberculosis infection in our study, indicating that this SNP locus might not contribute to the tuberculosis susceptibility among Chinese. Together, this phenomenon may in part be attributable to the heterogeneity in the pattern of genetic associations with diseases among different populations. For the first time, we investigated the relationships between rs72643819 and rs72643820 which both located in SFRP1 gene promoter region and TB susceptibility. Unexpectedly, the rs72643820 had no polymorphism among the study population. The unrelated results observed in the rs72643819 locus remain to be further confirmed in larger populations. For complex genetic diseases like TB for instance, linkage disequilibrium measure, haplotype analysis as well as gene–gene interaction analysis can be more advantageous over individual SNP analysis in determining disease-related genetic polymorphisms (Morris and Kaplan, 2002; Cordell, 2009). Haplotype CC of SFRP1 (representing rs3242 and rs4736958) displayed a decreased tuberculosis risk, which was in accordance with the individual SNP association analysis, providing evidence of the synthetic protective effect of these two SNPs against TB infection. Although gene–gene interactions between these two significant SNPs (rs7832767 and rs4736958) failed to yield additional positive information about their synergistic or antagonistic effects on TB risk, the gene– environment interaction analysis and high-order interaction study need to be performed in the future larger study. Studies have already demonstrated that the expression levels of CRP and ESR are in great values in assessing the activity and severity of TB to some extent (Furuhashi et al., 2012). Our former work has shown a phenomenon that rs7832767 T allele carriers are accompanied with elevated expression levels of CRP and ESR. We then identified TB patients with heterozygous genotypes of rs4736958 and rs7832767 who were more likely to have higher expression levels of CRP and ESR in this study. Albeit further functional experiments are required to support and explain these findings, these findings provide important evidences that SFRP1 genetic variances may affect the expression of inflammatory markers of TB patients as well as the host immune responses to MTB. One thing to note here is that the p-value of CRP concentrations in relation to rs7832767 was at borderline significant level (0.053), and a larger study will be more convincing to the underlying relationship. Our research still suffers from the limitations of the sample size and the SNP loci, and further prospective studies in larger populations with diverse ethnicities are required to verify these findings. More than that, our study is only an epidemiological genetic association survey, since functional exploration of significant polymorphisms is urgently warranted. In conclusion, rs4736958 and rs7832767 within SFRP1 gene were associated with susceptibility to tuberculosis in Chinese Han population,

and both SNPs were correlated with the expression levels of CRP and ESR, suggesting that they serve as novel predictive markers for the risk of developing TB in the Chinese Han population. Further epidemiological and functional studies in larger population are required to validate these results. Disclosure statement All authors have declared that no competing interests exist. Acknowledgments This work was supported by Grant #81472026 from the National Natural Science Foundation of China and by Grant #2014SZ0208 from the Projects in the Science and Technology Department of Sichuan Province Pillar Program of China. Appendix A. Supplementary data Supplementary materials are available in Infection, Genetics and Evolution (http://www.journals.elsevier.com/infection-genetics-andevolution/). Supplementary materials consist