American Journal of Respiratory Cell and Molecular Biology

We examined the association between single-nucleotide polymorphisms (SNPs) previously associated with chronic obstructive pulmonary disease (COPD) and/or lung function with COPD and COPD-related phenotypes in a novel cohort of patients with severe to very severe COPD. We examined 315 cases of COPD and 330 Caucasian control smokers from Poland. We included three SNPs previously associated with COPD: rs7671167 (FAM13A), rs13180 (IREB2), and rs8034191 (CHRNA 3/5), and four SNPs associated with lung function in a genome-wide association study of general population samples: rs2070600 (AGER), rs11134242 (ADCY2), rs4316710 (THSD4), and rs17096090 (INTS12). We tested for associations with severe COPD and COPD-related phenotypes, including lung function, smoking behavior, and body mass index. Subjects with COPD were older (average age 62 versus 58 years, P < 0.01), with more pack-years of smoking (45 versus 33 pack-years, P < 0.01). CHRNA3/5 (odds ratio [OR], 1.89; 95% confidence interval [CI], 1.5–2.4; P = 7.4 × 10−7), IREB2 (OR, 0.69; 95% CI, 0.5–0.9; P = 3.4 × 10−3), and ADCY2 (OR, 1.35; 95% CI, 1.1–1.7; P = 0.01) demonstrated significant associations with COPD. FAM13A (OR, 0.8; 95% CI, 0.7–1.0; P = 0.11) approached statistical significance. FAM13A and ADCY2 also demonstrated a significant association with lung function. Thus, in severe to very severe COPD, we demonstrate a replication of association between two SNPs previously associated with COPD (CHRNA3/5 and IREB2), as well as an association with COPD of one locus initially associated with lung function (ADCY2).

These data demonstrate the association of several genes with chronic obstructive pulmonary disease (COPD) in a novel population with severe COPD. This work additionally replicates a novel finding of ADCY2 with COPD. These data provide a platform for future investigations into the genetics of COPD.

Chronic obstructive pulmonary disease (COPD) is the third leading cause of death in the United States, and its prevalence continues to grow (1). Although cigarette smoking is the most significant risk factor for COPD, cigarette smoking differentially affects lung function decline, and not all smokers will develop COPD (2). The individual response to cigarette smoke and other environmental factors is affected in part by genetic factors, and the development of COPD is the culmination of the environment acting in concert with a complex array of genetic traits (3).

Although the development of COPD is mediated by multiple genetic factors, to date few susceptibility genes other than α1-antitrypsin (SERPINA1) have been convincingly identified. Early candidate gene studies yielded inconsistent results (4). These studies were limited by small sample sizes and a lack of uniformity in COPD phenotypes, including the degree of airflow obstruction and intensity of smoking exposure in subjects examined. However, several recent genome-wide association studies (GWAS) have identified genomic regions that demonstrate a highly significant and reproducible association with COPD in large studies including subjects with a range of impaired lung function and smoking history. These GWAS regions are located in FAM13A on chromosome 4q22 (5), on chromosome 4q31 near HHIP (6), and near several genes in the region of 15q25, including IREB2 (7), CHRNA3, and CHRNA5 (8). In addition to these studies examining COPD as an outcome, several recent large-scale GWAS meta-analysis identified genes associated with pulmonary function in general-population samples of subjects with a diverse range of lung function and smoking exposures (6, 9, 10). Because COPD is defined by a reduction in lung function, some of these genes seem likely to play a role in the pathogenesis of COPD. In fact, both HHIP and FAM13A were associated with lung-function levels in GWAS studies of general-population samples (6, 10). Using a candidate gene approach, several of these pulmonary-function GWAS regions recently demonstrated an association with COPD, including loci on 4q24 near INTS12, on 6p21 near AGER, on 5p15 near ADCY2, and on 15q23 near THSD4 (11).

In a previous analysis, we demonstrated that HHIP single-nucleotide polymorphisms (SNPs) were also associated with COPD in a cohort of Polish subjects with severe COPD (12). Given this finding, we examined the relationship between previously reported COPD and lung function GWAS SNPs in this novel population comprised of cases with severe to very severe COPD. In addition, we examined the relationship between these SNPs and COPD-related phenotypes such as lung function, smoking behavior, and body mass index (BMI) in severe COPD. We hypothesized that a subset of SNPs previously associated with moderate to severe COPD and/or pulmonary-function levels would be associated with severe to very severe COPD. By investigating these associations in a homogeneous population with severe COPD, we believed we would be be better able to determine the genes responsible for COPD. In addition, by investigating COPD-related phenotypes such as lung function and smoking behavior, we hoped to differentiate more conclusively whether these genes were directly acting upon COPD susceptibility, or were in fact mediated by COPD risk factors such as smoking behavior.

