- Top of page
- Materials and methods
- Supporting Information
Nicotine binds to nicotinic acetylcholine receptors and studies in animal models have shown that α4β2 receptors mediate many behavioral effects of nicotine. Human genetics studies have provided support that variation in the gene that codes for the α4 subunit influences nicotine dependence (ND), but the evidence for the involvement of the β2 subunit gene is less convincing. In this study, we examined the genetic association between variation in the genes that code for the α4 (CHRNA4) and β2 (CHRNB2) subunits of the nicotinic acetylcholine receptor and a quantitative measure of lifetime DSM-IV ND symptom counts. We performed this analysis in two longitudinal family-based studies focused on adolescent antisocial drug abuse: the Center on Antisocial Drug Dependence (CADD, N = 313 families) and Genetics of Antisocial Drug Dependence (GADD, N = 111 families). Family-based association tests were used to examine associations between 14 single nucleotide polymorphisms (SNPs) in CHRNA4 and CHRNB2 and ND symptoms. Symptom counts were corrected for age, sex and clinical status prior to the association analysis. Results, when the samples were combined, provided modest evidence that SNPs in CHRNA4 are associated with ND. The minor allele at both rs1044394 (A; Z = 1.988, P = 0.047, unadjusted P-value) and rs1044396 (G; Z = 2.398, P = 0.017, unadjusted P-value) was associated with increased risk of ND symptoms. These data provide suggestive evidence that variation in the α4 subunit of the nicotinic acetylcholine receptor may influence ND liability.
Smoking is the leading cause of self-induced morbidity and premature mortality, and 12.8% of the U.S. adult population meets criteria for nicotine dependence (ND; Grant et al. 2004). In the brain, nicotinic acetylcholine receptors (nAChRs) are the primary target for nicotine. Nicotinic acetylcholine receptors are ligand-gated ion channels that respond to the endogenous neurotransmitter acetylcholine as well as nicotine (Arias et al. 2006; McGehee 1999).
The most abundant nAChR found in the brain is the high-affinity α4β2* receptor (where * indicates that it can combine with other subunits; Picciotto et al. 1995). These receptors have been implicated in modulating the behavioral effects of nicotine in animal models. Pharmacological blockade of α4β2 nAChRs modulates nicotine self-administration and conditioned place aversion (Corrigall et al. 1994; Jackson et al. 2009; Liu et al. 2007). Furthermore, mice lacking the β2 subunit fail to self-administer nicotine (Picciotto et al. 1998; Pons et al. 2008) and do not exhibit nicotine-induced conditioned place preference (Walters et al. 2006). Moreover, use of genetically modified animals has shown that the α4 subunit modulates nicotine-induced hypothermia, reward, sensitization and tolerance (Pons et al. 2008; Tapper et al. 2004, 2007). These studies provide evidence for the involvement of α4β2* receptors in nicotine behaviors.
Given this literature, early human genetic studies of the genes coding for nAChR subunits focused on CHRNA4 and CHRNB2. For CHRNB2, results of association studies have been mixed. Early studies that were limited in sample size did not find an association between CHRNB2 variants and ND (Etter et al. 2009; Feng et al. 2004a; Li et al. 2005; Lueders et al. 2002; Silverman et al. 2000). However, more recent work examining other nicotine-related phenotypes (such as smoking initiation, subjective effects and abstinence rates) has detected evidence for association (Conti et al. 2008; Ehringer et al. 2007; Greenbaum et al. 2006; Hoft et al. 2011; King et al. 2012; Perkins et al. 2009). Furthermore, common and rare genetic variants in CHRNB2 have been associated with ND in individuals seeking smoking cessation treatment (Wessel et al. 2010).
In contrast to the human genetics studies examining an association between CHRNB2 and ND, the work with CHRNA4 has been more consistent. Feng et al. initially reported an association between rs1044396 and rs1044397 and ND in Chinese males (Feng et al. 2004a). Since this report, additional studies have reported associations between ND and CHRNA4 variation (Breitling et al. 2009; Chu et al. 2011; Han et al. 2011; Li et al. 2005; Saccone et al. 2009, 2010; Xie et al. 2011).
The goal of this study was to investigate the role of CHRNA4 and CHRNB2 in lifetime DSM-IV ND symptom counts in two large young adult samples ascertained through clinical referrals for drug treatment. These samples are unique because subjects were recruited as young adults and have been assessed longitudinally. Single nucleotide polymorphisms (SNPs) in each gene were selected to capture most common genetic variation and Family-Based Association Tests (FBAT) were employed to test for association with DSM-IV ND symptoms.
- Top of page
- Materials and methods
- Supporting Information
In this study, we investigated the role of genetic variation in the genes that code for the α4 and β2 nAChR subunits in a quantitative measure of lifetime DSM-IV ND symptoms in two separate samples. Six and eight SNPs in each gene (CHRNB2 and CHRNA4, respectively) were examined in the CADD sample using methods that account for LD between SNPs. Modest evidence for association with two SNPs was detected in CHRNA4. In CHRNA4, the minor alleles of rs1044394 and rs1044396 were associated with increased risk for ND symptoms (P = 0.010 and 0.027, respectively). This association was strongest in the non-Hispanic EA portion of the sample, and although the number of informative families increased with the addition of the Hispanic population, this did not provide additional support for an association. However, when these SNPs were examined in the GADD sample, the associations did not replicate.
