<<

CYP2A6 and CYP2B6: Sources of variation and their role in

by

Nael Al Koudsi

A thesis submitted in conformity with the requirements for the degree of

Doctor of Philosophy

Graduate Department of Pharmacology and Toxicology

University of Toronto

 Copyright by Nael Al Koudsi 2010

CYP2A6 and CYP2B6: Sources of variation and their role in nicotine metabolism

Nael Al Koudsi Doctor of Philosophy, 2010 Graduate Department of Pharmacology and Toxicology University of Toronto

Abstract

Nicotine is the primary substance in causing . In humans the majority

(70-80%) of nicotine is inactivated to in a reaction predominantly catalyzed by

CYP2A6 (80-90%), with a minor possible role for CYP2B6. Substantial interindividual variability is observed in the rate of nicotine’s inactivation to cotinine and this variation contributes to differences in behaviors (e.g. cigarette consumption). Twin studies suggest an important genetic contribution to the variability in nicotine metabolism. However in

2004, genetic variation in CYP2A6 and CYP2B6 accounted for only a small portion of the variability suggesting gaps in our knowledge. Our objective was to identify additional genetic and non-genetic sources of variability in CYP2A6 expression/activity, CYP2B6 expression, and nicotine to cotinine metabolism in vivo and/or in vitro. Participants included individuals from different world populations phenotyped for CYP2A6 activity either following oral nicotine administration or using metabolite ratios derived from baseline smoking. Genotyping and sequencing were utilized to identify and characterize multiple new CYP2A6 alleles. In total 17 novel CYP2A6 alleles were identified, many of which were found predominantly among individuals of black African descent and exhibited lower CYP2A6 activity. In addition, human livers were assessed for CYP2A6 and CYP2B6 expression and nicotine to cotinine metabolism.

The mechanisms underlying the lower CYP2A6 activity associated with some of the variant

CYP2A6 alleles included either a reduction in hepatic CYP2A6 expression, an alteration of CYP2A6’s structural property, or a combination of both. DNA methylation was not ii associated with altered hepatic CYP2A6 expression/activity. Livers from female donors were associated with higher CYP2A6 and CYP2B6 protein expression compared to male livers, while age did not influence the expression of either CYP. Finally, CYP2B6 and its prevalent altered function genetic variant (CYP2B6*6) did not influence nicotine to cotinine metabolism.

Identification of factors that contribute to the variability in CYP2A6 and nicotine metabolism is important to improve future association studies between CYP2A6 genotype, nicotine metabolism, and smoking behaviors. In addition, this information could provide the potential to personalize therapy in order to improve the clinical efficacy of nicotine, particularly as a smoking cessation aid.

iii Acknowledgments

My sincerest gratitude goes to Dr. Rachel Tyndale for her continuous support and guidance throughout my PhD program. Her knowledge, wisdom, and passion for science has inspired me immensely and contributed a great deal to my academic development. I am very grateful for her continuous encouragement and confidence in me to test novel methods and ideas. I will forever cherish my experience working in her lab.

I would also like to thank my colleagues in the laboratory who have shared their technical expertise and contributed to various parts of this work. These include, Ewa Hoffmann, Qian Zhou, Abbas Assadzadeh, Zhao Bin, Fariba Baghai Wadji, Sharon Miksys, and Linda Liu. I am also grateful for the support and friendship of all my lab and departmental colleagues, especially Ewa Hoffmann, Aman Mann, Rabindra Shivanaraine, Eric Siu, and Jill Mwenifumbo.

I am sincerely indebted to members of my PhD committee, Dr. Denis Grant and Dr. Anh Lê for their dedication, insight, and constructive criticism. In addition, I would like to thank the internal and external members of my committee Drs. David S. Riddick, Jose N. Nobrega, Albert Wong, James L. Kennedy and Kenneth E. Thummel for their time and helpful comments.

I would also like to acknowledge my sources of funding including OGS, CIHR-STPTR, CIHR- TUSP, CTCRI, and the University of Toronto.

Last but certainly not least, I would like to thank my friends and most importantly my family for their continuous support and love. This thesis is dedicated to my family, which without their endless inspiration and encouragement its completion would have not been possible.

iv Table of contents Abstract...... ii Acknowledgments ...... iv List of tables...... viii List of figures...... ix List of appendices...... xi Summary of abbreviations ...... xii Section 1 Introduction ...... 1 Statement of research problem...... 1 Purpose and objective of the thesis ...... 2 Review of the literature ...... 4 1.1 Tobacco and cigarette consumption ...... 4 1.2 Nicotine...... 5 1.2.1 Occurrence in nature and chemical properties...... 5 1.2.2 Nicotine pharmacology and addiction ...... 6 1.2.3 Nicotine pharmacokinetics ...... 8 1.2.3.1 Absorption...... 8 1.2.3.2 Distribution ...... 9 1.2.3.3 Metabolism...... 9 1.2.3.4 Excretion ...... 12 1.3 Cytochromes P450...... 13 1.3.1 Overview...... 13 1.3.2 2A subfamily ...... 15 1.3.2.1 Cytochrome P450 2A6...... 15 1.3.2.2 Cytochrome P450 2A7...... 16 1.3.2.3 Cytochrome P450 2A13...... 17 1.3.3 Cytochrome P450 2B subfamily...... 18 1.3.3.1 Cytochrome P450 2B6 ...... 18 1.4 Regulation of Cytochromes P450...... 19 1.4.1 Overview ...... 19 1.4.2 CYP2A6...... 20 1.4.3 CYP2B6...... 23 1.4.4 Genetic Variation...... 25 1.4.4.1 CYP2A6 genetic variation ...... 26 1.4.4.2 CYP2B6 genetic variation ...... 39 1.4.5 Epigenetic regulation ...... 49 1.5 Factors influencing nicotine metabolism...... 52 1.5.1 Genetic factors ...... 52 1.5.1.1 CYP2A6 ...... 52 1.5.1.2 CYP2B6 ...... 54 1.5.1.3 Additional liver ...... 54 1.5.1.4 Interethnic differences...... 57 1.5.2 Physiological influences ...... 59