Study Population

All subjects provided written, informed consent. Subjects participated at the Institute of Tuberculosis and Lung Disease (Warsaw, Poland), and genotyping and data analysis occurred at the Channing Laboratory of Brigham and Women's Hospital (Boston, MA). This study was approved by the institutional review boards at these institutions.

All subjects were white and from Poland, 40 to 80 years old, with at least 10 pack-years of smoking history. Cases manifested severe to very severe COPD (Guidelines for Obstructive Lung Disease Stage III or IV, post-bronchodilator FEV1 < 50% predicted, and FEV1/FVC < 0.7). Control subjects demonstrated normal lung function (post-bronchodilator FEV1, > 80% predicted; FEV1/FVC, > 0.7). Efforts were undertaken to frequency-match cases by age and sex. Subjects were excluded because of concomitant respiratory disorders, lung surgery, pregnancy, respiratory infection in the month before enrollment, or the presence of a first-degree or second-degree relative in the study.

Subjects performed standardized prebronchodilator and postbronchodilator spirometry, using the EasyOne Spirometer (NDD, Inc., Andover, MA), in accordance with American Thoacic Society/European Respiratory Society criteria (13). Postbronchodilator spirometry was performed approximately 20 minutes after administering 180 μg of albuterol via metered dose inhaler. In a few cases when subjects were unable to provide three acceptable maneuvers or slightly exceeded case spirometric criteria, those subjects' data were included at the investigators' discretion. FEV1 percent predicted was calculated using Hankinson equations (14). Subjects completed a modified American Thoacic Society Respiratory Epidemiology questionnaire (15), and blood was collected for genetic analyses.

Genetic Analysis

We chose genomic regions based on a review of the literature, and used the most significant reported SNP for each gene. We examined three candidate genes associated with COPD in populations with moderate to severe COPD (FAM13A, IREB2, and CHRNA3/5) and SNPs in or near four candidate genes previously associated with lung function and COPD (INTS12, ADCY2, THSD4, and AGER). SNPs were genotyped using TaqMan assays (Applied Biosystems, Foster City, CA).

Statistical Analysis

We investigated the association between the chosen SNPs and COPD case/control status, average cigarettes smoked per day (CPD), pack-years, lung function, and BMI. All demographic data were analyzed using SAS version 9.2 (SAS, Inc., Cary, NC). All genetic analyses were performed using PLINK, version 1.07 (16). Logistic regression analysis was performed to test the association between each SNP with severe COPD case/control status, controlling for age, sex, and pack-years of smoking. Linear regression was performed to assess the relationship between SNPs and quantitative phenotypes. Analyses of CPD and pack-years were adjusted for age and sex, and analyses of FEV1 percent predicted, FEV1/FVC, and BMI were additionally adjusted for pack-years. We performed a parallel analysis among all study subjects, with the additional use of COPD case/control status as a covariate. A conditional analysis between SNPs from CHRNA3/5 and IREB2 with COPD was performed by including the SNP that was adjusted for as a covariate in the logistic regression analysis. To assess the relationship between smoking behavior, candidate genes, and COPD, subjects were divided into equal tertiles of pack-years. In each tertile, the relationship between the genetic variants (in CHRNA3/5, IREB2, and ADCY2) with COPD was examined by means of logistic regression, using all smoking control subjects as the reference group, and adjusting for age and sex.

Three hundred fifteen smoking patients with COPD and 330 smoking control subjects were included in the genetic association analysis (Table 1). More males than females participated in this study. However, sex was equally distributed between severe COPD cases and control subjects. Cases on average were older than control subjects (aged 62 years versus 58 years, P < 0.0001) and demonstrated a lower BMI (26.7 kg/m2 versus 27.7 kg/m2, respectively; P < 0.05). Cases had greater average smoking exposure, including pack-years (44.5 versus 33.5 pack-years, P < 0.0001) and average number of cigarettes per day (23.6 versus 20.8, respectively; P < 0.0001), and had started smoking at a younger age (age 18.7 years versus 19.5 years, respectively; P < 0.0001). Fewer cases than control subjects were active smokers at the time of study enrollment (P < 0.0001). By study design, cases had worse lung function than control subjects, including a lower FEV1 percent predicted (30.36% versus 102.5%, respectively) and FEV1/FVC (0.36 versus 0.77, respectively).


CasesControlsP Value
Age, yr61.95 (7.33)58.28 (7.26)<0.0001
Male, %220 (69.84)222 (67.27)0.50
Current smoker, n (%)80 (25.40)163 (49.39)<0.0001
Pack-years44.50 (22.33)33.37 (14.94)<0.0001
CPD23.61 (9.3)20.75 (8.13)<0.0001
Age started smoking18.68 (3.94)19.49 (3.99)0.01
FEV1, L0.89 (0.32)3.13 (0.62)NA
FEV1% predicted, %30.36 (9.60)102.5 (12.11)NA
FEV1/FVC0.36 (0.098)0.77 (0.045)NA
BMI, kg/m226.68 (6.52)27.72 (6.08)0.037

Definition of abbreviations: BMI, body mass index; CPD, cigarettes per day; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; NA, not available.