Our findings are consistent with early human genetics studies that provided no evidence of an association with variation in CHRNB2 and ND (Etter et al. 2009; Feng et al. 2004a; Li et al. 2005; Lueders et al. 2002; Silverman et al. 2000). Two distinct genetic loci in CHRNB2 have recently been associated with nicotine behaviors in multiple studies. Two highly correlated SNPs in the 3′ UTR region rs2072660/rs2072661 have been associated with smoking initiation (Greenbaum et al. 2006), smoking cessation (Conti et al. 2008; Perkins et al. 2009) and ND (Wessel et al. 2010). We attempted to genotype rs2072660 in this study, but the error rate in our sample was 1.3%; thus, we did not include this SNP in our analyses.
Similarly, SNP rs2072658 in the 5′ UTR of CHRNB2 has been associated with the subjective effects reported when subjects were asked to retrospectively recall how nicotine made them feel shortly after taking the drug (Ehringer et al. 2007) and reported subjective responses to cigarettes when smoked in a controlled laboratory setting (Hoft et al. 2011). However, there was no evidence for association with rs2072658 in this study of ND symptoms.
The two CHRNA4 SNPs implicated in this study with ND symptoms have been implicated previously in smoking-related phenotypes. The SNP rs1044396 has been nominally associated with nicotine addiction, ND, heavy and regular smoking vs. control, and smoking quantity (Breitling et al. 2009; Feng et al. 2004b; Li et al. 2005), although a recent study found no association between this SNP and smoking status (Tsai et al. 2012). Together with the data presented here, there are now three studies showing evidence that the minor allele of rs1044396 increases the risk of ND (Breitling et al. 2009; Feng et al. 2004b).
For rs1044394, one other study has provided evidence of an association with ND as measured both by DSM-IV ND criteria and the Fagerström Test for ND (Han et al. 2011). In contrast to our results where the minor allele was associated with increased risk of ND symptoms, Han et al. found evidence that the minor allele lowered the risk of ND (Han et al. 2011). Differences do exist between these studies that could account for the difference in effect. The Han et al. study examined both African Americans and EA, whereas we examined non-Hispanic EA and self-identified Hispanics. The association with rs1044394 and ND was driven by the African American sample in their study (although there was suggestive evidence in the EA sample). It is also possible that the differences observed between these studies are due to differences in rare functional variants associated with rs1044294 that differ between our samples. There is evidence that CHRN SNPs show opposite allelic effects with different drugs (Grucza et al. 2008; Sherva et al. 2010; Wang et al. 2008) and in some cases the major allele in EA samples is the minor allele in African American samples, so this is another factor that may contribute to directional differences in effect. Both rs1044394 and rs1044396 are synonymous mutations located in the fifth exon of CHRNA4. However, possible functional effects of these alleles remain unknown.
Our clinical subjects form an extreme sample in terms of drug abuse in line with the ascertainment criteria (having at least one drug dependence symptom, excluding ND criteria). Over 70% of the CADD clinical subjects have at least one substance use disorder other than ND. The GADD sample is similar in that 67% of the sample has one or more substance use disorders. We did not control for the use of other substances in our study, and given the extreme nature of this sample, this may limit the generalizability of these results. Additionally, in this analysis, we have included subjects who have not tried tobacco products. While many groups limit their samples to subjects meeting a certain use threshold (e.g. Saccone et al. 2009, 2010) others have found associations with the CHRNA4 gene among samples including individuals who have never smoked (Han et al. 2011). This is an important limitation to this study and it is possible that this association is driven by tobacco use rather than dependence specifically, but unfortunately we do not have enough subjects to test this hypothesis.
Our results provide evidence that common genetic variation in CHRNA4 is associated with ND. Rare genetic variation in this region has also been associated with this trait (Wessel et al. 2010; Xie et al. 2011). Together, these results show that both common and rare variations in this region of CHRNA4 are likely to influence ND.
Although we found evidence of an association between SNPs in CHRNA4 and ND symptoms in the CADD sample, our results with non-Hispanic EA were stronger than when Hispanics were also included in the analysis. We may have observed this pattern because in our samples the non-Hispanic EA show more severe ND compared with the Hispanics (see Table 1). In our samples, the EA have a greater number of DSM-IV ND criteria as well as a greater number of substance use disorders compared with the corresponding Hispanic sample.
Furthermore, we were unable to replicate our results in the GADD sample, and for rs1044394 the ‘risk’ allele was different. We chose the GADD as a replication sample because it was ascertained in a manner similar to the CADD with recruitment being through drug abuse treatment centers. Overall, the samples were phenotypically similar (see Table 1). However, the entire CADD sample was ascertained in the Denver metro area, whereas the GADD sample was recruited in both Denver and San Diego. Furthermore, there were fewer informative families in the GADD analysis compared with the CADD; thus, there was less power to detect an association in this sample (see Table 5). In the GADD non-Hispanic EA sample there were only 14–43 informative families per SNP (compared with 32–116 for the same SNPs in CADD). This is quite a small number of informative families and would only provide limited power to detect an association. Given this limitation, future studies that have greater power to detect an association are warranted.
Taken together, results from this study follow the pattern of complexity typically observed for psychiatric genetics, which is illustrative of how both genetic diversity (ethnic differences) and phenotypic heterogeneity (comorbidity with other disorders) make it difficult to interpret non-replications. In the CADD sample, we found suggestive evidence of associations between genetic variation in CHRNA4 and lifetime DSM-IV ND symptoms, but this result must be interpreted with caution. As the results did not replicate in the GADD, which is phenotypically quite similar, it is not clear whether this represents a true non-replication or if it is due to reduced sample size and statistical power. It is encouraging that the nominal associations reported here are consistent and replicate prior results with CHRNA4 in separate samples. In conclusion, this work adds to the emerging literature exploring the role of these genes in smoking-related behaviors, suggesting that the common α4β2* nicotinic receptors remain an important possible target for pharmacogenetic therapies.