v 1.5.2.1 Gender ...... 59 1.5.2.2 Age ...... 60 1.5.3 Pathological conditions...... 61 1.5.4 Medications...... 61 1.5.4.1 Inducers ...... 61 1.5.4.2 Inhibitors...... 62 1.5.5 Meals and diet...... 63 1.5.6 Smoking...... 64 1.6 Importance of understanding variability in nicotine metabolism and CYP2A6 activity……………………………………………………………………………………..66 Statement of research hypotheses ...... 68 Section 2 Thesis Chapters...... 71 Chapter 1: Novel and established CYP2A6 alleles impair in vivo nicotine metabolism in a population of black African descent ...... 71 Abstract...... 72 Introduction ...... 73 Materials and Methods ...... 75 Results ...... 82 Discussion...... 96 Acknowledgments ...... 106 Significance of chapter ...... 107 Chapter 2: A Novel CYP2A6 allele (CYP2A6*35) resulting in an amino acid substitution (Asn438Tyr) is associated with lower CYP2A6 activity in vivo ...... 108 Abstract...... 109 Introduction ...... 110 Results ...... 112 Discussion...... 119 Materials and Methods ...... 124 Acknowledgments ...... 131 Significance of Chapter ...... 132 Chapter 3: Hepatic CYP2A6 levels and nicotine metabolism: impact of genetic, physiological, environmental and epigenetic factors ...... 133 Abstract...... 134 Introduction ...... 135 Material and Methods...... 138 Results ...... 145 Discussion...... 156 Acknowledgments ...... 162 Significance of Chapter ...... 163 Chapter 4: Hepatic CYP2B6 is altered by genetic, physiologic and environmental factors but plays little role in nicotine metabolism...... 165 Abstract...... 166 Introduction ...... 167 Materials and Methods ...... 170 Results ...... 177 Discussion...... 185 Conclusions ...... 191

vi Acknowledgments ...... 192 Significance of Chapter ...... 193 Section 3 General Discussion...... 194 3.1 CYP2A6...... 194 3.1.1 Missing genetic variability in CYP2A6...... 194 3.1.2 CYP2A6 sequencing and identification of novel genetic variants...... 194 3.1.3 CYP2A6 haplotype structures ...... 197 3.1.4 Unique CYP2A6 haplotypes, CYP2A6*1B and CYP2A6*4 as examples ...... 199 3.1.5 CYP2A6 genetic variation major conclusions...... 202 3.1.6 CYP2A6 genetic variation and smoking behavior ...... 203 3.1.7 Nicotine metabolism: in vivo-in vitro correlations ...... 206 3.1.8 CYP2A6 genetic variation: in vivo-in vitro correlations ...... 209 3.1.9 Physiological factors: in vivo-in vitro correlations...... 212 3.1.10 Remaining variability in hepatic CYP2A6 expression and nicotine metabolism.. 215 3.1.11 Utility of understanding variability in nicotine metabolism...... 216 3.2 CYP2B6 ...... 218 3.2.1 Genetic and nongenetic factors underlying variability in hepatic CYP2B6 protein expression ...... 218 3.2.2 Association between CYP2B6 genetic variation, nicotine metabolism and smoking behavior ...... 220 3.3 Conclusions ...... 225 References ...... 227 List of publications and abstracts...... 260 Appendices...... 263

vii List of tables

Table 1. A detailed illustration of the haplotype structure of currently numbered CYP2A6 alleles, pages 28-31.

Table 2. CYP2A6 allele frequencies among different world populations, pages 37-38.

Table 3. A detailed illustration of the haplotype structure of currently numbered CYP2B6 alleles, pages 40-42.

Table 4. CYP2B6 allele frequencies in different world populations, pages 43-44.

Table 5S. Supplementary table 5. Polymerase chain reaction first-step amplification conditions, page 77.

Table 6S. Supplementary table 6. Polymerase chain reaction second-step amplification conditions for CYP2A6*4A&D, *14, *15, *17, *20, 594G>C, 1672T>C, 2162_2163GC>A, and 5750G>C, page 77.

Table 7S. Supplementary table 7. Primers utilized in the study, page 78.

Table 8S. Supplementary table 8. Primer combinations utilized in the study, page 79.

Table 9. Novel characterized CYP2A6 alleles (A) and novel uncharacterized CYP2A6 alleles (B), pages 83-84.

Table 10. CYP2A6 genotypes and their mean in vivo adjusted 3HC/COT, page 85.

Table 11. CYP2A6 allele frequencies, page 86.

Table 12S. Supplementary table 12. DNA sequence variations in the CYP2A6 detected through sequencing, pages 97-100.

Table 13. CYP2A6*35(A and B), CYP2A6*36, and CYP2A6*37 alleles, page 113.

Table 14. CYP2A6*35 is associated with lower in vivo CYP2A6 activity, page 115.

Table 15. Kinetic parameters of the wildtype and variant CYP2A6 constructs for the metabolism of nicotine, page 116.

Table 16. Primers utilized in the study, page 125.

Table 17. Association of CYP2B6*6 with nicotine C-oxidation activity and CYP2A6, page 184.

viii List of figures

Figure 1. Trends in per capita consumption of various tobacco products (in pounds) in the United States from 1880 to 2000 among persons aged 18 years or older, page 4.

Figure 2. Chemical structure of nicotine, page 5.

Figure 3. Activation of brain nAChRs enhances the release of various key neurotransmitters that might mediate various smoking behaviors, page 7.

Figure 4. Schematic of nicotine and its metabolites recovered from human urine over 24hrs following nicotine administration, page 11.

Figure 5. CYP2 gene cluster on human 19q13.2, page 15.

Figure 6. Regulation of cytochromes P450 levels by different mechanisms, page 20.

Figure 7. Transcriptional regulation of CYP2A6, Page 21.

Figure 8. Schematic of the nuclear receptors involved in regulating CYP2B6 expression, Page 24.

Figure 9. Schematic of gene conversions thought to occur between CYP2A6 and CYP2A7, page 27.

Figure 10. 3HC/COT varies with CYP2A6 genotype, Page 88.

Figure 11. Normal, intermediate, and slow metabolism groups were associated with in vivo CYP2A6 activity and the extent of the systemic exposure to oral nicotine, page 91.

Figure 12. Levels of CYP2A6 and in vitro nicotine C-oxidation catalytic activity for heterologously expressed CYP2A6.24, .25, .26, .27, and .28 compared to wild type CYP2A6.1, page 93.