Values are reported as means (standard deviations) or means (percentages). Sex and current smoking status were assessed according to the Fisher exact test. All continuous variables were compared using t tests.

SNPs located in or near three genes previously associated with COPD (CHRNA3/5 and IREB2 on chromosome 15q25, and FAM13A on chromosome 4q22) and four genes previously associated with lung function (INTS12 on chromosome 4q24, ADCY2 on chromosome 5p15, THSD4 on chromosome 15q23, and AGER on chromosome 6p21) were examined for associations with severe COPD case–control status (Table 2). Of these, the minor (C) allele of the IREB2 SNP, rs13180 (odds raio [OR], 0.69; 95% confidence interval [CI], 0.53–0.88; P = 0.003) and the minor (C) allele from the CHRNA3/5 SNP rs8034191 (OR, 1.89; 95% CI, 1.47–2.43; P = 7.4 × 10−7) were associated with COPD, in the same direction as previously reported. The minor allele of the FAM13A SNP, rs7671167, did not reach statistical significance when examining the association with COPD. However, the OR was in the same direction of effect as in previous studies (OR, O.82; 95% CI, 0.65–1.04; P = 0.11). Of the genes originally associated with lung-function levels, an SNP in ADCY2, rs11134242, was associated with an increased risk for COPD (OR, 1.35; 95% CI, 1.1–1.7; P = 0.01), which was in the same direction as previously reported in an analysis of COPD affection status (11).


SNPChrGeneOR (95% CI)P Value
rs76711674q22.1FAM13A0.82 (0.65–1.04)0.107
rs1318015q25IREB20.69 (0.53–0.88)3.4 × 10−3
rs803419115q25CHRNA3/51.89 (1.47–2.43)7.4 × 10−7
rs170360904q24INTS120.84 (0.48–1.46)0.54
rs111342425p15ADCY21.35 (1.06– 1.7)0.01
rs431671015q23THSD41.16 (0.84–1.58)0.37
rs20706006p21AGER1.41 (0.47–4.19)0.54

Definition of abbreviations: Chr, chromosome; CI, confidence interval; COPD, chronic obstructive pulmonary disease; OR, odds ratio; SNP, single-nucleotide polymorphism.

All analyses were performed using logistic regression in PLINK, with age, sex, and pack-years as covariates. Each OR represents odds of COPD case status, with minor allele as the reference allele.

The SNPs from IREB2 (rs13180) and CHRNA3/5 (rs8034191) were in weak linkage disequilibrium (LD) with each other (R2 = 0.21), consistent with previous reports. To determine the independent effect of each SNP on COPD case status, we performed a conditional analysis examining the association of these SNPs with COPD. We examined the relationship between each SNP with COPD individually, while controlling for the presence of the other linked SNP (Table 3). When controlling for the presence of IREB2, CHRNA3/5 remained significantly associated with COPD (P = 0.0001). However, when controlling for the presence of CHRNA3/5, the IREB2 SNP was no longer significantly associated with COPD (P = 0.45).


GeneSNPConditioned onOR (95% CI)P Value
IREB2rs13180rs80341910.89 (0.67–1.19)0.45
CHRNA3/5rs8034191rs131801.75 (1.31–2.32)1.4 × 10−4

Definition of abbreviations: CI, confidence interval; OR, odds ratio; SNP, single-nucleotide polymorphism.

Logistic regression was performed in PLINK, using conditional analysis with the conditioned SNP included as a covariate. Age, sex, and pack-years comprised additional covariates.

We next examined the relationship between these SNPs and COPD-related phenotypes, including lung function (FEV1 % predicted and FEV1/FVC), smoking history (pack-years and average cigarettes per day), and BMI (Table 4). Regarding measures of lung function, the presence of the FAM13A minor allele was associated with a 1.39% increase in FEV1/FVC ratio among all study subjects (P = 8.0 × 10−4), as well as a 1.8% increase in FEV1/FVC among cases only (P = 0.02). Among all subjects examined together, the minor allele of the ADCY2 SNP was associated with a decrease in FEV1/FVC of 2.5% (P = 0.03). When examining control subjects only, no significant association was evident between these SNPs and any of the measures of lung function. No significant relationships were observed between any of these SNPs and measures of smoking behavior, FEV1 percent predicted, or BMI (data not shown), both in cases only and in the entire cohort.