of data provided by the authors are published to help reader. Supplementary data associated with this article can be found in the online version, at doi:http://dx. doi.org/10.1016/j.meegid.2015.11.031. References Barandon, L., Casassus, F., Leroux, L., Moreau, C., Allières, C., Lamazière, J.M., Dufourcq, P., Couffinhal, T., Duplàa, C., 2011. Secreted frizzled-related protein-1 improves postinfarction scar formation through a modulation of inflammatory response. Arterioscler. Thromb. Vasc. Biol. 31, e80–e87. Blumenthal, A., Ehlers, S., Lauber, J., Buer, J., Lange, C., Goldmann, T., Heine, H., Brandt, E., Reiling, N., 2006. The Wingless homolog WNT5A and its receptor Frizzled-5 regulate inflammatory responses of human mononuclear cells induced by microbial stimulation. Blood 108, 965–973. Bovolenta, P., Esteve, P., Ruiz, J.M., Cisneros, E., Lopez-Rios, J., 2008. Beyond Wnt inhibition: new functions of secreted Frizzled-related proteins in development and disease. J. Cell Sci. 121, 737–746. Caldwell, G.M., Jones, C., Gensberg, K., Jan, S., Hardy, R.G., Byrd, P., Chughtai, S., Wallis, Y., Matthews, G.M., Morton, D.G., 2004. The Wnt antagonist sFRP1 in colorectal tumorigenesis. Cancer Res. 64, 883–888. Clevers, H., Nusse, R., 2012. Wnt/β-catenin signaling and disease. Cell 149, 1192–1205. Cordell, H.J., 2009. Detecting gene-gene interactions that underlie human diseases. Nat. Rev. Genet. 10, 392–404. Dahl, E., Wiesmann, F., Woenckhaus, M., Stoehr, R., Wild, P.J., Veeck, J., Knüchel, R., Klopocki, E., Sauter, G., Simon, R., Wieland, W.F., Walter, B., Denzinger, S., Hartmann, A., Hammerschmied, C.G., 2007. Frequent loss of SFRP1 expression in multiple human solid tumours: association with aberrant promoter methylation in renal cell carcinoma. Oncogene 26, 5680–5691. Feng, W.X., Mokrousov, I., Wang, B.B., Nelson, H., Jiao, W.W., Wang, J., Sun, L., Zhou, S.R., Xiao, J., Gu, Y., Wu, X.R., Ma, X., Shen, A., 2011. Tag SNP polymorphism of CCL2 and its role in clinical tuberculosis in Han Chinese pediatric population. PLoS ONE 6, e14652. Furuhashi, K., Shirai, T., Suda, T., Chida, K., 2012. Inflammatory markers in active pulmonary tuberculosis: association with Th1/Th2 and Tc1/Tc2 balance. Kekkaku 87, 1–7. Gabriel, S., Ziaugra, L., Tabbaa, D., 2009. SNP genotyping using the Sequenom MassARRAY iPLEX platform. Curr. Protoc. Hum. Genet. 2, 12. Hu, X., Shang, M., Zhou, J., Ye, Y., Lu, X., Tao, C., Ying, B., Wang, L., 2014. Association of genetic variants in Wnt signaling pathway with tuberculosis in Chinese Han population. PLoS ONE 9, e93841. Iwai, H., Funatogawa, K., Matsumura, K., Kato-Miyazawa, M., Kirikae, F., Kiga, K., Sasakawa, C., Miyoshi-Akiyama, T., Kirikae, T., 2015. MicroRNA-155 knockout mice are susceptible to Mycobacterium tuberculosis infection. Tuberculosis (Edinb.) 95, 246–250. Lee, Y.S., Lee, K.A., Yoon, H.B., Yoo, S.A., Park, Y.W., Chung, Y., Kim, W.U., Kang, C.Y., 2012. The Wnt inhibitor secreted Frizzled-Related Protein 1 (sFRP1) promotes human Th17 differentiation. Eur. J. Immunol. 42, 2564–2573. Li, D., Wang, T., Song, X., Qucuo, M., Yang, B., Zhang, J., Wang, J., Ying, B., Tao, C., Wang, L., 2011. Genetic study of two single nucleotide polymorphisms within corresponding microRNAs and susceptibility to tuberculosis in a Chinese Tibetan and Han population. Hum. Immunol. 72, 598–602. Li, Y., Shi, J., Yang, J., Ma, Y., Cheng, L., Zeng, J., Hao, X., Ma, C., Wang, Y., Liu, X., 2014. A Wnt/β-catenin negative feedback loop represses TLR-triggered inflammatory responses in alveolar epithelial cells. Mol. Immunol. 59, 128–135. Lin, K., Wang, S., Julius, M.A., Kitajewski, J., Moos Jr., M., Luyten, F.P., 1997. The cysteine-rich frizzled domain of Frzb-1 is required and sufficient for modulation of Wnt signaling. Proc. Natl. Acad. Sci. U. S. A. 94, 11196–11200.