Figure 13. CYP2A6*28A confounds the traditional CYP2A6*4A&D assay, page 95.

Figure 14. Thermal stability of the CYP2A6 wild-type and variant constructs at 37°C, page 118.

Figure 15. Correlations among CYP2A6 protein levels, CYP2A6 mRNA levels, and nicotine C-oxidase activity, page 146.

Figure 16. Effect of CYP2A6 variants on CYP2A6 protein levels, CYP2A6 mRNA levels, nicotine C-oxidation activity, and the apparent affinity for nicotine, page 148.

Figure 17. Impact of gender on CYP2A6 protein levels, mRNA levels, and nicotine pharmacokinetics, page 150.

ix Figure 18. Impact of age on CYP2A6 protein levels and nicotine C-oxidation activity (Vmax), page 152.

Figure 19. DNA methylation status of the DR-4 and TF sites among liver samples with either very high or very low CYP2A6 levels/activity, page 154.

Figure 20. Comparison of DNA methylation levels at the CpG island, DR-4 site, and TF site between human liver samples and HepG2 cells, page 155.

Figure 21. RFLP assay to detect the 1459C>T SNP in CYP2B6, page 173.

Figure 22. Representative immunoblot of CYP2B6, page 175.

Figure 23. Impact of genetic variation in CYP2B6 on CYP2B6 protein expression, page 178.

Figure 24. Impact of gender and inducers on CYP2B6 protein expression, page 179.

Figure 25. Impact of age on CYP2B6 protein expression, page 180.

Figure 26. Correlations between (A) CYP2B6 and Vmax and (B) CYP2B6 and CYP2A6, page 182.

Figure 27. An illustration of a haplotype structure, page 198.

x List of appendices

The thesis is based on four original research articles outlined in Section 2. Additional original research articles that were completed during the PhD program but not included in the thesis are found in the appendix.

Appendix A: Al Koudsi N, Mwenifumbo JC, Sellers EM, Benowitz NL, Swan GE, Tyndale RF. Characterization of the novel CYP2A6*21 allele using in vivo nicotine kinetics. Eur J Clin Pharmacol. 2006 Jun;62(6):481-4.

Appendix B: Audrain-McGovern J, Al Koudsi N, Rodriguez D, Wileyto EP, Shields PG, Tyndale RF. The role of CYP2A6 in the emergence of nicotine dependence in adolescents. Pediatrics. 2007 Jan;119(1):264-74.

xi Summary of abbreviations 3’-UTR 3’-untraslated region 3HC Trans-3’-hydroxycotinine 5-AzaC 5-Aza-2′-deoxycytidine 7-EFC 7-Ethoxy-4-trifluoromethylcoumarin AHR Aryl receptor ANOVA Analysis of variance AP-1 Activator protein-1 AUC Area under the curve Constitutive androstane receptor C/EBPα CCAAT-box/enhancer binding protein-α CI Confidence interval CNS Central nervous system CNV Copy number variant CO monoxide COT Cotinine CPD Cigarettes per day CYP Cytochrome P450 DA Dopamine DNMT DNA methyltransferase DR-4 Direct repeat separated by four nucleotides ER-α receptor-α ERE Estrogen response element FMO3 Flavin-containing 3 GABA γ-aminobutyric acid GH Growth hormone GI Gastrointestinal GR Glucocorticoid receptor HNF-4α Hepatocyte nuclear factor-4α hnRNPA1 Heterogeneous nuclear ribonucleoprotein HPLC High performance chromatography Indels Small (<1 kb) genetic insertions and deletions LD Linkage disequilibrium MDMA Methylenedioxymethamphetamine NADH adenine dinuclotide NADPH Nicotinamide adenine dinuclotide phosphate nAChR Nicotinic Acetylcholine Receptor NF-Y Nuclear factor-Y NNK 4-(methylnitrosoamino)-1-(3-pyridyl)-1-butanone NNN N’-nitrosonornicotine NRT Nicotine replacement therapy Oct-1 Octamer transcription factor-1 OR Odds ratio PCR Polymerase chain reaction PBREM -responsive enhancer module PGC-1α Peroxisome proliferator-activated receptor-γ coactivator 1α POR NADPH-cytochrome P450 PXR Pregnane X receptor

xii RFLP Restriction fragment length polymorphism RXR Retinoid X receptor SNP Single nucleotide polymorphism SRC-1 receptor coactivator-1 TF Transcription factor TSA Trichostatin A UGT UDP-glucuronosyltransferase VDR Vitamin D receptor VTA Ventral tegmental area XREM Xenobiotic responsive enhancer module

xiii Section 1 Introduction

Statement of research problem

Tobacco smoking is the most important preventable cause of disease and premature death worldwide (World Health Organization 2008). The primary substance in cigarettes responsible for tobacco dependence is nicotine and dependent smokers titrate their cigarette consumption, at least in part, to maintain desired levels of nicotine (Benowitz 2008). In humans, nicotine is primarily metabolized and inactivated to cotinine (COT) by CYP2A6 (Benowitz and

Jacob 1994; Messina et al. 1997). Twin studies suggest a significant (~60%) genetic contribution to this metabolic pathway, however controlling for known CYP2A6 genetic variants at the time only modestly (~10%) accounted for this relationship (Swan et al. 2005). This suggested a gap in our knowledge of the genetic factors influencing nicotine metabolism, which could include unknown/uncharacterized CYP2A6 variants or possibly other (e.g.

CYP2B6).

Genetic variation in CYP2A6 has been associated with altered smoking behaviors (e.g. smoking frequency) and the risk for lung cancer (Malaiyandi et al. 2005), however not all studies agree (Carter et al. 2004). The discordance and lack of replication between studies is likely due to a number of factors, one of which is the lack of a strong correlation between

CYP2A6 genotype and phenotype. This is largely due to the small number of CYP2A6 alleles investigated and characterized in some populations. Thus, greater knowledge of CYP2A6 genetic variation is essential to improve the accuracy of genetic association studies between

CYP2A6, smoking behavior, and disease.