SNPCPDPack-YearsFEV1% PredictedFEV1/FVC
Cases only
FAM13A0.06 (0.76), P = 0.940.85 (1.80), P = 0.640.19 (0.78), P = 0.801.79 (0.8), P = 0.02
IREB2−0.55 (0.78), P = 0.48−1.15 (1.83), P = 0.53−0.45 (0.81), P = 0.58−0.38 (0.8), P = 0.65
CHRNA3/50.37 (0.73), P = 0.620.70 (1.72), P = 0.68−0.83 (0.75), P = 0.270.15 (0.8), P = 0.85
ADCY20.86 (0.73), P = 0.242.53 (1.69), P = 0.14−1.10 (0.76), P = 0.15−0.15 (0.78), P = 0.85
All subjects
FAM13A−0.53 (0.5), P = 0.270.29 (1.0), P = 0.78−0.38 (0.61), P = 0.541.39 (0.41), P = 0.0008
IREB2−0.62 (0.51), P = 0.22−0.45 (1.1), P = 0.67−0.82 (0.64), P = 0.20−0.38 (0.45), P = 0.40
CHRNA3/50.16 (0.5), P = 0.75−0.40 (1.1), P = 0.710.19 (0.63), P = 0.770.09 (0.4), P = 0.83
ADCY2−0.33 (0.5), P = 0.49−1.16 (1.0), P = 0.250.91 (0.61), P = 0.13−2.5 (1.2), P = 0.03

Definition of abbreviations: COPD, chronic obstructive pulmonary disease; CPD, cigarettes per day; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; SNP, single-nucleotide polymorphism.

Data presented are β (standard error) with two-sided P values. Cigarettes per day and pack-years were adjusted for age and sex, FEV1, FEV1/FVC, and body mass indices were adjusted for age, sex, and pack-years. The analysis including all subjects was additionally adjusted for COPD case–control status.

Because both IREB2 and CHRNA3/5 are located in a region that was also associated with nicotine addiction and smoking behavior (1721), we investigated whether these SNPs demonstrated an independent effect on COPD case–control status, or if their effect was in fact the result of their impact on smoking behavior. When examining each SNP in relation to smoking behavior, neither CHRNA3/5 nor IREB2 was significantly associated with average cigarettes per day or pack-years, either in cases only or in all subjects.

We examined the relationship between each of the three SNPs significantly associated with severe COPD case–control status within tertiles of smoking intensity, based on pack-years of smoking, using control subjects as the reference group (Table 5). If the effect of the SNP is independent from smoking behavior, we might expect to see either a similar association within each tertile or potentially the strongest association in the lowest tertile of smoking behavior, in which COPD genetic susceptibility unrelated to nicotine addiction may be most important. The association between both IREB2 and CHRNA3/5 and COPD became more significant between tertiles 1 and tertiles 2–3 (P = 0.18, 0.004, and 0.008, respectively, for IREB2, and P = 0.002, 3.8 × 10−5, and 7.4 × 10−5, respectively, for CHRNA3/5). When the association between IREB2 and COPD was conditioned on the presence of CHRNA3/5, the relationship was no longer significant (P = 0.93, 0.3, and 0.24), whereas the association between CHRNA3/5 and COPD remained significant with increasing tertile (P = 0.01, 0.007, and 0.003008, respectively), despite conditioning for IREB2. We performed a similar analysis with ADCY2 and smoking behavior, and found a significant association with COPD only in the lowest tertile of smoking behavior (OR, 1.49; 95% CI, 1.1–2.1; P = 0.018).


GeneTertile 1Tertile 2Tertile 3
IREB20.79 (0.6–1.1), P = 0.180.59 (0.4–0.8), P = 0.0040.6 (0.4–0.9), P = 0.008
IREB2*1.02 (0.68–1.53), P = 0.930.8 (0.52–1.22), P = 0.30.78 (0.51–1.19), P = 0.24
CHRNA3/51.7 (1.2–2.4), P = 0.0022.1 (1.5–2.9), P = 3.8 × 10−52.1 (1.5–3.1), P = 7.4 × 10−5
CHRNA3/5*1.7 (1.1–2.5), P = 0.011.8 (1.2–2.6), P = 0.0071.86 (1.2–2.8), P = 0.003
ADCY21.49 (1.1–2.1), P = 0.0181.24 (0.89–1.7), P = 0.21.21 (0.9–1.7), P = 0.26

Definition of abbreviations: CI, confidence interval; COPD, chronic obstructive pulmonary disease; OR, odds ratio.

Data are presented as OR (95% CI), with P value (two-sided). All cases were divided into equal tertiles of smoking behavior (pack-years of smoking), and compared against all control subjects as the reference group for COPD case–control status.

*IREB2 was additionally adjusted for the presence of the CHRNA3/5 minor allele. CHRNA3/5 was additionally adjusted for the presence of the IREB2 minor allele.

We have demonstrated that SNPs in IREB2, CHRNA3/5, and ADCY2 are associated with severe COPD in a Polish case–control population. Although FAM13A was not significantly associated with COPD in this study, the effect measure was in the same direction as previously reported, and therefore the lack of significance may be related to our small sample size. This is the first demonstration, to the best of our knowledge, of an association between an SNP near ADCY2 and severe COPD, and confirms a recent finding of an association between this SNP and COPD in a series that included a more heterogeneous population with COPD (11).

