The ratio of 3HC to cotinine (3HC/COT; also known as the nicotine metabolite ratio (NMR)) is a phenotypic biomarker of CYP2A6 enzymatic activity and, like CYP2A6 genotype, is associated with smoking behaviors. Before discussing the impact of CYP2A6 genetic variation on smoking phenotypes, it is useful to understand the utility and characteristics of this biomarker when investigating genetic associations. The utility of NMR as a biomarker of CYP2A6 activity among smokers derives from the exclusive metabolism of COT to 3HC by CYP2A6, as well as the relatively long half-life of COT (~16 h) and the formation dependence of 3HC (Benowitz et al. 2009; Benowitz and Jacob 2001; Nakajima et al. 1996). The long half-life of COT promotes stability of the relative levels of both COT and 3HC in regular smokers, irrespective of the level of cigarette consumption, making this a useful biomarker for assessing the rate of CYP2A6 activity and nicotine metabolism in both heavy and light smokers. In addition, the NMR is stable over time in regular smokers (Lea et al. 2006; Mooney et al. 2008; St Helen et al. 2012, 2013). NMR is highly correlated with the rate of nicotine clearance, due to the major role of CYP2A6-mediated inactivation of nicotine in nicotine clearance (Dempsey et al. 2004; Levi et al. 2007). NMR is also strongly associated with CYP2A6 genotype, as decrease- and loss-of-function CYP2A6 variants result in slower CYP2A6 activity and corresponding decreases in NMR (Malaiyandi et al. 2006b). In addition to its association with CYP2A6 genotype, NMR also captures environmental sources of variability in CYP2A6 activity, and is often used together with CYP2A6 genotype to investigate contributions of nicotine metabolism variation to smoking behaviors.

Blood nicotine concentration is negatively correlated with cigarette craving (Jarvik et al. 2000), and smokers are known to titrate their smoking levels and intensity to maintain desirable nicotine levels in the body (McMorrow and Foxx 1983). Variation in nicotine pharmacokinetics can contribute to variation in smoking acquisition, the level of consumption (cigarettes smoked per day), puff volume, nicotine dependence, cravings and withdrawal, and cessation success between smokers. A significant proportion of variation in these behaviors is captured by variation in CYP2A6 genotype and phenotype via altered nicotine metabolism rates.

The vast majority of smoking initiation occurs in adolescence (Giovino 1999), and CYP2A6 variation influences smoking acquisition during this developmental period. Relative to faster CYP2A6 nicotine metabolizers, slower CYP2A6 metabolizers are at increased risk of acquiring nicotine dependence, described below. The mechanism(s) underpinning this elevated risk is not known, but may relate to longer durations of action of nicotine in the CNS among those with slower nicotine clearance, and/or genotype group differences in initial smoking experiences. A longer duration of action of nicotine in the CNS may in turn lead to prolongation of the reinforcing properties of nicotine, thus increasing the risk for the development of dependence. Adolescents with one or two copies of the loss-of-function CYP2A6 alleles *2 and *4, classified as slow nicotine metabolizers (Table 1), are at significantly greater risk of developing nicotine dependence defined by the International Classification of Diseases (ICD-10) and the modified Fagerström Tolerance Questionnaire (mFTQ) after initiating smoking compared to normal metabolizers (Al Koudsi and Tyndale 2010; Al Koudsi 2010; Karp et al. 2006; O’Loughlin et al. 2004). Although initial smoking experiences have been shown to be predictive of nicotine dependence (Pomerleau et al. 1998), the proportion of CYP2A6 faster and slower metabolizers experiencing early smoking symptoms (e.g., rush, buzz, dizziness, nausea, relaxation) appears similar among adolescents (Audrain-McGovern et al. 2007; O’Loughlin et al. 2004). Whether the intensity of early smoking symptoms differs according to CYP2A6 genotype group remains to be determined and, if differences exist, may help to explain these findings.

