The role of WWOX polymorphisms on COPD susceptibility and pulmonary function traits in Chinese: a case-control study and family-based analysis

Single nucleotide polymorphisms (SNPs) in the WW domain containing oxidoreductase (WWOX) gene were recently identified to be quantitative trait loci for lung function and thus likely to be susceptible biomarkers for COPD. However, the associations between WWOX SNPs and COPD risk are still unclear. Here, by conducting a two-center case-control study including 1511 COPD cases and 1677 controls and a family-based analysis comprising 95 nuclear pedigrees, we tested the associations between five SNPs that are rs10220974C >T, rs3764340C >G, rs12918952G >A, rs383362G >T, rs12828G >A of WWOX and COPD risk as well as the hereditary inclination of these loci among COPD families. We found that the SNP rs383362G >T was significantly associated with an increased risk of COPD in a T allele-number dependent-manner (OR = 1.30, 95%CI = 1.11 - 1.52). The T allele was more prone to over transmit to sick children and sibs than the G allele (Z = 2.900, P = 0.004). Moreover, the forced expiratory volume in one second/forced vital capacity (FEV1/FVC), FEV1/predicted-FEV1 and annual FEV1 also significantly decreased in the rs383362T carriers compared to the rs383362GG carriers. For other SNPs, no significant association was observed for COPD and pulmonary function. Taken together, our data demonstrated that the SNP rs383362G >T of WWOX plays a role in COPD inheritance.

Recent results from the genome-wide association studies (GWASs) verified above opinion. Two GWASs conducted in European reported that single nucleotide polymorphisms (SNPs), which are located in the WW domain containing oxidoreductase (WWOX) gene, were significantly associated with pulmonary function traits including the forced expiratory volume in one second (FEV1), forced vital capacity (FVC) and FEV1/FVC 15,16 . WWOX is a potential antioxidant enzyme with respect to its oxidoreductase functional domain. WWOX plays important roles in the regulation of a wide variety of cellular functions such as protein degradation, transcription and RNA splicing. Stimulus such as tobacco smoking can directly trigger oxidative stress by virtue of decrease in WWOX expression 7 . WWOX also can interact with P53 that is an important regulator involving the pathogenesis of COPD 17 . Taken together, these evidences suggest the WWOX gene to be a susceptible gene of COPD.
However, association between WWOX SNPs and COPD risk are still unknown by now. Also, the GWASs reported SNPs are located in the introns of WWOX, which are lack of possibly biological function. The human WWOX gene (NM_016373) maps to the chromosome 16q23.3-24.1 region with a 1.13Mb DNA sequence. Through the fine mapping analysis, we had previously reported that the WWOX SNPs were associated with lung cancer risk in Chinese 18 . Overlap of susceptible loci between COPD and lung cancer had been widely observed 13,[18][19][20] . Therefore, we aimed to test the hypothesis that WWOX SNPs were associated with COPD risk in Chinese.
In the current study, we conducted a two-stage case-control study including 1511 COPD cases and 1677 controls to test the associations between five putatively functional tagSNPs of WWOX (i.e., rs10220974C > T, rs3764340C > G, rs12918952G > A, rs383362G > T and rs12828G > A) and COPD risk. We further performed a family-based analysis comprising 95 pedigrees to verify the associations and to assess the hereditary inclination of susceptible loci among COPD pedigrees. Table S1 summarizes frequency distribution of demographic characteristics, possible risk factors and the Global Initiative for Chronic Obstructive Lung Disease (GOLD) stages in COPD cases and controls of both populations. Table 1 presents genotypes of the five WWOX SNPs between cases and controls in the southern Chinese population as a discovery set. All genotype frequency distributions of these SNPs matched the Hardy-Weinberg equilibrium (HWE) in the controls (P > 0.05 for all). As shown, only a significant deviation was found in genotype frequency distribution of the SNP rs383362G > T between the COPD cases and controls (P = 0.025), but was not of others (P > 0.05 for all). Compared to the rs383362GG genotype, the rs383362TG genotype exerted a significantly increased COPD risk (OR = 1.25, 95%CI = 1.01-1.55), while the the rs383362TT genotype conferred a non-significant increased COPD risk (OR = 2.23, 95%CI = 0.84-5.93). According to the genetic model selection strategy 21 , this SNP best fitted the additive model that the rs383362G > T conferred a significantly increased risk to COPD in a T allele dose-dependent manner (OR = 1.28, 95%CI = 1.06-1.56). However, the Bonferroni correction test did not support the significance of the association (P = 0.125). In addition, because one previous study had reported the haplotype rs3764340G-rs383362T with significant association with lung cancer risk 18 , we tested the association between the haplotype and COPD risk. However, no significant association was observed (Data not shown).