IREB2 and CHRNA3/5 are located on a region of chromosome 15q25 that is particularly compelling for investigating the genetic components of COPD. This region contains a number of genes with biological plausibility for a role in the development of COPD. However, the significant linkage disequilibrium in this region complicates the ability to isolate causal genes and variants.

The gene cluster CHRNA3/CHRNA5/CHRNB4 encodes nicotinic acetylcholine receptor subunits expressed in the central nervous system and bronchial epithelium, and is responsive to nicotine (22). These receptors have been implicated in nicotine addiction, and appear to be up-regulated with chronic tobacco use (23). CHRNA3/5 has been associated with the risk for nicotine dependence through mRNA brain expression levels (24) and through candidate gene analyses (25), dense genotyping (19), and GWAS (17) (21, 26) in diverse populations, including smokers with normal lung function and lung disease as well as European and African-American populations. The CHRNA3/5 SNP that we investigated, rs8034191, was found to be associated with heavier smoking behaviors (18) as well as increased susceptibility to lung cancer (27), COPD (8), and radiographically determined emphysema (28).

Although these studies suggest that CHRNA3/5 acts on COPD case–control status by influencing smoking behavior, some evidence also exists for an independent effect of these genes on the development of COPD (8, 29). In the first GWAS of COPD, the CHRNA3/5 SNP rs8034191 was associated with an increased risk for COPD (8). In that study, investigators did identify a gene-by-environment interaction between current smoking and this variant on COPD, indicating that some of the increased risk for COPD associated with this region could be mediated by smoking behavior. However, as in the present study, they did not identify an association between pack-years and this locus, and the association between CHRNA3/5 and COPD remained robust even after adjusting for current smoking status and pack-years. Using mediation analysis, other investigators examined the relationship between rs1051730 and COPD, which is in LD with the CHRNA3/5 SNP we examined, and found that smoking may mediate the relationship between this SNP and COPD (30).

In our study, we demonstrated that the CHRNA3/5 SNP is associated with an increased risk for COPD, even after adjusting for pack-years of smoking. In addition, we did not identify an association between the CHRNA3/5 SNP and pack-years of smoking or cigarettes per day in cases, control subjects, or the entire group taken together. Although we are limited in our ability to assess the functional role that CHRNA3/5 plays in increasing the risk for COPD, our findings suggest an effect that may be independent from smoking behavior. On the other hand, we note that the association between CHRNA3/5 and COPD is strongest in the top tertile of smoking intensity, suggesting a gene-by-environment interaction. Further investigation, including the fine mapping of this region and improved smoking intensity phenotypes, may help tease apart these effects.

IREB2 is also located in the same genomic region as CHRNA3/5, and the SNP in our study was in moderate LD with rs8034191 (r2 = 0.21). The protein product of this gene encodes the iron-response protein IRP2. This protein, together with IRP1, is involved in iron metabolism and the response to hypoxemia, and has been demonstrated to affect mitochondrial iron stores (3133). IREB2 protein concentrations are increased in emphysematous lungs when compared with control samples, as are IREB2 mRNA concentrations, and IREB2 was associated with FEV1 level and COPD in candidate gene studies (7).

In this study, we demonstrated an association between an IREB2 SNP and severe COPD case status. After adjustment for the presence of the CHRNA3/5 SNP, this association is no longer significant, indicating that the association may be driven by linkage disequilibrium with CHRNA3/5 SNPs. It is also possible that IREB2 is not involved in severe COPD, at least in this Polish population. However, because the direction of effect remained the same despite a loss of significance, our study may more likely have been underpowered to demonstrate a persistent impact from IREB2.

Although we did not demonstrate a significant association between a FAM13A SNP and COPD, we did demonstrate that the minor allele of this SNP was associated with improved lung function, both in cases only and in the entire study population. These findings are consistent with those of a previous COPD association study that identified an association between FAM13A and lung function in two family-based studies of COPD, demonstrating a reduced odds of COPD associated with the minor allele of this SNP (5). This association was also evident in a recent population-based meta-analyses of lung function GWAS (10). However, other lung function GWAS have failed to identify this association (9).

When examining SNPs previously associated with lung function in general-population samples, we found that an ADCY2 SNP was associated with COPD in our population. None of these previously reported lung function SNPs were associated with lung-function measures when examining cases only. However, when examining all subjects together, the ADCY2 SNP demonstrated a significant association with spirometric measures of lung function.