CYP2A6 also influences the progression in nicotine dependence among adolescent smokers. Once dependent, those smokers with slower CYP2A6 activity demonstrate slower progression in nicotine dependence scores (on mFTQ) over time, relative to those with faster CYP2A6 activity (Audrain-McGovern et al. 2007). Thus, while their initial risk for nicotine dependence is greater, those with slower nicotine metabolism do not appear to escalate in nicotine dependence as rapidly as faster metabolizers (Audrain-McGovern et al. 2007). When looking at a single measure of nicotine dependence, rather than progression in nicotine dependence, a cross-sectional study demonstrated that adolescent smokers with slower nicotine metabolism (lower NMR) displayed greater mFTQ dependence relative to those with higher NMR (Rubinstein et al. 2013). In contrast, a separate study showed a trend toward higher scores on the Hooked on Nicotine Checklist (HONC), a measure of nicotine dependence, among adolescent light smokers (smoking 1–6 cigarettes per day; CPD) with higher NMR (Rubinstein et al. 2008). The different patterns of dependence scores across NMR groups in the separate studies may reflect, in part, that the specific study populations were at different stages of smoking onset, as well as differences in the aspects of dependence captured by the two scales.

To achieve desirable levels of nicotine in the brain and bloodstream, smokers regulate the number of cigarettes smoked per day, as well as the intensity (i.e., puff volume) with which they smoke each cigarette (McMorrow and Foxx 1983). Several biomarkers of smoking consumption have been established, including the levels of exhaled carbon monoxide (CO), plasma and urinary cotinine, and urinary total nicotine equivalents (TNE) (Hartz et al. 2012). Slower CYP2A6 nicotine metabolizers smoke fewer CPD than faster CYP2A6 metabolizers, as indicated by lower breath CO, TNE, and/or depth of inhalation (Ariyoshi et al. 2002; Benowitz et al. 2002; Liu et al. 2011; Malaiyandi et al. 2006b; Rao et al. 2000; Schoedel et al. 2004; Zhu et al. 2013a). The lower levels of smoking exhibited by slower nicotine metabolizers likely stems from their slower elimination of nicotine. It appears, however, that the titration of nicotine levels may be more pronounced in nicotine-dependent smokers as Schoedel et al. (2004), in adults, showed that slow metabolizers of European descent smoke fewer CPD only when they meet criteria for tobacco dependence.

In a separate study of smokers of European ancestry, slower CYP2A6 metabolizers smoked fewer cigarettes while maintaining a level of smoking intensity (determined by the CO to cotinine ratio) similar to that of individuals with a wild-type (CYP2A6*1/*1) genotype (Rao et al. 2000). Individuals with the CYP2A6 gene duplication (associated with an increased rate of nicotine metabolism, Table 1) smoked a similar number of CPD as individuals with the CYP2A6*1/*1 genotype, but displayed increased smoking intensity (higher CO to cigarette and nicotine to cigarette ratios) (Rao et al. 2000). Different titration patterns are also seen among adolescent smokers. Although adolescents who are slower CYP2A6 nicotine metabolizers are faster to acquire nicotine dependence (O’Loughlin et al. 2004), dependent slower CYP2A6 metabolizers smoke fewer CPD than dependent faster CYP2A6 metabolizers (Al Koudsi 2010; Audrain-McGovern et al. 2007; O’Loughlin et al. 2004). Similar findings were obtained in studies of Chinese and Japanese adult smokers (Ariyoshi et al. 2002; Benowitz et al. 2002; Liu et al. 2011). Chinese and Japanese smokers living in their native countries smoke relatively high quantities of cigarettes overall, but a similar impact of CYP2A6 variability on CPD remains apparent. For example, Japanese slow metabolizers smoke ~15 CPD compared to ~20 CPD smoked by intermediate metabolizers, and ~25–30 CPD smoked by normal metabolizers (Ariyoshi et al. 2002; Fujieda et al. 2004). In addition to the association of CYP2A6 genotype with cigarette consumption, variation in NMR also influences daily tobacco consumption, with lower NMR, indicative of slower nicotine metabolism, corresponding to lower levels of smoking (Benowitz et al. 2003). This has been demonstrated in Europeans, African Americans, and Alaska Natives (Benowitz et al. 2003; Mwenifumbo et al. 2007; Zhu et al. 2013a).