Nicotine, in the form of nicotine replacement therapies (NRTs) (e.g. patch), increases the chances of quitting smoking by approximately 50 to 70% (Stead et al. 2008). However, only a

1 small proportion (~20-30%) of smokers are able to quit (Silagy et al. 2004). The need to improve the efficacy of NRTs is warranted. NRTs were produced to replace the nicotine delivered by smoking. Thus, variability in nicotine metabolism can alter the amount of nicotine maintained from these products and as a result either enhance or reduce their efficacy.

Identifying genetic and non-genetic sources of variability in CYP2A6 mediated nicotine metabolism could allow for personalizing treatment in order to improve abstinence rates.

Purpose and objective of the thesis

A common feature of nicotine pharmacokinetic studies is the large interindividual variability observed in the rate of nicotine metabolism (Hukkanen et al. 2005). This variability in nicotine metabolism has largely been attributed to variability in CYP2A6 activity (Dempsey et al. 2004). The main objective of this thesis is to identify and characterize genetic and non- genetic sources of variability in CYP2A6 expression and activity in vivo and in vitro. In 2003-5 several large sequencing studies identified a number of single nucleotide polymorphisms (SNPs) in CYP2A6 that resulted in amino acid changes (Saito et al. 2003; Solus et al. 2004; Haberl et al.

2005). We characterized several of these SNPs in terms of their haplotype structure, allele frequency, and functional impact on CYP2A6 activity in vivo (Chapter 1). By sequencing

CYP2A6, this work also led to the identification of a nonsynonymous SNP (6458A>T, N438Y) that has not previously been described in the literature. In Chapter 2 we characterized this

SNP (6458A>T, N438Y) with respect to its haplotype structure, allele frequency, and functional impact in vitro and in vivo.

To date, many CYP2A6 genetic variants have been associated with lower rates of nicotine clearance and CYP2A6 activity in vivo. However, the exact mechanisms by which this is mediated in human livers is currently not very understood or studied. For example, do

CYP2A6 variant alleles associated with lower in vivo activity, particularly for nicotine C-

2 oxidation, encode with lower mRNA/protein expression, intrinsic activity, and/or apparent affinity for the ? To answer these questions CYP2A6 expression (protein and mRNA) and nicotine metabolic parameters (Vmax and Km) were assessed in human livers genotyped for the majority of known CYP2A6 variants (Chapter 3). Gender and age have also been shown to alter nicotine metabolism in vivo and whether these differences are due to metabolic factors (i.e. different CYP2A6 levels/activity) or other physiological factors (e.g. volume of distribution or liver blood flow) was not clear. Thus, the effect of age, gender, and medication use on CYP2A6 expression and nicotine metabolism was also tested (Chapter

3). The advantage of this study was our ability to assess the effect of age, gender, and medication use while controlling for genetic variation in CYP2A6. This is particularly important since CYP2A6 is highly polymorphic.

The identification and characterization of CYP2A6 variants in Chapters 1 and 2 helped to improve the correlation between CYP2A6 genotype and its phenotype, however it is evident that unaccounted variability in phenotype still exists, suggesting the presence of more unidentified

CYP2A6 variants and/or other sources of variability (e.g. epigenetic regulation of CYP2A6 or the role of other enzymes). In accord, we conducted a pilot study to investigate the potential role of epigenetics in regulating CYP2A6 expression in human livers (Chapter 3). Finally, in Chapter 4 we identified multiple (genetic and non-genetic) sources underlying variability in CYP2B6 protein expression in human livers and tested its putative role in nicotine C-oxidation.

3 Review of the literature

1.1 Tobacco and cigarette consumption

Tobacco is composed of the dried leaves of the cultivated plant Nicotiana tabacum

(Koob and Le Moal 2006). Humans consume tobacco in many forms; it can be combusted (e.g. cigarettes, cigars, bidis, kreteks), heated (e.g. waterpipes, nargile, hookah), or taken orally or nasally (e.g. snuff, betel quid, ) (Hammond 2009). All forms of tobacco use are toxic (Boffetta et al. 2008; World Health Organization 2008) and the predominant mode of use in the past century has been cigarettes (Figure 1), although this varies among cultures.

Figure 1 Trends in per capita consumption of various tobacco products (in pounds) in the United States from 1880 to 2000 among persons aged 18 years or older. The latest data on per capita cigarette consumption (in pounds) per year is 2.9 in 2006 compared to 3.5 in 2000. The data is from the U.S. Department of Agriculture Tobacco Outlook Report series. This figure is adapted from Koob and Le Moal 2006 and reprinted with permission from Elsevier.

Even though smoking rates in developed countries have been declining (e.g. from 42 to

21% during 1965-2008 in the U.S.) (Dube et al. 2009), smoking remains the most important avoidable cause of disease and premature death (World Health Organization 2008). It was estimated that cigarette smoking killed approximately 5 million individuals in 2000 (Ezzati and

Lopez 2003), with projections of up to 10 million deaths per year by the year 2020 if current smoking prevalence persists globally (Mackay and Eriksen 2002). The primary causes of death 4 associated with smoking are lung cancer, ischemic heart disease, and chronic obstructive pulmonary disease (Adhikiari et al. 2008). Although the deadly effects of smoking are related to other compounds in cigarette smoke, it is nicotine that establishes tobacco addiction and maintenance of cigarette use thereby influencing an individual’s lifetime exposure to these harmful chemicals (CDC 1988; Stolerman and Jarvis 1995).

1.2 Nicotine

1.2.1 Occurrence in nature and chemical properties

Nicotine (C10H14N2) is a naturally occurring alkaloid that acts as an in plants of the genus Nicotiana (e.g. tobacco) (Soloway 1976; Steppuhn et al. 2004). Other species of plants that contain nicotine include Camellia sinensis (), Solanum melongena (eggplant),

Solanum lycoperiscum (tomato), and Solanum tuberosum (potato) (Davis et al. 1991; Domino et al. 1993). Ingestion of these dietary products results in urinary cotinine (a major metabolite of nicotine) levels (0.6-6.2 ng/ml) (Davis et al. 1991) that are lower than the urinary levels observed in smokers (~650 ng/ml) (Wall et al. 1988) and the urinary cutoff established for defining nonsmokers (<15 ng/ml) (Benowitz et al. 2009).