ADCY2 encodes adenylyl cyclase–2, a member of the adenylyl cyclase family involved in G protein–coupled receptor signaling. The mRNA from this gene was found to be expressed in the human brain, smooth muscle, and testes, and mildly expressed in lung tissue (34). This gene was identified as part of a gene cluster associated with the ability to quit smoking in a previous genetic analysis (35). In their lung function meta-analysis, Hancock and colleagues (10) demonstrated that SNPs from this gene were associated with FEV1 in the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) Consortium, although their findings did not reach genome-wide significance. In a follow-up analysis, examining the ADCY2 gene in greater depth, Castaldi and colleagues identified the minor allele of rs11134242 from ADCY2 as significantly associated with COPD. Interestingly, this SNP is only in weak LD with the CHARGE SNPs (rs6555465 and rs7710510) near the same gene (11). To the best of our knowledge, the present study represents the first replication of the novel finding of association by Castaldi and colleagues (11).

We additionally examined the relationship of IREB2, CHRNA3/5, and ADCY2 with COPD among tertiles of smoking behavior. Although we did not identify an association between these SNPS and smoking behavior when smoking intensity was measured as a quantitative variable (i.e., pack-years), we were also interested in determining if these associations were more significant in heavy versus light smokers. By dividing smokers into tertiles of smoking behavior, we attempted to distinguish whether the noted association between the SNPs was related to smoking behavior or COPD susceptibility. If the association were stronger in the lowest smoking tertile (light smokers), this would provide evidence of a greater relationship between the selected SNPs and COPD susceptibility. If the relationship became stronger with increased tertiles of smoking (heavy smokers), we felt this would indicate that smoking behavior was more likely to be driving the relationship. In our study, the association between CHRNA3/5 and COPD was strongest among the upper two tertiles of smokers, which may indicate that this association is driven by smoking-related behavior. In contrast, the association between ADCY2 and COPD was strongest among the lowest tertile of smokers. Taken in combination with our finding of an association between ADCY2 and FEV1/FVC in our entire study population, ADCY2 appears to affect COPD case–control status through a nonsmoking-related effect on the development of airflow obstruction.

This study contains several limitations. As in any candidate gene analysis, we are limited by our a priori hypothesis in regard to which SNPs and genes to investigate, and which phenotypes to study. Using a Bonferroni correction for the multiple testing of seven individual SNPs and six separate outcomes, a P value of 6.25 × 10−3 would be required for statistical significance. With this more conservative approach, the associations between IREB2 and CHRNA3/5 and COPD remain significant, and the association between CHRNA3/5 and COPD remained robust after adjusting for the IREB2 association. The association between FAM13A and FEV1/FVC also remained significant after adjusting for multiple comparisons. Although we were able to identify associations, we are not able to identify causal mechanisms. For example, although our data provide intriguing evidence for a role of CHRNA3/5 and IREB2 in the development of COPD, we were not able to demonstrate an independent effect of IREB2 on COPD susceptibility. Although limited by our small sample size, we were still able to demonstrate several significant associations with COPD. A larger sample size may have allowed for greater evidence of the role of FAM13A in COPD. We also lacked an additional replication population of severe smokers. However, our study functions as a confirmation of prievous associations with COPD, and further strengthens the association of CHRNA3/5, IREB2, and ADCY2 with COPD, especially given our more extreme case phenotype. Future studies should include investigations in additional populations of smokers with severe to very severe COPD.

We have attempted to distinguish distinct roles for IREB2 and CHRNA3/5 in the development of COPD that are independent from nicotine addiction. Here we demonstrate that CHRNA3/5 is associated with COPD even while controlling for cigarette smoking exposure. However, our measures of nicotine addiction (i.e., cigarettes per day and pack-years) may not have been adequately replicated the toxic effects of tobacco smoke. Nevertheless, these same measures were used in many previous GWAS assessing smoking behavior, and we feel cigarettes per day to be a highly heritable, robust marker of nicotine addiction (26).

We examined the relationship between our SNPs and COPD-related phenotypes, including spirometry and smoking behavior, in our study population. We analyzed these traits in cases only as well as in our entire population using linear regression, including COPD case–control status as a covariate. Although this procedure is not a substitute for population-based sampling, it avoids potential biases in analyzing quantitative phenotypes in a case–control–based population (36, 37).

In conclusion, in a population with severe to very severe COPD, we confirm the previously identified association between CHRNA3/5 and COPD, and this association remained robust after adjusting for smoking behavior. We also demonstrate that ADCY2 is associated with severe COPD and with lung-function levels. Our results also provide some evidence that IREB2 and FAM13A are associated with COPD. Further studies will be required to elucidate the functional variants that confer a risk for COPD in these associated genomic regions.