In contrast to smokers of European descent, African American and Alaska Native smokers (predominantly reporting smoking less than 10 CPD) and smokeless tobacco users do not display CYP2A6 genotype associations with reported levels of use (CPD, chews per day) (Ho et al. 2009a; Zhu et al. 2013a). For example, the number of cigarettes smoked per day by these two populations does not differ between slow and normal nicotine metabolizers (Ho et al. 2009a; Zhu et al. 2013a). However, when using a more robust measure of consumption, levels of urinary TNE, light smokers, and smokeless tobacco users who are slow metabolizers have lower levels of tobacco consumption (Zhu et al. 2013a). Light smokers with low NMR may achieve lower doses of tobacco by smoking less intensely. This suggests that self-reported CPD may be a relatively poor biomarker of total tobacco exposure, especially among lighter smokers.

Cotinine, the primary metabolite of nicotine, has been used as a biomarker of tobacco exposure in a large number of studies, including those focused on CYP2A6 genetic influences on smoking . In heavy smokers, plasma cotinine is correlated with tobacco consumption (Benowitz et al. 2011). However, cotinine levels can overestimate tobacco exposure in CYP2A6 slower metabolizers because slower CYP2A6 activity influences cotinine clearance more than cotinine formation (Zhu et al. 2013b). In support of this, slow metabolizers exhibit greater plasma cotinine levels than normal metabolizers, despite smoking fewer cigarettes and having lower TNE (Ho et al. 2009a; Rao et al. 2000; Zhu et al. 2013a, b). The higher plasma cotinine levels in slower versus normal metabolizers does not accurately reflect their differences in smoking. Therefore, other biomarkers should be used to determine the level of tobacco consumption, especially when contrasting groups with differing CYP2A6 activity, such as those with differing genotypes (e.g., ethnic groups with different CYP2A6 genotype frequencies) and those of different gender (females are faster nicotine metabolizers than males) (Zhu et al. 2013b).

The measurement of TNE, as opposed to plasma cotinine, self-reported CPD, and breath CO, is likely the most accurate biomarker of nicotine exposure available thus far. Urinary TNE is advantageous over these other markers of dose because TNE is associated with total nicotine exposure (Benowitz et al. 2010; Scherer et al. 2007), and accounts for multiple metabolites of nicotine (~90 % of nicotine dose) (Benowitz et al. 1994), thus making it a measure of dose which is independent of variation in these metabolic pathways. CYP2A6 genotype is associated with differences in TNE, even among light smokers, whereas the other relatively weaker markers of smoking are not (Zhu et al. 2013a).

Several assessments are available to evaluate nicotine dependence in smokers. The Fagerström Test of Nicotine Dependence (FTND) (Heatherton et al. 1991), HONC, and DSM-IV criteria are common examples of scales that measure the level of nicotine dependence. CYP2A6 genotype or NMR variability have not consistently been shown to be associated with nicotine dependence, with one assessment measure in particular yielding differing results—FTND. Several studies have shown higher FTND scores, indicative of greater nicotine dependence, in faster nicotine metabolizers compared to slower nicotine metabolizers when characterized by CYP2A6 genotype (Kubota et al. 2006; Wassenaar et al. 2011) and NMR (Sofuoglu et al. 2012). In contrast, other studies do not show a significant association between the rate of nicotine metabolism and the level of dependence (Benowitz et al. 2009; Ho et al. 2009b; Johnstone et al. 2006; Malaiyandi et al. 2006a; Schnoll et al. 2009b). When measuring dependence in those with lower cigarette consumption, however, HONC may be a more appropriate measure to ascertain nicotine dependence (Wellman et al. 2006).