Associations between the WWOX SNPs and COPD risk. Supplementary
Despite the Bonferroni correction showed no significant association between the SNP rs383362G > T and COPD risk, the above data still supported a promising association between them considering that the Bonferroni correction sometimes cause excessive correction. Thus, we still validated the SNP in the eastern Chinese population. The results were consistent. The SNP was associated with an increased COPD risk under the additive model (OR = 1.36, 95%CI = 1.04-1.77; Bonferroni correction: P = 0.130; Table 2). We then merged the two populations to increase the study power. Under the additive model, the risk for developing COPD increased by 1.30-times accompanying with the number of rs383362T allele (OR = 1.30, 95%CI = 1.11-1.52; Bonferroni correction: P = 0.005; Table 2).  Table 3 shows results from stratification analysis of the association between the SNP rs383362G > T and COPD risk, as well as the interaction between the SNP and variables on COPD risk. The rs383362G > T significantly interacted with biomass as fuels on increasing COPD risk (P = 0.021), because there was an intuitionistic  Table 3. Stratification analysis of the association between the SNP rs383362G >T and COPD risk in the merged population. a ORs were adjusted for age, sex, pack-years smoked, biomass as fuels and sample source region in a logistic regression model. b P value from the multiple interaction test.
higher OR that equaled to 2.32 (95%CI = 1. 28-4.22) in subjects who used biomass as fuels than the OR that equaled to 1.23 (95%CI = 1.04-1.45) in subjects who did not use biomass as fuels. Further three-way interaction analysis showed no significant interaction between biomass as fuels, gender and the SNP on COPD risk (P = 0.548). No other significant interaction was observed (P > 0.05 for all). We also conducted a multivariable logistic regression analysis including above covariates and interaction. The data showed a 1.31-fold risk of COPD in rs383362T carriers compared to rs383362GG carriers (OR = 1.31, 95%CI = 1.12-1.55). Table 4, the single-marker analysis indicated that the T risk allele of rs383362G > T was more prone to over transmitted to sick children and sibs under the additive genetic model (Z = 2.900, P = 0.004), so other variants were not (P > 0.05 for all). The haplotype analysis involving the two SNPs rs3764340C > G and rs383362G > T further demonstrated that compared to the other three haplotypes, the T-C haplotype exerted significantly excessive transmission from parents to affected offspring under the additive genetic model (Z = 2.550, P = 0.011), so other haplotypes did not (P > 0.05 for all). In addition, the logistic regression model showed that the rs383362T allele and T-C haplotype conferred significantly increased risk of COPD compared to the rs383362G allele (OR = 2.04, 95%CI = 1.   hsa-miR-134 and hsa-miR-758 were experimentally proved to be translational regulators of WWOX (http://www. genecards.org/cgi-bin/carddisp.pl?gene = WWOX).

Discussion
In the current study, based on a two-stage case-control and a family-based study, we found that the T genotypes of WWOX SNP rs383362G > T were significantly associated with an increased risk of COPD and decreased pulmonary function traits in Chinese. The T allele exerted significantly excessive transmission from parents to sick offspring. The genotypes also interacted with biomass as fuels on increasing COPD risk. Bioinformatics analysis further showed that the SNP has a possible regulation on WWOX expression. COPD is a well-established disease involving both genetic and environmental factors. Multiple association studies especially the GWASs, have reported an abundant of SNPs to be susceptible loci for COPD or pulmonary function. These SNPs majorly belong to genes that involve the major pathogenic mechanisms of COPD, such as inflammation 22,23 , oxidative stress 15,24 , DNA damage 25,26 and epithelial-mesenchymal transition (EMT) 26,27 . However, most of these SNPs are considered to be lack of function and non-causal variant with regard to their genomic location in gene introns or intergenic region. An effective strategy to discover causal variants was fine-mapping based on the results from GWASs. Fine mapping majorly analyzes those putatively function variants considering their location in cis-acting element of genes that harbor susceptible loci of disease. Based on this strategy, we found a 3′ -UTR SNP rs383362G > T of WWOX, a gene that has been reported to harbor susceptible loci of pulmonary function by GWASs 15,16 , was associated with COPD risk and pulmonary function traits. The rs383362T risk allele was more prone to over transmitted to sick children and sibs. This SNP got Bioinformatics data support to have potentiality on affecting WWOX expression. Taken together, all these evidences suggested the rs383362G > T polymorphism to be a susceptible loci for COPD in Chinese.