1. Minino A, Xu J, Kochanek K. Deaths: preliminary data for 2008. Natl Vital Stat Rep 2010;59:152.
2. Wan ES, Silverman EK. Genetics of COPD and emphysema. Chest 2009;136:859866.
3. Silverman EK, Vestbo J, Agusti A, Anderson W, Bakke PS, Barnes KC, Barr RG, Bleecker ER, Boezen HM, Burkart KM, et al.. Opportunities and challenges in the genetics of COPD 2010: an International COPD Genetics Conference report. COPD 2011;8:121135.
4. Hersh CP, Demeo DL, Lange C, Litonjua AA, Reilly JJ, Kwiatkowski D, Laird N, Sylvia JS, Sparrow D, Speizer FE, et al.. Attempted replication of reported chronic obstructive pulmonary disease candidate gene associations. Am J Respir Cell Mol Biol 2005;33:7178.
5. Cho MH, Boutaoui N, Klanderman BJ, Sylvia JS, Ziniti JP, Hersh CP, DeMeo DL, Hunninghake GM, Litonjua AA, Sparrow D, et al.. Variants in FAM13A are associated with chronic obstructive pulmonary disease. Nat Genet 2010;42:200202.
6. Wilk JB, Chen TH, Gottlieb DJ, Walter RE, Nagle MW, Brandler BJ, Myers RH, Borecki IB, Silverman EK, Weiss ST, et al.. A genome-wide association study of pulmonary function measures in the Framingham Heart Study. PLoS Genet 2009;5:e1000429.
7. DeMeo DL, Mariani T, Bhattacharya S, Srisuma S, Lange C, Litonjua A, Bueno R, Pillai SG, Lomas DA, Sparrow D, et al.. Integration of genomic and genetic approaches implicates IREB2 as a COPD susceptibility gene. Am J Hum Genet 2009;85:493502.
8. Pillai SG, Ge D, Zhu G, Kong X, Shianna KV, Need AC, Feng S, Hersh CP, Bakke P, Gulsvik A, et al.. A genome-wide association study in chronic obstructive pulmonary disease (COPD): identification of two major susceptibility loci. PLoS Genet 2009;5:e1000421.
9. Repapi E, Sayers I, Wain LV, Burton PR, Johnson T, Obeidat M, Zhao JH, Ramasamy A, Zhai G, Vitart V, et al.. Genome-wide association study identifies five loci associated with lung function. Nat Genet 2010;42:3644.
10. Hancock DB, Eijgelsheim M, Wilk JB, Gharib SA, Loehr LR, Marciante KD, Franceschini N, van Durme YM, Chen TH, Barr RG, et al.. Meta-analyses of genome-wide association studies identify multiple loci associated with pulmonary function. Nat Genet 2010;42:4552.
11. Castaldi PJ, Cho MH, Litonjua AA, Bakke P, Gulsvik A, Lomas DA, Anderson W, Beaty TH, Hokanson JE, Crapo JD, et al.. The association of genome-wide significant spirometric loci with COPD susceptibility. Am J Respir Cell Mol Biol 2011;45:11471153.
12. Zhou X, Baron RM, Hardin M, Cho MH, Zielinski J, Hawrylkiewicz I, Sliwinski P, Hersh CP, Mancini JD, Lu K, et al.. Identification of a chronic obstructive pulmonary disease genetic determinant that regulates HHIP. Hum Mol Genet 2012;21:13251335.
13. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A, Crapo R, Enright P, van der Grinten CP, Gustafsson P, et al.. Standardisation of spirometry. Eur Respir J 2005;26:319338.
14. Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S. population. Am J Respir Crit Care Med 1999;159:179187.
15. Ferris BG. Epidemiology standardization project (American Thoracic Society). Am Rev Respir Dis 1978;118:1120.
16. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, Maller J, Sklar P, de Bakker PI, Daly MJ, et al.. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007;81:559575.
17. Thorgeirsson TE, Geller F, Sulem P, Rafnar T, Wiste A, Magnusson KP, Manolescu A, Thorleifsson G, Stefansson H, Ingason A, et al.. A variant associated with nicotine dependence, lung cancer and peripheral arterial disease. Nature 2008;452:638642.
18. Stevens VL, Bierut LJ, Talbot JT, Wang JC, Sun J, Hinrichs AL, Thun MJ, Goate A, Calle EE. Nicotinic receptor gene variants influence susceptibility to heavy smoking. Cancer Epidemiol Biomarkers Prev 2008;17:35173525.
19. Saccone NL, Wang JC, Breslau N, Johnson EO, Hatsukami D, Saccone SF, Grucza RA, Sun L, Duan W, Budde J, et al.. The CHRNA5CHRNA3CHRNB4 nicotinic receptor subunit gene cluster affects risk for nicotine dependence in African-Americans and in European-Americans. Cancer Res 2009;69:68486856.
20. Saccone NL, Saccone SF, Hinrichs AL, Stitzel JA, Duan W, Pergadia ML, Agrawal A, Breslau N, Grucza RA, Hatsukami D, et al.. Multiple distinct risk loci for nicotine dependence identified by dense coverage of the complete family of nicotinic receptor subunit (CHRN) genes. Am J Med Genet B Neuropsychiatr Genet 2009;150B:453466.
