Variation in CYP2A6 and tobacco dependence throughout adolescence and in young adult smokers

Highlights

• In adolescence, CYP2A6 slow nicotine metabolizers have a higher risk of becoming tobacco dependent.

• The role of CYP2B6 variation in smoking acquisition and dependence remains to be clarified.

• The findings highlight the role of genetic risk factors in adolescent tobacco dependence.

Abstract

Background

Smoking is influenced by genetic factors including variation in CYP2A6 and CYP2B6, which encode nicotine-metabolizing enzymes. In early adolescence, CYP2A6 slow nicotine metabolism was associated with higher dependence acquisition, but reduced cigarette consumption. Here we extend this work by examining associations of CYP2A6 and CYP2B6 with tobacco dependence acquisition in a larger sample of smokers followed throughout adolescence.

Methods

White participants from the Nicotine Dependence in Teens cohort that had ever inhaled (n = 421) were followed frequently from age 12–18 years. Cox's proportional hazards models compared the risk of ICD-10 tobacco dependence acquisition (score 3+) for CYP2A6 and CYP2B6 metabolism groups. Early smoking experiences, as well as amount smoked at end of follow-up, was also computed. At age 24 (N = 162), we assessed concordance between self-reported cigarette consumption and salivary cotinine.

Results

In those who initiated inhalation during follow-up, CYP2A6 slow (vs. normal) metabolizers were at greater risk of dependence (hazards ratio (HR) = 2.3; 95% CI = 1.1, 4.8); CYP2B6 slow (vs. normal) metabolizers had non-significantly greater risk (HR = 1.5; 95% CI = 0.8, 2.6). Variation in CYP2A6 or CYP2B6 was not significantly associated with early smoking symptoms or cigarette consumption at end of follow-up. At age 24, neither gene was significantly associated with dependence status. Self-reported consumption was associated with salivary cotinine, a biomarker of tobacco exposure, acquired at age 24 (B = 0.37; P < 0.001).

Conclusions

Our findings extend previous work indicating that slow nicotine metabolism mediated by CYP2A6, and perhaps CYP2B6, increases risk for tobacco dependence throughout adolescence.

Introduction

Approximately 90% of smokers begin smoking in adolescence (O’Loughlin et al., 2014b, U.S. Department of Health and Human Services, 2012). A substantial proportion (∼40–75%) of smoking behaviour is influenced by genetics (Broms et al., 2006, Vink et al., 2005). CYP2A6 inactivates nicotine, the principle psychoactive compound in cigarette smoke, to cotinine (Nakajima et al., 1996). Genetic variation in CYP2A6 that reduces the rate of nicotine metabolism is associated with lower cigarette consumption (Malaiyandi et al., 2006, Wassenaar et al., 2011), dependence scores (Schnoll et al., 2014, Sofuoglu et al., 2012, Wassenaar et al., 2011), brain response to smoking cues (Tang et al., 2012), and greater cessation (Gu et al., 2000, Lerman et al., 2006, Schnoll et al., 2009), even in adolescence (Chenoweth et al., 2013). In adolescents, CYP2A6 slow nicotine metabolism was also associated with an increased risk of tobacco dependence acquisition at young ages (from age 12 to 16 years; Al Koudsi et al., 2010, O’Loughlin et al., 2004), but slower escalation in nicotine dependence (Audrain-McGovern et al., 2007) and reduced cigarette consumption (O’Loughlin et al., 2004). In young adults, CYP2A6 slow (vs. normal) metabolizers were less likely to be smokers (Schoedel et al., 2004). Together these findings suggest that while CYP2A6 slow metabolism increases the risk of becoming a smoker in younger adolescence, slow metabolism also increases cessation, and reduces cigarette consumption in dependent smokers. However, it is not known whether CYP2A6 slow metabolism increases smoking acquisition in later adolescence, a period during which a substantial amount of smoking uptake occurs (O’Loughlin et al., 2014b).

A small proportion (∼10%) of nicotine's metabolism to cotinine occurs via a second enzyme, CYP2B6 (Al Koudsi and Tyndale, 2010). The CYP2B6*6 allele, a prevalent haplotype (∼25% frequency in Whites (Rotger et al., 2007)) is associated with lower CYP2B6 hepatic protein levels (Al Koudsi and Tyndale, 2010) and slower CYP2B6-mediated metabolism of bupropion and efavirenz (reviewed in (Thorn et al., 2010)). In adult smokers, CYP2B6*6 was associated with lower abstinence rates in the placebo arm of a bupropion smoking cessation clinical trial; 15% of individuals with one or two copies of CYP2B6*6 achieved abstinence, compared to 32% of CYP2B6*1/*1 individuals (Lee et al., 2007a). In a separate study, the CYP2B6*6 allele was more frequent in nicotine dependent individuals compared to those that were not dependent (32% vs. 22%, respectively; Riccardi et al., 2015). Whether CYP2B6*6 also influences the risk for acquiring nicotine dependence in adolescence is not known.

Here we examined associations for CYP2A6 and CYP2B6 with tobacco dependence acquisition in a larger (n > 400) sample of adolescent smokers assessed four times each year across the entire adolescent period (age 12–18 years). We hypothesized that CYP2A6 slow (vs. CYP2A6 normal), and that CYP2B6 slow (i.e., individuals with one or two copies of CYP2B6*6) (vs. CYP2B6 normal) metabolizers would be at increased risk of acquiring dependence. We also hypothesized that a larger proportion of slow (vs. normal) metabolizers for each gene would report early smoking experiences, which are associated with the development of nicotine dependence (DiFranza et al., 2004). We also assessed cigarette consumption at the end of follow-up among dependent smokers, hypothesizing that CYP2A6 slow (vs. CYP2A6 normal) metabolizers would smoke fewer cigarettes; no association was expected between CYP2B6 genotype groups. For both CYP2A6 and CYP2B6, we further hypothesized that slow (vs. normal) metabolizers would be more likely to be dependent at end of follow-up. At age 24, we expected CYP2A6 slow (vs. normal) metabolizers to be at lower risk of dependence, as CYP2A6 slow metabolizers are less likely to be dependent smokers (vs. non-smokers) as adults (Schoedel et al., 2004). Finally, we hypothesized that CYP2B6 slow (vs. normal) metabolizers would be more likely to be dependent at age 24, consistent with the higher frequency of CYP2B6*6 in dependent (vs. non-dependent) adults (Riccardi et al., 2015).

Finally, an adjunct biochemical analysis to assess the validity of the self-reported cigarette consumption data was undertaken. We examined the construct-related validity of self-reported cigarette consumption against salivary cotinine, widely used as an objective biomarker of tobacco consumption (Connor Gorber et al., 2009), and also assessed its relationships with nicotine dependence and withdrawal scores.