Potential explanations for the different findings with FTND include the heavy weighting toward CPD and the lack of assessment of withdrawal or craving, which can be influenced by the rate of nicotine metabolism (Jarvik et al. 2000; Kubota et al. 2006). Another potential explanation is the possibility of increased selection bias in clinical smoking cessation trials compared to observational studies. In clinical trials, one may be less likely to see differences in dependence between slow and normal metabolizers as this particular group of slow metabolizers may differ as they are smokers who have not been successful in quitting smoking . This could manifest as greater dependence than would be observed in a random sample of smokers who are not seeking help for cessation. Discordant findings involving FTND may also result from sampling a mixed gender population, as no association between NMR and FTND was seen in females (Schnoll et al. 2014). In contrast, male smokers with higher NMR have higher FTND scores than males with lower NMR (Schnoll et al. 2014). The lack of association of NMR and nicotine dependence in females suggests that non-nicotinic factors may have a relatively stronger influence on dependence in females than in males (Schnoll et al. 2014).

Several studies have directly assessed relationships between CYP2A6 genotype and NMR with cessation, where slower nicotine metabolism, as measured by genotype or phenotype, predicts increased quit rates in adult (Chen et al. 2014; Gu et al. 2000; Ho et al. 2009b; Lerman et al. 2006; Patterson et al. 2008; Schnoll et al. 2009) and adolescent (Chenoweth et al. 2013) smokers. Variation in CYP2A6 activity can influence spontaneous smoking cessation success as well as response to smoking cessation pharmacotherapies.

Relatively higher quit rates observed among slower nicotine metabolizers compared to faster nicotine metabolizers may be due in part to their less severe withdrawal symptoms (Kubota et al. 2006), lower numbers of cigarettes smoked (resulting in fewer learning trials and possibly less ingrained smoking behaviors) (Ariyoshi et al. 2002; Audrain-McGovern et al. 2007; Benowitz et al. 2002; Liu et al. 2011; O’Loughlin et al. 2004; Rao et al. 2000), as well as reduced brain response to smoking cues as demonstrated through functional magnetic resonance imaging (Tang et al. 2012). Relative to faster nicotine metabolizers, slower nicotine metabolizers likely experience lower fluctuations in blood nicotine concentrations throughout the day. These latter imaging findings regarding decreased response to smoking cues among slower CYP2A6 metabolizers suggest slower metabolizers likely form weaker conditioned responses to smoking cues, perhaps due to relatively lower surges in brain nicotine concentration, and resulting dopamine levels, achieved during ad lib smoking (Tang et al. 2012).

Slower nicotine metabolizers have higher spontaneous quit rates compared to normal nicotine metabolizers (Gu et al. 2000) as well as increased quitting on certain smoking cessation pharmacotherapies. Cessation trials involving nicotine replacement therapies (NRTs) and bupropion indicate that the rate of nicotine metabolism can predict cessation success (retrospective analyses). NRTs include nicotine gum, patch, nasal spray, and lozenge, which act to replace nicotine that would normally be consumed from cigarettes, with the intention of helping to reduce a smoker’s cravings and likelihood to smoke. Smokers with lower NMR display greater cessation success when treated with transdermal nicotine (nicotine patch) compared to faster metabolizers, but the same effect is not seen when using nicotine nasal spray (Lerman et al. 2006; Schnoll et al. 2009). The lack of effect observed for subjects using nasal spray is likely due to the smokers’ differential titration of the dose, as usage in this trial differed by their rate of nicotine metabolism (Malaiyandi et al. 2006a). In contrast, Chen et al. (2014) showed that NRT increases cessation success compared to placebo in fast, but not slow, nicotine metabolizers as defined by CYP2A6 genotype. When treated with bupropion, a dopamine and norepinephrine reuptake inhibitor and weak nAChR antagonist (Warner and Shoaib 2005), which is not metabolized by CYP2A6 , there was no difference in quit rates by NMR in those receiving bupropion therapy. In contrast, in the placebo arm, slow nicotine metabolizers experienced higher quit rates relative to faster nicotine metabolizers (Patterson et al. 2008). A more recent study, however, demonstrated that bupropion was effective in prolonging abstinence in both slower and faster nicotine metabolizers relative to placebo, suggesting no effect of CYP2A6 genotype on response to bupropion (Chen et al. 2014). Varenicline, a newer pharmacotherapy for smoking cessation that acts as a partial agonist at α4β2 nAChRs , competes with nicotine for α4β2 binding sites, blocking nicotine -evoked dopamine release (Garrison and Dugan 2009). King et al. (2012) assessed whether certain genes play an important role in cessation success with this drug as well as with bupropion. Continuous abstinence when taking varenicline was not associated with CYP2A6 genotype.