21. Saccone NL, Culverhouse RC, Schwantes-An TH, Cannon DS, Chen X, Cichon S, Giegling I, Han S, Han Y, Keskitalo-Vuokko K, et al.. Multiple independent loci at chromosome 15q25.1 affect smoking quantity: a meta-analysis and comparison with lung cancer and COPD. PLoS Genet 2010;6(8):e1001053.
22. Greenbaum L, Rigbi A, Teltsh O, Lerer B. Role of genetic variants in the CHRNA5CHRNA3CHRNB4 cluster in nicotine dependence risk: importance of gene–environment interplay. Mol Psychiatry 2009;14:828830.
23. Dani JA. Overview of nicotinic receptors and their roles in the central nervous system. Biol Psychiatry 2001;49:166174.
24. Wang JC, Cruchaga C, Saccone NL, Bertelsen S, Liu P, Budde JP, Duan W, Fox L, Grucza RA, Kern J, et al.. Risk for nicotine dependence and lung cancer is conferred by mRNA expression levels and amino acid change in CHRNA5. Hum Mol Genet 2009;18:31253135.
25. Weiss RB, Baker TB, Cannon DS, von Niederhausern A, Dunn DM, Matsunami N, Singh NA, Baird L, Coon H, McMahon WM, et al.. A candidate gene approach identifies the CHRNA5A3B4 region as a risk factor for age-dependent nicotine addiction. PLoS Genet 2008;4:e1000125.
26. Berrettini W, Yuan X, Tozzi F, Song K, Francks C, Chilcoat H, Waterworth D, Muglia P, Mooser V. Alpha-5/alpha-3 nicotinic receptor subunit alleles increase risk for heavy smoking. Mol Psychiatry 2008;13:368373.
27. Hung RJ, McKay JD, Gaborieau V, Boffetta P, Hashibe M, Zaridze D, Mukeria A, Szeszenia-Dabrowska N, Lissowska J, Rudnai P, et al.. A susceptibility locus for lung cancer maps to nicotinic acetylcholine receptor subunit genes on 15q25. Nature 2008;452:633637.
28. Pillai SG, Kong X, Edwards LD, Cho MH, Anderson WH, Coxson HO, Lomas DA, Silverman EK. Loci identified by genome-wide association studies influence different disease-related phenotypes in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2010;182:14981505.
29. Schwartz AG, Cote ML, Wenzlaff AS, Land S, Amos CI. Racial differences in the association between SNPs on 15q25.1, smoking behavior, and risk of non–small cell lung cancer. J Thorac Oncol 2009;4:11951201.
30. Wang J, Spitz MR, Amos CI, Wilkinson AV, Wu X, Shete S. Mediating effects of smoking and chronic obstructive pulmonary disease on the relation between the CHRNA5A3 genetic locus and lung cancer risk. Cancer 2010;116:34583462.
31. LaVaute T, Smith S, Cooperman S, Iwai K, Land W, Meyron-Holtz E, Drake SK, Miller G, Abu-Asab M, Tsokos M, et al.. Targeted deletion of the gene encoding iron regulatory protein–2 causes misregulation of iron metabolism and neurodegenerative disease in mice. Nat Genet 2001;27:209214.
32. Leipuviene R, Theil EC. The family of iron responsive RNA structures regulated by changes in cellular iron and oxygen. Cell Mol Life Sci 2007;64:29452955.
33. Galy B, Ferring-Appel D, Sauer SW, Kaden S, Lyoumi S, Puy H, Kolker S, Grone HJ, Hentze MW. Iron regulatory proteins secure mitochondrial iron sufficiency and function. Cell Metab 2010;12:194201.
34. Ludwig MG, Seuwen K. Characterization of the human adenylyl cyclase gene family: cDNA, gene structure, and tissue distribution of the nine isoforms. J Recept Signal Transduct Res 2002;22:79110.
35. Uhl GR, Liu QR, Drgon T, Johnson C, Walther D, Rose JE, David SP, Niaura R, Lerman C. Molecular genetics of successful smoking cessation: convergent genome-wide association study results. Arch Gen Psychiatry 2008;65:683693.
36. Monsees GM, Tamimi RM, Kraft P. Genome-wide association scans for secondary traits using case–control samples. Genet Epidemiol 2009;33:717728.
37. Lin DY, Zeng D. Proper analysis of secondary phenotype data in case–control association studies. Genet Epidemiol 2009;33:256265.
Correspondence and requests for reprints should be addressed to Megan Hardin, M.D., Channing Laboratory, Brigham and Women's Hospital, 181 Longwood Ave., Boston, MA 02115. E-mail:

This work was supported by the National Institutes of Health grants T32 HL07427–29 (M.H.), R01HL084323 (E.K.S.), R01HL075478 (E.K.S.), P01 HL083069 (E.K.S.), P01 HL105339 (E.K.S.), and 5R01HL084323-04.

Originally Published in Press as DOI: 10.1165/rcmb.2012-0011OC on March 29, 2012

Author disclosures


No related items
American Journal of Respiratory Cell and Molecular Biology

Click to see any corrections or updates and to confirm this is the authentic version of record