Z. Zhao et al. / Infection, Genetics and Evolution 37 (2016) 259–265 Liu, Y., El-Serag, H.B., Jiao, L., Lee, J., Moore, D., Franco, L.M., Tavakoli-Tabasi, S., Tsavachidis, S., Kuzniarek, J., Ramsey, D.J., White, D.L., 2013. WNT signaling pathway gene polymorphisms and risk of hepatic fibrosis and inflammation in HCV-infected patients. PLoS ONE 8, e84407. Lönnroth, K., Castro, K.G., Chakaya, J.M., Chauhan, L.S., Floyd, K., Glaziou, P., Raviglione, M.C., 2010. Tuberculosis control and elimination 2010-50: cure, care, and social development. Lancet 375, 1814–1829. Lönnroth, K., Raviglione, M., 2008. Global epidemiology of tuberculosis: prospects for control. Semin. Respir. Crit. Care Med. 29, 481–491. Lopez-Rios, J., Esteve, P., Ruiz, J.M., Bovolenta, P., 2008. The Netrin-related domain of Sfrp1 interacts with Wnt ligands and antagonizes their activity in the anterior neural plate. Neural Dev. 3, 19. Möller, M., de, Wit, E., Hoal, E.G., 2010. Past, present and future directions in human genetic susceptibility to tuberculosis. FEMS Immunol. Med. Microbiol. 58, 3–26. Morris, R.W., Kaplan, N.L., 2002. On the advantage of haplotype analysis in the presence of multiple disease susceptibility alleles. Genet. Epidemiol. 23, 221–233. Neumann, J., Schaale, K., Farhat, K., Endermann, T., Ulmer, A.J., Ehlers, S., Reiling, N., 2010. Frizzled1 is a marker of inflammatory macrophages, and its ligand Wnt3a is involved in reprogramming Mycobacterium tuberculosis-infected macrophages. FASEB J. 24, 4599–46612. Ohnaka, K., Yamamoto, K., Nakamura, K., Adachi, M., Kawate, H., Kono, S., Takayanagi, R., 2009. Association of single nucleotide polymorphisms in secreted frizzled-related protein 1 gene with bone mineral density in Japanese women. Geriatr. Gerontol. Int. 9, 304–309. Peng, R., Yue, J., Han, M., Zhao, Y., Liu, L., Liang, L., 2013. The IL-17F sequence variant is associated with susceptibility to tuberculosis. Gene 515, 229–232. Rogler, A., Hoja, S., Socher, E., Nolte, E., Wach, S., Wieland, W., Hofstädter, F., Goebell, P.J., Wullich, B., Hartmann, A., Stoehr, R., 2013. Role of two single nucleotide polymorphisms in secreted frizzled related protein 1 and bladder cancer risk. Int. J. Clin. Exp. Pathol. 6, 1984–1998. Salanti, G., Amountza, G., Ntzani, E.E., Ioannidis, J.P., 2005. Hardy–Weinberg equilibrium in genetic association studies: an empirical evaluation of reporting, deviations, and power. Eur. J. Hum. Genet. 13, 840–848.

265

Schaale, K., Neumann, J., Schneider, D., Ehlers, S., Reiling, N., 2011. Wnt signaling in macrophages: augmenting and inhibiting mycobacteria-induced inflammatory responses. Eur. J. Cell Biol. 90, 553–559. Sims, A.M., Shephard, N., Carter, K., Doan, T., Dowling, A., Duncan, E.L., Eisman, J., Jones, G., Nicholson, G., Prince, R., Seeman, E., Thomas, G., Wass, J.A., Brown, M.A., 2008. Genetic analyses in a sample of individuals with high or low BMD shows association with multiple Wnt pathway genes. J. Bone Miner. Res. 23, 499–506. Thulin, P., Wei, T., Werngren, O., Cheung, L., Fisher, R.M., Grandér, D., Corcoran, M., Ehrenborg, E., 2013. MicroRNA-9 regulates the expression of peroxisome proliferator-activated receptor δ in human monocytes during the inflammatory response. Int. J. Mol. Med. 31, 1003–1010. Torres-García, D., Cruz-Lagunas, A., García-Sancho Figueroa, M.C., Fernández-Plata, R., Baez-Saldaña, R., Mendoza-Milla, C., Barquera, R., Carrera-Eusebio, A., RamírezBravo, S., Campos, L., Angeles, J., Vargas-Alarcón, G., Granados, J., Gopal, R., Khader, S.A., Yunis, E.J., Zuñiga, J., 2013. Variants in toll-like receptor 9 gene influence susceptibility to tuberculosis in a Mexican population. J. Transl. Med. 11, 220. Veeck, J., Niederacher, D., An, H., Klopocki, E., Wiesmann, F., Betz, B., Galm, O., Camara, O., Dürst, M., Kristiansen, G., Huszka, C., Knüchel, R., Dahl, E., 2006. Aberrant methylation of the Wnt antagonist SFRP1 in breast cancer is associated with unfavourable prognosis. Oncogene 25, 3479–3488. WHO, 2014. Global Tuberculosis Report Available: http://www.who.int/tb/publications/ global-report/en/. Wu, X., Deng, G., Hao, X., Li, Y., Zeng, J., Ma, C., He, Y., Liu, X., Wang, Y., 2014. A caspasedependent pathway is involved in Wnt/β-catenin signaling promoted apoptosis in Bacillus Calmette–Guerin infected RAW264.7 macrophages. Int. J. Mol. Sci. 15, 5045–5062. Wu, X., Deng, G., Li, M., Li, Y., Ma, C., Wang, Y., Liu, X., 2015. Wnt/β-catenin signaling reduces Bacillus Calmette–Guerin-induced macrophage necrosis through a ROSmediated PARP/AIF-dependent pathway. BMC Immunol. 16, 16.