Nicotine is a tertiary composed of a pyrrolidine and pyridine ring (Figure 2). It is a dibasic base due to its pyrolidine (pKa=7.84) and pyridine (pKa=3.04) (Gorrod and

Jacob III 1999). The pH of a nicotine solution can significantly alter its protonated state and consequently determine its absorption, distribution, pharmacology, and toxicology. Nicotine per se is not mutagenic or genotoxic (Clarke et al. 1995), instead it exhibits a number of important peripheral and central effects discussed below.

Figure 2 Chemical structure of nicotine.

5 1.2.2 Nicotine pharmacology and addiction

Nicotine produces its major pharmacological effects by binding to the nicotinic acetylcholine receptors (nAChRs). nAChRs are ligand-gated ion channels that are composed of five subunits and are present in both the central and peripheral (autonomic nervous system and the skeletal neuromuscular junction) nervous systems (Albuquerque et al. 2009). In the periphery, nicotine activates the sympathetic nervous system to increase heart rate, blood pressure, myocardial contractility, and cardiac output (Henningfield et al. 1985; Benowitz et al.

2002). Alterations in blood flow resulting in lowered skin temperature also occur due to cutaneous vasoconstriction (Waeber et al. 1984; Benowitz 1986). As for the respiratory and gastrointestinal (GI) systems, nicotine results in bronchial dilation and initial stimulation of GI motility respectively (Benowitz 1986; Coulie et al. 2001). Nicotine also suppresses appetite

(Grunberg et al. 1985; Perkins et al. 1990a) and increases metabolic rate (Perkins et al. 1990b;

Arcavi et al. 1994). Weight gain is often cited, especially among women, as a major concern for not trying to quit and for relapse following cessation (Filozof et al. 2004).

Nicotine readily crosses the blood brain barrier and binds to brain nAChRs. Activation of brain nAChRs results in the enhanced release of various key neurotransmitters including dopamine (DA), norepinephrine, acetylcholine, glutamate, serotonin, γ-aminobutyric acid

(GABA) and endorphins, all of which may contribute to various smoking behaviors (Figure 3)

(Benowitz 2008). Of importance is the increase in DA release by the mesolimbic

(dopaminergic) neurons that project from the ventral tegmental area (VTA) to the nucleus accumbens, as this pathway is involved in the rewarding properties of nicotine and other of abuse (e.g. and ) (Di Chiara and Imperato 1988; Koob 2000). Indeed, lesioning of the DA projections to the nucleus accumbens attenuates self-administration of nicotine (Corrigall et al. 1992) and other drugs of abuse (e.g. cocaine and amphetamine) in rats

6 (Lyness et al. 1979; Roberts and Koob 1982). It is important to note that GABA and glutamate also modulate the activation of the dopaminergic neurons in the VTA (Walaas and Fonnum

1980a; Walaas and Fonnum 1980b; Kalivas et al. 1993; McGehee et al. 1995); therefore, the effect of nicotine on DA release is complex and involves an interplay between direct effects of nicotine and modulatory effects of GABA and glutamate. The release of DA is associated with pleasurable experiences. As an indicator of its addictive properties nicotine is intravenously self- administered producing pleasurable effects that are related to its dose (Henningfield and

Goldberg 1983; Harvey et al. 2004). Following a cigarette smokers report improved mood, cognitive enhancement, and decreased anxiety (Benowitz 1988; Xu et al. 2007; Mendelson et al.

2008).

Figure 3 Activation of brain nAChRs enhances the release of various key neurotransmitters that might mediate various smoking behaviors. The figure is adapted from Benowitz 2008 and reprinted with permission from Nature Publishing Group.

Repeated nicotine exposure results in neuroadaptive processes that include nAChR inactivation and desensitization (short-term) (Corringer et al. 1998) followed by long-term upregulation (i.e. increase in number) (Marks et al. 1992; Perry et al. 1999). This contributes to the development of tolerance and subsequent withdrawal symptoms following abrupt abstinence from cigarettes or nicotine administration (Epping-Jordan et al. 1998; Teneggi et al. 2002).

Some of the neuroadaptive changes (e.g. nAChR upregulation) are long lasting (1-3 months) contributing to persistent craving and difficulties with stopping smoking (Cosgrove et al. 2009).

7 The pharmacological effects of nicotine are dependent on its pharmacodynamics and pharmacokinetics (discussed below).

1.2.3 Nicotine pharmacokinetics

1.2.3.1 Absorption

The amount of nicotine that enters the systemic circulation is highly dependent on: 1) the variety of ways by which nicotine-containing products are administered and consumed and 2) the pH of the body fluids and surfaces that come in contact with nicotine. Nicotine can be absorbed via the lungs (cigarettes), skin (transdermal patch), gastrointestinal tract (lozenge), nasal mucosa (snuff), and oral mucosa (chewing tobacco) (Hukkanen et al. 2005). In addition, nicotine undergoes reabsorption across the renal tubular membranes (Hukkanen et al. 2005).

The degree of absorption of nicotine across cell membranes is highly dependent on pH (Russell and Feyerabend 1978). Since nicotine is a weak base (pKa of 8.0), acidic environments result in its ionization and poor absorption, while in basic environments nicotine is unionized facilitating its absorption (Russell and Feyerabend 1978). For example, the pH of cigarette smoke is acidic

(pH 5.5-6.0) (Brunnemann and Hoffmann 1974) resulting in minimal buccal absorption of nicotine (Gori et al. 1986), while the pH of cigar smoke is more basic (pH 6.5 or higher) allowing a greater amount of nicotine buccal absorption (Armitage et al. 1978).

Cigarettes contain approximately 10-14 mg of nicotine (Kozlowski et al. 1998) much of which (~75%) is lost during smoking due to side stream smoke and retention of nicotine in the cigarette butt (Armitage et al. 1975). Once tobacco smoke reaches the alveoli of the lungs, about

80-90% of nicotine is rapidly absorbed (Armitage et al. 1975). The efficient and rapid absorption of nicotine by the lungs is essentially due to the large surface area of the alveoli and small airways (Hukkanen et al. 2005). Depending on how a cigarette is smoked (i.e. inhalation depth and the number, duration, and volume of puffs) an average of 1 mg (range 0.3-2 mg) of

8 nicotine is absorbed per cigarette (Isaac and Rand 1972; Benowitz and Jacob 1984). Nicotine blood concentration rises quickly during cigarette consumption and is estimated to reach the brain within 10-20 seconds (Benowitz et al. 1988a; Gourlay and Benowitz 1997). Due to this rapid absorption smoking is the one of the most reinforcing and dependence-producing forms of nicotine administration (Henningfield and Keenan 1993; Benowitz 2008). In addition, smoking allows the smoker to control (i.e. “titrate”) the level of nicotine and related effects by adjusting factors such as inhalation depth and the volume and rate of puffs (Herning et al. 1983; Benowitz

2008).