The associations between CYP2A6 and NMR and smoking cessation outcomes were elucidated from retrospective analyses of smoking cessation clinical trial data. The utility of NMR as a predictive biomarker of smoking cessation outcomes is being prospectively investigated in a phase III clinical trial (NCT0131001) involving varenicline, nicotine patch, and placebo. Randomization to treatment group was stratified prospectively according to participant NMR. Varenicline, compared to nicotine patch, was associated with greater quitting among normal nicotine metabolizers, whereas for slow metabolizers patch worked well and had fewer side effects than Varenicline for slow metabolizers (Lerman et al. 2015).

In addition to interindividual differences in smoking behavior within ethnicities, smoking patterns also vary between ethnicities. The variable smoking patterns between different ethnic groups are associated with differences in the rate of nicotine metabolism, which are consistent with population differences in CYP2A6 allele frequencies. Multiple studies have shown that CYP2A6 allele frequencies vary significantly between European, African American, Asian, and Alaska Native populations (Table 1). In ethnic groups that have a higher frequency of CYP2A6 decrease- and loss-of-function variants, a lower overall NMR and slower rate of nicotine metabolism is observed. For example, relative to those of European descent, African Americans have a higher frequency of several CYP2A6 alleles that result in a decrease or loss of CYP2A6 enzyme function (Al Koudsi et al. 2009; Fukami et al. 2005; Mwenifumbo et al. 2008, 2010; Nakajima et al. 2006). Consistent with the higher frequency of decrease- and loss-of-function of CYP2A6 alleles in African Americans, African American adolescent smokers display slower nicotine metabolism, as measured by NMR, relative to adolescent smokers of European ancestry (Rubinstein et al. 2013).

Similarly, Asian smokers metabolize nicotine more slowly than European smokers, due in part to the higher frequency of specific CYP2A6 slower metabolism variants in Asian populations, relative to Europeans (Nakajima et al. 2006; Schoedel et al. 2004). Japanese subjects exhibited significantly lower cotinine/nicotine ratios compared to Europeans (Nakajima et al. 2006), and Chinese-American smokers had significantly lower nonrenal clearance of nicotine and cotinine compared to European smokers, which was associated with lower nicotine intake from smoking in Chinese-Americans than Europeans (Benowitz et al. 2002). Consistent with these findings, Asian adolescent smokers also possess a lower average population NMR, relative to adolescent smokers of European descent (Rubinstein et al. 2013).

Unlike African American and Asian populations, Alaska Natives metabolize nicotine more rapidly than Europeans as shown by a greater NMR in the overall population and among the wild-type (CYP2A6*1/*1) subgroup (Binnington et al. 2012). This is consistent with higher levels of nicotine intake during tobacco use in Alaska Native smokers compared to the general United States population (Benowitz et al. 2012). The observed higher NMR among Alaska Natives does not correspond to a lower frequency of CYP2A6 decrease- and loss-of-function variants compared to Europeans; the reason(s) for this higher NMR has not yet been determined, but may relate to uncharacterized genetic variation or to dietary inducers.

Similar to CYP2A6, CYP2B6 is involved in the metabolism of nicotine (Yamazaki et al. 1999), however its role in hepatic nicotine metabolism to cotinine is minor (~10 %) relative to CYP2A6 (~90 %) (Al Koudsi and Tyndale 2010). Similar to CYP2A6, the CYP2B6 gene is highly polymorphic (variants characterized to date found at http://​www.​cypalleles.​ki.​se/​cyp2b6.​htm). However, consistent with its minor role in hepatic nicotine metabolism there is no detectable impact of CYP2B6 gene variants on the modulation of peripheral nicotine metabolism (Lee et al. 2007b). Conversely, when CYP2A6 expression or activity is reduced, CYP2B6 may play a relatively more prominent role in nicotine metabolism, and CYP2B6 genetic variation may represent a source of variation in nicotine metabolism in those with reduced CYP2A6 activity (Ring et al. 2007). CYP2B6 is expressed in the liver, but it is also expressed in the brains of nonhuman primates and humans (Ferguson et al. 2013; Miksys et al. 2003), thus potentially modulating CNS nicotine metabolism and the duration of action of nicotine in the brain.