Low expression of WWOX has been reported to be an inducement of human diseases such as lung cancer [28][29][30] . The SNP rs383362G > T is located in the 3′ -UTR of WWOX, the transversion of G to T might influence four microRNAs' binding ability to WWOX mRNA as bioinformatics analysis shown. Of the four SNPs, the hsa-miR-134 is an identified microRNA to regulate WWOX expression 31 . Meaningfully, expression data in lymphoblastic cells of CHB from microarray platforms showed that the rs383362T variants have lower WWOX expression than the rs383362GG genotype. These indicated a conjectural mechanism underlying the effect of rs383362G > T to COPD susceptibility. The SNP also interacted with biomass as fuels. Biomass exposure is an important risk factor of COPD, from which smoke can produce environmental stimuluses and induce oxidative damage 32 . WWOX can be downregulated by such exposures 33 . Therefore, subjects with T genotypes are more vulnerable to suffer oxidative damage and thus predispose to develop COPD.
Associations between the WWOX SNPs and several human diseases have been reported including thyroid carcinoma 34 , prostate cancer 35,36 , breast cancer 37 , and lung cancer 18 . Now, we identified the WWOX SNP rs383362G > T to be associated with COPD risk. Meanwhile, the adverse effect driven by the rs383362T allele was increased by a slight in patients with GOLD stage III than those with GOLD stage II or I. The effect was not significant in The oneway ANOVA test was used to test the differences in pulmonary function between the rs383362G > T genotype carriers. As shown, the rs383362T variants significantly exerted lower FEV1 and FEV1/predicted-FEV1, and caused more annual average decline in FEV1 than the rs383362GG genotype.
stage IV, which may be due to the limited sample size. Furthermore, the SNP rs383362G > T was also correlated with FEV1/FVC and FEV1/predicted-FEV1, both of which are essential on COPD diagnosis. The rs383362T variants also caused more annual average decline in FEV1 than the rs383362GG genotype. Decline in FEV1 is a marker for airflow obstruction and serves as primary symptom of both developing COPD and exacerbation. These correlations suggested that the SNP has direct effect on pulmonary function, further supporting its association with COPD risk. Taken together, The SNPs might be not only a genetic biomarker for COPD onset but also a predictor for COPD severity.
Some limitations may affect the validity of current study. First, the portion of GOLD stage in our study was a little inconsistent with the parent population as suggested with respect to the high frequency of stage I COPD in the current study, suggesting a serious selection bias in the current study that influenced the external validity 38 . Second, some important confounders such as passive smoking were not concerned in this study, which might cause confounding bias and thus affect internal validity. Third, the relatively small sample size of the nuclear pedigrees in family based analysis might also cause error results. However, the family-based analysis effectively controlled some confounding biases. Because consistent associations were observed in both the case-control study and family based analysis, and the SNP got Bioinformatics data support to be functional, it is conceivable that our finding was not achieved by chance.
To conclude, our study identified a putatively functional SNP rs383362G > T of WWOX that confers an increased risk of COPD, the rs383362T risk allele was more prone to over transmitted from parents to sick offspring. This SNP might be a genetic biomarker to predict risk of COPD in Chinese.