1.2.3.2 Distribution

After consuming one cigarette, within 5-10 minutes, plasma levels of nicotine rise to reach a maximum level that ranges from 15 to 30 ng/ml (Benowitz et al. 1988a; Henningfield et al. 1993a; Gourlay and Benowitz 1997). Nicotine levels then decline rapidly in a biphasic manner due to its rapid distribution and metabolism (Isaac and Rand 1972). The distribution half-life of nicotine averages about 8 minutes (Feyerabend et al. 1985; Benowitz et al. 1991) and its steady-state volume of distribution averages 2.6 L/kg (Benowitz et al. 1991; Benowitz and

Jacob 1994). Typical of many basic drugs, nicotine distributes extensively into all tissues with the highest affinity being in the liver, kidney, spleen, and lung, while adipose tissue has the lowest affinity (Kyerematen and Vesell 1991; Urakawa et al. 1994; Kemp et al. 1997). Binding of nicotine to plasma proteins is estimated to be less than 5% (Benowitz et al. 1982); therefore its distribution is unlikely to be significantly altered by disorders that alter plasma protein concentrations.

1.2.3.3 Metabolism

In humans, the total clearance of nicotine is relatively high averaging about 1.2 L/min

(range 1.1-2.5 L/min), 85 to 95% of which is accounted for by metabolic (i.e. nonrenal) 9 clearance (Benowitz et al. 1991; Benowitz and Jacob 1994; Benowitz et al. 1999; Benowitz and

Jacob 2000; Benowitz et al. 2002a). The rapid and extensive metabolism of nicotine results in its short half-life of 120 minutes (range 100-150 minutes) and minor (8-10% of a dose) recovery as unchanged nicotine in the urine of smokers (Benowitz and Jacob 1994; Benowitz et al.

1994a). The major site of nicotine metabolism is the liver, with very limited contribution by the kidneys and lungs (Turner et al. 1975; Kyerematen and Vesell 1991; Hukkanen et al. 2005).

Approximately 70-80% of nicotine is metabolized to cotinine in a reaction that occurs in two sequential steps (Figure 4) (Benowitz and Jacob 1994). The first step is catalyzed by the cytochrome P450 (CYP) system (Peterson et al. 1987) and involves the production of an intermediate ion known as nicotine-Δ1’(5’)-iminium ion (Murphy 1973). CYP2A6 is the main that contributes to the formation of nicotine-Δ1’(5’)-iminium ion (80-90%) with a minor

(10-20%) role of CYP2B6 (Nakajima et al. 1996a; Messina et al. 1997; Dicke et al. 2005). The second step involves the production of cotinine from the nicotine-Δ1’(5’)-iminium ion in a reaction catalyzed by the cytosolic enzyme aldehyde oxidase (Brandange and Lindblom 1979;

Gorrod and Hibberd 1982). Recently, human liver microsomes, cDNA expressed CYP2A6, but not cDNA expressed CYP2B6 have all been shown to produce “significant” amounts of cotinine from nicotine in the absence of aldehyde oxidase (Dicke et al. 2005; von Weymarn et al. 2006) suggesting a role for CYP2A6 in catalyzing the conversion of nicotine-Δ1’(5’)-iminium ion to cotinine. Another member of the CYP2A subfamily, CYP2A13, can also catalyze nicotine’s metabolism to cotinine (Bao et al. 2005; Murphy et al. 2005), however its negligible expression in the liver limits its overall contribution (Koskela et al. 1999; Su et al. 2000). The production of cotinine from nicotine in human livers can be almost completely (80-99%) eliminated by antibody inhibition of CYP2A6 and CYP2B6 (Dicke et al. 2005).

10

Figure 4. Schematic of nicotine and its metabolites recovered from human urine over 24hrs following cigarette consumption. The circled compounds represent urinary metabolites and the associated numbers indicate the percent of systemic dose of nicotine accounted for by the specific metabolite. This figure is adapted and modified from Hukkanen et al. (2005).

Although 70-80% of nicotine is metabolized to cotinine, only 10-15% of the nicotine absorbed by smokers appears in urine as unchanged cotinine (Figure 4) (Benowitz et al. 1994a).

Thus similar to nicotine, cotinine is extensively metabolized although at a much slower rate

(0.042-0.055 L/min) resulting in a longer half-life that averages 960 minutes (range 770-1130 minutes) (Benowitz and Jacob 1994; Benowitz and Jacob 2000). The main nicotine metabolite detected in smoker’s urine is trans-3’-hydroxycotinine (3HC) accounting for approximately 33-

40% of the nicotine absorbed by smokers (Figure 4) (Benowitz et al. 1994a). 3HC is produced from cotinine (COT) via a reaction that is thought to be mediated exclusively by CYP2A6

11 (Nakajima et al. 1996; Dempsey et al. 2004). This has allowed the use of the metabolite ratio

3HC/COT as a phenotypic measure of CYP2A6 activity (Dempsey et al. 2004). The advantage of using the 3HC/COT ratio is that it is fairly stable (Lea et al. 2006) due to: 1) COT’s long half- life (~16 hrs), and 2) the formation of 3HC is limited by COT levels and will therefore be stable as a consequence of COT’s long half-life. In addition, the 3HC/COT ratio significantly correlates (r=0.78) with the total rate of nicotine clearance (Dempsey et al. 2004), suggesting its use as a proxy measure for the rate of nicotine metabolism.

Nicotine and cotinine that is not metabolized via the major pathways described above are excreted either unchanged or are metabolized to other minor metabolites (Figure 4). Nicotine and cotinine undergo N-glucuronidation by the same enzyme (UDP-glucuronosyltransferase

2B10, UGT2B10) (Chen et al. 2007; Kaivosaari et al. 2007), while trans-3’-hydroxycotinine undergoes O-glucuronidation by the enzymes UGT2B7 and UGT1A9 (Yamanaka et al. 2005).