The impact of CYP2B6 variation on nicotine metabolism appears negligible (Al Koudsi and Tyndale 2010; Lee et al. 2007b), suggesting that the role of CYP2B6 in the modulation of smoking behaviors is also minor. However, since CYP2B6 metabolizes the smoking cessation drug bupropion (Faucette et al. 2000), the association between variation in CYP2B6 and cessation success has been investigated. There is inconsistency regarding whether slower CYP2B6 activity promotes smoking cessation when taking bupropion. The unpredictability of treatment outcome based on CYP2B6 genotype may stem from the fact that both bupropion and the CYP2B6-mediated metabolite hydroxybupropion are therapeutically active (Carroll et al. 2014; Damaj et al. 2004; Zhu et al. 2012).

In a placebo-controlled smoking cessation clinical trial involving bupropion, subjects taking placebo who possessed the CYP2B6*5 variant, resulting in decreased CYP2B6 expression (Lang et al. 2001), reported more cravings post-quit and lower quit rates compared to those without the CYP2B6*5 allele (Lerman et al. 2002). In the bupropion arm, male smokers possessing the CYP2B6*5 variant also exhibited decreased abstinence, but the association with CYP2B6 genotype was not observed in females. However, interpretation of the CYP2B6*5 data is unclear as the impact of this variant in vivo is not well established (Burger et al. 2006; Kirchheiner et al. 2003; Wyen et al. 2008). If this allele does reduce activity, the slower bupropion metabolism may result in decreased cessation success, potentially due to lower levels of bupropion’s pharmacologically active metabolite hydroxybupropion, which similarly inhibits dopamine and norepinephrine transporters (Carroll et al. 2014; Damaj et al. 2004). In support of this, smokers with higher levels of hydroxybupropion, suggestive of greater CYP2B6 activity, have a greater likelihood of being abstinent than smokers with low levels of hydroxybupropion (Zhu et al. 2012). It is worth noting several things about this study. The first is that only about 60 % of the participants were actually taking bupropion at week 3—thus without biomarkers it is difficult to assess adherence and thus the effect of a gene on the drug effect. Also, while CYP2B6 genotype altered hydroxybupropion levels, there was no direct association between CYP2B6 genotype and cessation rates, suggesting that the genotype effects are not large enough to alter smoking cessation, that the study was insufficiently powered for this comparison, or that other undetected alleles or environmental influences on CYP2B6 may have muted the association. A separate placebo-controlled bupropion clinical trial found that smokers with one or two copies of the CYP2B6*6 haplotype, which is associated with decreased bupropion metabolism (Zhu et al. 2012), had similar rates of abstinence on bupropion (Lee et al. 2007a). Conversely, King et al. (2012) showed an association between continuous smoking abstinence at weeks 9–12 of bupropion treatment and through nondrug follow-up from weeks 12–52 and several CYP2B6 SNPs. However, these SNPs are not yet functionally described, and a direct impact on treatment outcome could not be determined. Thus it remains to be clarified whether CYP2B6 genetic variation has a substantial impact on quitting, in both placebo and bupropion treatment arms.

Lee et al. (2007b) assessed the association between CYP2B6*6 and abstinence rates when subjects were administered different forms of NRT (nicotine patch and nasal spray), and found no difference in levels of nicotine , or abstinence on nicotine patch or nasal spray, between wild-type and CYP2B6*6 groups. This suggests a negligible effect of CYP2B6 on nicotine kinetics or cessation using NRTs.