Materials and Methods
Case-control study. A two-stage case-control study was conducted in a southern Chinese population and an eastern Chinese population, which have been described in previously published studies 13,14 . COPD patient was defined as subjects with FEV1/FVC < 70% after inhalation of 400μ g salbutamol and with some chronic airway symptoms including chronic cough, dyspnea, sputum production or wheezing. In brief, the southern Chinese population included 1025 COPD patients and 1061 age (± 5 years) and sex frequency-matched controls, of which 697 cases were recruited from three communities (Liwan, Xicun and Zhanqian communities) in Guangzhou city based on annual cross-sectional surveys of COPD, 328 cases were recruited from physical examination centers of three hospitals in Guangzhou city including the Guangzhou Chest Hospital, the third Affiliated Hospital of Guangzhou Medical University and the third Affiliated Hospital of Sun Yat-sen University. All controls were selected from individuals with normal lung function (i.e., FEV1/FVC > 70%) who participated in the annual cross-sectional surveys of COPD. The eastern Chinese population comprised 486 COPD patients and 616 normal controls, of which the cases were enrolled from the physical examination center of the Second Affiliate Hospital of Soochow University and the controls were selected from a database consisting of 3,500 individuals based on a physical examination in Suzhou city. All subjects are generally unrelated Han Chinese. Moreover, a long-term follow-up on lung function monitoring was performed in part of the southern Chinese population for years as previously reported 39 . Heretofore, 427 individuals were successfully followed up and had at least four years' data of lung function between 2002 and 2010 with annual spirometry test. Each subject was scheduled for an interview with a structured questionnaire to provide data on age, sex, smoking status, pack-years smoked, biomass using and to donate 5 mL peripheral blood after a written informed consent had obtained. The definitions of smoking status and biomass as fuels were described in our previous publications 13 . Briefly, those participants who had smoked < 100 cigarettes in their lifetime were defined as never smokers. Otherwise, they were classified as ever smokers. Biomass as fuels means that people use bio-crop stalks and wood as fuels. The study was approved by the institutional review boards of Guangzhou Medical University and Soochow University. The methods were carried out in accordance with the approved guidelines.
Family-based analysis. A family-based analysis was conducted in a southern Chinese population between September 2010 and September 2012. COPD diagnosis was consistent with the above COPD definition. Briefly, 148 COPD probands were originally recruited from the Fifth People's Hospital of Dongguan City. Among them, 113 were males with a mean age as 67.9 ± 12.9 year-old and 35 were females with a mean age as 65.6 ± 16.3 year-old. The probands' immediate family members who were mostly the probands' brothers, sisters and offspring were then asked to take a diagnostic test of COPD. To the exclusion of those who did not finish or achieve lung function test, 342 individuals were enrolled included 69 COPD cases and 273 healthy people. According to the definition of nuclear family, there were 95 nuclear families in the total population (n = 308), of which 26 (27.4%, n = 88) families had both parents, 29 (30.5%, n = 95) families had one single parent and at least one brother or sister, and 40 (42.1%, n = 125) had no available parents but at least two brothers or sisters. For these 95 COPD probands of nuclear families, 70 were males with a mean age as 67.0 ± 12.5 year-old and 25 were females with a mean age as 65.8 ± 17.9 year-old. All subjects are generally unrelated Han Chinese. Each subject provided data using a questionnaire as above described and donated 5 mL peripheral blood after a written informed consent was obtained. The analysis was approved by the institutional review boards of Guangzhou Medical University. The methods were carried out in accordance with the approved guidelines.
SNPs selection and genotyping. We selected 5 putatively functional tagSNPs that are rs10220974C > T in promoter, rs3764340C > G, rs12918952G > A in exons, rs383362G > T, rs12828G > A in 3′ -UTR as previously described 18 . The 5 SNPs were genotyped using the TaqMan allelic discrimination Assay on the ABI7500 system with self-designed primes and probes as shown in Supplementary Table S2 18 .
Statistical analysis. The chi-square test was used to assess differences in the frequency distribution of SNPs' genotypes between the cases and controls. The PROC HAPLOTYPE procedure in SAS/Genetics software was used to infer haplotype frequencies from the observed genotypes of SNPs. The unconditional logistic regression model without or with adjustment for these covariates, which were significantly associated with COPD risk including age, sex, pack-years smoked, biomass as fuels and sample source region, was used to estimate the odds ratio (OR) and 95% confidence interval (95%CI) to measure effect of each SNP as well as haplotype on COPD risk. The multiplicative interaction analysis was applied for assessing the interaction between each SNP and surrounding factors 40 . The family base-association test (FBAT) software was used to perform transmission disequilibrium test (TDT) and transmission disequilibrium of patient-normal (SDT) to analysis the effect of WWOX SNPs and haplotypes on susceptibility of COPD 41 . The oneway ANOVA test was performed to assess the effect of WWOX SNPs on pre-bronchodilator pulmonary function traits. All tests were two-sided by using the SAS software (version 9.2; SAS Institute, Cary, NC). P < 0.05 was considered to be statistically significant.