The glucuronide conjugates of nicotine, cotinine, and trans-3’-hydroxycotinine account for approximately 3-5%, 12-17%, and 7-9% of the nicotine dose, respectively (Benowitz et al.

1994a). Other minor metabolites of nicotine and cotinine include nicotine N-oxide (4-7%), cotinine N-oxide (2-5%), nornicotine (0.4-0.8%), and norcotinine (1-2%) (Benowitz et al.

1994a). Nicotine N-oxide formation is catalyzed by flavin-containing monooxygenase 3

(FMO3) (Cashman et al. 1992), while a cytochrome P450 mediated reaction forms cotinine N- oxide (Gorrod and Wahren 1993). Nornicotine and norcotinine can be produced by cDNA expressed CYP2A6, CYP2A13, and CYP2B6 (Murphy et al. 2005; Yamanaka et al. 2005a).

1.2.3.4 Excretion

Nicotine excretion occurs primarily in the kidneys where tubular reabsorption, depending on urinary pH, plays an important role followed by glomerular filtration and tubular secretion (Hukkanen et al. 2005). Compared to its metabolic clearance, nicotine’s renal

12 clearance (0.035-0.090 L/min) plays only a minor role, accounting for approximately 5% of its total clearance in uncontrolled urinary pH settings (Benowitz and Jacob 1994). Nonetheless, in acidic urine (pH<5), nicotine is mostly ionized and tubular reabsorption is minimized, resulting in higher rates of renal clearance (~23% of total clearance), shorter half-life (~86 minutes), and a greater amount of nicotine recovered unchanged (Beckett et al. 1972; Rosenberg et al. 1980).

Conversely in alkaline urine (pH>7), nicotine is mostly unionized and tubular reabsorption is facilitated resulting in lower rates of renal clearance (~2% of total clearance), slightly shorter half-life (~110 minutes), and a smaller amount of nicotine recovered as unchanged (Beckett et al. 1972; Rosenberg et al. 1980). Similar to nicotine, cotinine’s renal clearance is a minor route of elimination, averaging approximately 12% of cotinine’s total clearance (Benowitz et al. 1983;

Benowitz and Jacob 2000). However, unlike nicotine, cotinine’s excretion is less influenced by pH because it is less basic and therefore is predominantly in the unionized form at physiological pH (Benowitz et al. 1983; Hukkanen et al. 2005). On the other hand the renal clearance of trans-

3’-hydroxycotinine plays an important role, accounting for approximately 62% of trans-3’- hydroxycotinine’s total clearance in uncontrolled urinary pH settings (Benowitz and Jacob

2001). The effect of altering urinary pH on the renal clearance of trans-3’-hydroxycotinine has not been investigated yet. The rate of renal excretion for nicotine, cotinine, and trans-3’- hydroxycotinine is influenced by urinary flow rate (Beckett et al. 1972; Hukkanen et al. 2005).

1.3 Cytochromes P450

1.3.1 Overview

It is now more than fifty years since Garfinkel and Klingenberg (Garfinkel 1958;

Klingenberg 1958) first identified an unknown cytochrome, which was later named P450 due to its characteristic spectral peak at 450 nm when bound to (Omura and Sato

1962; Omura and Sato 1964). Since then cytochromes P450 (CYP) have been identified in most 13 organisms including mammals, birds, fish, reptiles, amphibians, insects, plants, fungi, and bacteria, suggesting a common ancestral gene that has evolved over a period of 1-3 billion years

(Nebert et al. 1989a). In general, CYP enzymes oxidize hydrophobic compounds to more polar compounds for further biotransformation or subsequent excretion (Wrighton and Stevens 1992).

This process includes multiple steps that require the transfer of two electrons from nicotinamide adenine dinuclotide phosphate (NADPH) mediated by the partner NADPH-cytochrome

P450 oxidoreductase (Degtyarenko and Archakov 1993; Lewis and Hlavica 2000). Depending on the CYP isoform, the second of the two electrons required could be donated by nicotinamide adenine dinuclotide (NADH) via cytochrome b5 (Degtyarenko and Archakov 1993; Lewis and

Hlavica 2000). In humans, CYP enzymes and their redox partners are membrane bound proteins localized predominantly in the smooth endoplasmic reticulum of the liver and other tissues

(Stier 1976). The large surface area to volume ratio of the endoplasmic reticulum makes it an ideal and efficient surface for supporting multiple CYP mediated reactions (Lewis and Pratt

1998). To date, 57 functional CYP genes and 58 belonging to 18 families and 42 subfamilies have been identified in humans (Nelson 2009). Among the 18 CYP families, CYP1,

CYP2, and CYP3 members contribute to the oxidative metabolism of more than 90% of clinical drugs (Nebert and Russell 2002). The CYP2 family has the largest number of members, many of which have been mapped to a CYP2 gene cluster of about 350kb on chromosome 19q13.2

(Hoffman et al. 1995; Hoffman et al. 2001) (Figure 5). This cluster contains CYP gene loci from the CYP2A, 2B, 2F, 2G, 2S, and 2T subfamilies (Di et al. 2009).

14 110 Zhou et al.