CYP2A13 is also capable of metabolizing nicotine to cotinine in humans (Bao et al. 2005). However, the role that CYP2A13 plays in nicotine clearance is relatively minor as its low levels and extrahepatic localization to the nasal mucosa, lung, and trachea limits its overall contribution to systemic nicotine metabolism (Su et al. 2000). Although CYP2A13 genetic variation does not substantially affect the rate of nicotine metabolism, CYP2A13 variants are implicated in lung cancer risk in smokers, likely via CYP2A13-mediated metabolic activation of the tobacco-specific nitrosamine NNK (Su et al. 2000). CYP2A13 has the highest affinity for NNK of all P450 enzymes investigated (Jalas et al. 2005) and is located within the lung where it may mediate this activation. Subjects having one or two copies of a CYP2A13 variant that result in markedly reduced metabolic activity toward NNK are associated with a lower risk of developing lung adenocarcinoma (Wang et al. 2003).

FMOs are involved in the metabolism of nicotine , although the relative contribution depends on the subtype. FMO3, the most common form of FMO present in human liver (Hines et al. 1994), metabolically inactivates nicotine to nicotine N′-oxide (Cashman et al. 1992). FMO1 is predominantly expressed in the human kidneys and potentially in the brain, and has been shown to metabolize nicotine more efficiently than FMO3 in vitro (Hinrichs et al. 2011). However, the extrahepatic expression of FMO1 likely limits its overall impact on nicotine N′-oxidation. The genes encoding both FMO1 and FMO3 are polymorphic, indicating a potential role of FMO variation in interindividual differences in nicotine metabolism.

FMO3 may play a relatively larger role in nicotine metabolism in subjects with reduced CYP2A6 expression or activity compared to those with normal CYP2A6 function (Yamanaka et al. 2004). A larger proportion of an oral nicotine dose is excreted as nicotine N′-oxide in urine among those homozygous for the CYP2A6 gene deletion (*4) (~30 %) relative to those with wild-type CYP2A6 genotype (~5–7 %) (Yamanaka et al. 2004). In addition, the FMO3 SNP rs2266782 (G>A, E158K), a decrease-of-function variant, was associated with slower nicotine metabolism in slower CYP2A6 metabolizers, but not in faster CYP2A6 metabolizers, in African Canadian nonsmokers receiving oral nicotine (Chenoweth et al. 2014). Despite the association between FMO3 rs2266782 and modestly slower nicotine metabolism among those with slower CYP2A6 activity, FMO3 rs2266782 did not alter cigarette consumption or TNE, a measure of daily nicotine intake and smoke exposure, examined among African North American light smokers (<10 CPD) (Chenoweth et al. 2014).

Bloom et al. (2013) investigated several FMO3 haplotypes prevalent in subjects of European descent, which included the following coding SNPs: rs1800822 (C>T), rs2266782 (G>A, E158K, previously discussed), rs1736557 (G>A, V257M), rs909530 (T>C), and rs2266780 (A>G, E308G). Variation in FMO3 haplotype was associated with differences in cigarette consumption among European American heavier smokers (~20–25 CPD), with the FMO3 haplotypes CGGCA and CAGTG corresponding to an increase in consumption of ~3 cigarettes per day, relative to the FMO3 haplotypes CAGCA, CGACA, and TGGTA (Bloom et al. 2013). However, this association was apparent only in those with faster, rather than slower, CYP2A6 activity, and it is unclear which SNPs comprising the haplotypes are causal (Bloom et al. 2013).

The role of FMO1 genetic variation in smoking behaviors has not been widely investigated. In European American smokers, two FMO1 SNPs, rs10912765 and rs7877, were associated with nicotine dependence (Hinrichs et al. 2011). These SNPs are present in the 5′ and 3′ regions of FMO1 and are thought to play a role in FMO1 regulation (Hinrichs et al. 2011). Together these findings suggest that variation in FMO activity may influence the rate of nicotine metabolism. However, it is not yet clear if the changes are sufficiently large to alter smoking behaviors, including cigarette consumption and nicotine dependence.

Nicotine and cotinine are metabolized by UGTs through the process of N-glucuronidation, which results in the excretion of nicotine- and cotinine-glucuronide in the urine. Nicotine-glucuronide and cotinine-glucuronide account for 4 and 17 % of the total nicotine metabolites excreted in the urine, respectively (Benowitz et al. 2009; Benowitz and Jacob 1994; Byrd et al. 1992). Variation in UGT genes may result in altered rates of nicotine metabolism, potentially impacting smoking behaviors.

