SNPs in the chromosome 15 CHRNA3/CHRNA5/LOC123688/IREB2 region have been shown to have associations with lung cancer and COPD unrelated to AAT deficiency. In our current analysis, SNPs in IREB2, LOC123688 and CHRNA3 genes were shown to be associated with lung function phenotypes in AAT deficient subjects (all PI ZZ) from the AAT Genetic Modifiers Study, and suggested a potential sex-specific effect. Replication in another cohort of AAT deficient subjects from the UK showed that a SNP in IREB2 was also associated with emphysema in men. This suggests that chromosome 15q region genes that were found by GWA studies and gene expression analysis of lung tissue samples may also be modifier genes of COPD and emphysema in AAT deficient subjects.
CHRNA3 was associated with lung cancer in three separate large GWA studies. This gene was associated with COPD by GWA and the association was replicated in two other COPD cohorts. There have also been recent reports of an association with smoking addiction [
27], so it is unclear whether the lung cancer and COPD associations relate to smoking behavior, another aspect of lung biology, or both.
CHRNA3 is a subunit gene of the nicotinic cholinergic receptor and expressed in autonomic ganglia and brain but is also expressed in bronchial and non-bronchial epithelial cells [
28]. Expression in lung cancer cells and signal transduction and apoptosis studies suggests a potential role in carcinogenesis [
29]. Interestingly, there are not many observations of lung cancer in patients with AAT deficiency, perhaps because of mortality associated with the development of severe COPD at an early age. Whether there is a common mechanism unrelated to smoking in the pathogenesis of lung cancer and COPD, or whether these previously reported associations relate to smoking addiction is unclear.
IREB2 is a protein of iron-responsive elements (IREs) and is regulated in response to iron and oxygen supply [
30]. IREB2-/- mice have aberrant iron homeostasis and accumulate iron in the intestine and the central nervous system(CNS); the CNS accumulation may lead to neurodegenerative disease [
31]. Excess iron can be toxic, but the mechanism of neurodegenerative disease is unclear; work is in progress to further characterize the functional pathways impacted by
IREB2 in the lung.
IREB2 was found to be differentially expressed according to lung function by microarray experiments, and the SNPs in
IREB2 showed associations in both a COPD case-control study and family-based studies including the Boston Early-onset COPD and International COPD Genetics Network studies [
18]. In a recent report,
IREB2 polymorphisms were associated with COPD susceptibility in a European population [
32]. Interestingly, rs2568494 was significantly associated with COPD in three studies including our current study.
Previous studies of AAT deficient subjects showed that lung function was lower in men than women [
33], and previous analyses in the AAT Genetic Modifiers Study also showed lower lung function in men [
19]. Our current study suggests that genetic modifier effects of
IREB2 and
CHRNA3 may be more prominent in males--potentially contributing to some of the sex-specific features of COPD susceptibility and severity among PI ZZ subjects, although a larger sample size is needed to verify a gene-by-sex interaction.
In this study, there was no association between
IREB2 and
CHRNA3 genes and smoking intensity. In the AAT Genetic Modifiers Study, results showed no association when the cohort was stratified by smoking history (ever smokers versus never smokers). However, there was a marginal interaction of rs1051730 with smoking. In the UK study, there were significant smoking interactions of rs2568494 and rs8034191. Smoking markedly increases the risk of COPD and lowers the age-of-onset of COPD in AAT deficient subjects [
6,
19], and despite small sample sizes, we found reasonable evidence for gene-by-smoking interactions in the chromosome 15q region.
There are several limitations in this study. Multiple statistical comparisons are a potential concern in any complex disease genetics study. Adjusting for either 3 genes or 9 SNPs tested, a p value of 0.02 is marginal. Additionally, the association with pulmonary function did not replicate in the UK population, potentially due to phenotypic heterogeneity between the two cohorts. Specifically, the UK subjects have lower mean FEV1 and potentially more emphysema, both of which could influence non-replication. Of note, the association with emphysema was investigated only in the UK population as chest CT scan data collection was not part of the AAT Genetic Modifiers Study. Considering that these SNPs (rs2568494 in IREB2, rs8034191 in LOC123688, and rs1051730 in CHRNA3) were associated with intermediate phenotypes of COPD in other populations and that we include an independent AAT deficient replication cohort, our result are likely meaningful. Additionally, this test- replication approach is even more appealing since all subjects were homozygous recessive for the AAT risk locus (PI ZZ). Also, replication of our results showed association with emphysema, a less heterogeneous pulmonary phenotype. The associated SNPs included two intronic (rs2568494, rs8034191) and one synonymous exonic (rs1051730) SNP. The exonic SNP was not associated with COPD-related phenotypes in the UK cohort. Another limitation of our current study is that rare functional variants in this chromosome 15 region may be contributing to the role of these genes in COPD; genome sequencing efforts in AAT deficient cohorts would be valuable to study rare variant associations. Functional data for associated variants are currently lacking, but many groups are pursuing functional work on this chromosome 15 region.