Interindividual variation in nicotine glucuronidation rates has been observed, with genetic factors contributing to this variation (Benowitz et al. 1999; Lessov-Schlaggar et al. 2009). Multiple UGTs may be responsible for nicotine glucuronidation, in particular UGT1A4 and UGT2B10. Variation in UGT1A4 alters the glucuronidation of non-nicotinic substrates, but potential effects on nicotine metabolism and resulting changes in smoking behavior appear to be negligible (Kaivosaari et al. 2007). Two SNPs in UGT1A4 (rs6755571 C>A, P24T and rs2011425 T>G, L48V), known as UGT1A4*2 and UGT1A4*3, respectively, exhibit decreased glucuronidation activity toward the steroid dihydrotestosterone. UGT1A4*3 is also associated with increased glucuronidation efficiency (V _max/K _m ) for the substrate clozapine (Ehmer et al. 2004; Mori et al. 2005). Considering that UGT1A4 does not appear to play a significant role in hepatic nicotine glucuronidation (Kaivosaari et al. 2007), it is unlikely that UGT1A4 variants would alter smoking behaviors, although this requires confirmation.

A SNP occurring in the UGT2B10 gene (rs61750900 G>T, D67R), referred to as UGT2B10*2, exhibits significantly reduced N-glucuronidation of nicotine and cotinine (Chen et al. 2007, 2010). One study has found that individuals with the UGT2B10*2 allele consume less nicotine , exemplified by lower urinary TNE relative to those with the UGT2B10*1/*1 genotype (Berg et al. 2010), while another did not (Zhu et al. 2013c). This suggests that the UGT2B10*2 allele may lower smoking levels in some circumstances or populations.

Renal elimination is responsible for approximately 5 % of total nicotine clearance (Benowitz et al. 2009; Benowitz and Jacob 1985). A twin study provides evidence of substantial genetic influence (approximately 40 % for nicotine , 60 % for cotinine) on variation in nicotine and cotinine renal clearance (Benowitz et al. 2008). The gene(s) contributing to the observed genetic variation remain to be characterized. Organic cation transporters (OCTs), which have been shown to transport nicotine in a human carcinoma cell line (Takami et al. 1998; Urakami et al. 2002; Zevin et al. 1998), may play a role in nicotine elimination through active transport of nicotine across kidney cells. The gene encoding OCT2 is polymorphic, and OCT2 variation alters the renal clearance of metformin, a substrate of OCT2 (Yoon et al. 2013). Increased renal clearance of nicotine in smokers, achieved through urine acidification (results in ionic trapping of nicotine, a weak base), was previously associated with an increase in daily nicotine intake (Benowitz and Jacob 1985). Thus, if a role for OCT2 in nicotine transport in vivo is elucidated, OCT2 variation may influence the renal clearance of nicotine, and lead to compensatory changes in smoking behaviors. OCT2 is also expressed at the blood–brain barrier; however, it is not clear if variation in OCT2 at this site would be sufficient to alter smoking behaviors. Thus genetic variation in nicotine transport systems, while a very minor pathway for nicotine clearance, may represent another source of variation in smoking behaviors.

In addition to variation in the rate of nicotine clearance, alteration of nicotine’s CNS targets and downstream signaling pathways can modify nicotine’s psychoactive effects. As previously mentioned, the binding of nicotine to neuronal nAChRs (Liu et al. 2012) leads to a cascade of downstream signaling events involving the release of dopamine , serotonin , norepinephrine, acetylcholine, γ-aminobutyric acid, glutamate, and endorphins (Benowitz 2008). Dopamine release, which mediates the primary reinforcing effects of nicotine intake, contributes to the development of nicotine dependence (Dani and Heinemann 1996), and variation in the neurobiological pathways that regulate response to nicotine may contribute to differences in multiple smoking behaviors, including the level of smoking and dependence. Polymorphic genes encode nAChR subunits, dopamine transporters, dopamine receptors, and dopamine-metabolizing enzymes, as well as enzymes responsible for serotonin synthesis and transport. The impact of genetic variation in the CNS on smoking behaviors will be described in the following section.