from the CYP2A, 2B, 2F, 2G, 2S, and 2T subfamilies with Substrate speci! of CYP2A6 six functional genes (CYP2A6, 2A7, 2B6, 2A13, 2F1, and CYP2A6 plays a crucial role in the metabolism of many 2S1) and seven pseudogenes, including CYP2A18P-T, therapeutic drugs, environmental toxicants, as well as 2A18P-N, 2B7P, 2G1P, 2G2P, 2T2P, and 2T3P (Figure 12). metabolic activation of procarcinogens, such as nitro- CYP2G1P contains a single-nucleotide deletion in exon samines (e.g., NNK) (Brown et al., 2007), 1,3-butadiene 2 and a 2.4-kbp deletion between exons 3 and 7, whereas (Duescher and Elfarra, 1994), 2,6-dichlorobenzonitrile CYP2G2P contains two nonsense mutations in exons (a ), and a$atoxin B (Le Gal et al., 2003; 1 and 3, respectively (Sheng et al., 2000). !is cluster 1 Oscarson, 2001). !e preference of CYP2A6 for small- is considered to be created by duplication events in molecule substrates, such as and nicotine, evolution. suggests that its may be small, in comparison In humans, there are three functional genes in the with those of the members of CYP2C and 3A families. CYP2A subfamily: CYP2A6, 2A7, and 2A13 (Fernandez- CYP2A6 metabolizes about 1% of clinical drugs. CYP2A6 Salguero and Gonzalez, 1995; Ho"man et al., 1995; is involved in the metabolism of valproic acid, with sub- Raunio et al., 1999). !e CYP2A6 and 2A7 genes have stantial contribution from CYP2B6 and 2C9 (Sadeque a 96% similarity in the nucleotide sequence and a 94% et al., 1997). !e reactive metabolite, 4-ene-valproic identity at the amino-acid sequence ( et al., 1989). acid, is a hepatotoxin. is metabolized by CYP2A6 codes a functional enzyme that is polymorphi- CYP2A6 as well as 3A4 (Spracklin et al., 1996). CYP2A6 is cally expressed in the human liver accounting for about responsible for the sulfoxidation and thiono-oxidation of 1–10% of total CYPs, and only trace amounts are found diethyldithiocarbamate methyl ester to form S-methyl- in extrahepatic tissues (Koskela et al., 1999), while the N,N-diethylthiolcarbamate sulfoxide, the putative active product of CYP2A7 has been shown to not incorporate metabolite responsible for the deterrent e"ects heme and is thus inactive (Ding et al., 1995; Yamano of disul#ram (Madan et al., 1998). CYP2A6, 2B6, and 3A4 et al., 1990). CYP2A13 is not expressed in the liver, but are the high K components for and expressed in the olfactory bulb and respiratory tract m ifosfamide 4-hydroxylation, while CYP2C8 and 2C9 are (Fernandez-Salguero and Gonzalez, 1995; Ho"man the low K components (Chang et al., 1993). et al., 1995; Raunio et al., 1999). In addition, the CYP2A m is a cholinergic agonist that is metabolized to pilocarpic subfamily contains two identical copies of a , acid by serum esterase. Formation of 3-hydroxypilo- CYP2A7PT and CYP2A7PC (or CYP2A7P1), which con- carpine from pilocarpine, a cholinergic agonist, is mainly tain putative CYP2A coding sequences corresponding metabolized by CYP2A6 (Endo et al., 2007). Coumarin For personal use only. to exons 1 through 5 (Fernandez-Salguero et al., 1995). strongly inhibited the formation of 3-hydroxypilocarpine CYP2A7 mRNA is expressed in liver at similar levels as by >90%. 2n-propylquinoline, a newly developed CYP2A6. for the treatment of visceral leishmaniasis, is hydroxy- !e CYP2A6 has 494 amino acids and shows a crys- lated by CYP2A6, with a contribution from CYP2E1 and tal structure with a compact and small active site (Yano 2C19 (Belliard et al., 2003). In humans, coumarin is pri- et al., 2005). !e inhibitor, methoxsalen, e"ectively #lls marily (70–80%) metabolized to 7-hydroxycoumarin by the active site cavity without substantially altering the CYP2A6 (Li et al., 1997; Miles et al., 1990; Soucek, 1999; enzyme structure. q13.4 q13.3 q13.2 q13.1 q13.1 q13.2 q13.3 q12 p12

Drug Metabolism Reviews Downloaded from informahealthcare.com by University of Toronto on 05/12/10

2A18P-C 2A18P-N 2F1 2F1P 2A7 2G1P 2B7P 2B6 2T3P 2T2P 2A6 2G2P 2A13 2S1

Figure 12. !e CYP2 cluster on human chromosome 19 13q.2. !is cluster includes loci from the CYP2A, 2B, 2F, 2G, 2S, and 2T families with six functional genes Figure(CYP2A6 ,5 2A7 CYP2, 2B6 , gene2A13, cluster2F1, and on2S1 )human and seven chromosome pseudogenes. 19q13.2. Dark boxes indicate pseudogenes. Arrow indicates the direction of transcription. This figure is adapted from Zhou et al. (2009) and reprinted with permission from Informa Healthcare.

1.3.2 Cytochrome P450 2A subfamily

The human CYP2A subfamily consists of the three genes CYP2A6, CYP2A7, and

CYP2A13 and one split pseudogene (CYP2A18P-C and CYP2A18P-N) (Hoffman et al. 2001).

CYP2A6, CYP2A7, and CYP2A13 are highly homologous exhibiting >90% identity at the

amino acid level (Hoffman et al. 2001).

1.3.2.1 Cytochrome P450 2A6

The CYP2A6 gene consists of 9 exons spanning approximately 6kb of coding and

noncoding regions. The CYP2A6 protein, formerly termed CYP2A3 (Nelson et al. 1993),

contains 494 amino acids and has been purified from human liver microsomes (Maurice et al.

1991; Yun et al. 1991). CYP2A6 is predominantly expressed in the liver with much lower

expression in extrahepatic tissues (e.g. nasal mucosa, esophageal mucosa, trachea, lung, and

skin) (Su et al. 1996; Koskela et al. 1999).

The crystal structure of CYP2A6 confirms its preference for small molecules since it has

the second smallest active site of all currently crystallized CYPs (Yano et al. 2005). Its

15 substrates include approximately 3% of therapeutic drugs (e.g. nicotine, valproic acid, halothane, tegafur, and SM-12502), environmental toxicants (e.g. gasoline additives), and many procarcinogens (e.g. B1 and tobacco specific nitrosamines such as 4-

(methylnitrosoamino)-1-(3-pyridyl)-1-butanone (NNK) and N’-nitrosonornicotine (NNN)) (Di et al. 2009). The 7-hydroxylation of coumarin has been used as a probe reaction to measure

CYP2A6 activity in vitro and in vivo (Miles et al. 1990; Xu et al. 2002). Clinically, CYP2A6 is of significance due to its major role in the metabolism of nicotine, the main addictive compound in cigarette smoke (Benowitz 2009b). In human liver microsomes, cotinine formation from nicotine correlates significantly with immunochemically determined CYP2A6 levels (r=0.60-

0.90) and coumarin 7-hydroxylase activities (r=0.83) (Nakajima et al. 1996a; Messina et al.

1997; Yamazaki et al. 1999). In addition the immunochemical and chemical inhibition of

CYP2A6 by anti-CYP2A6 IgG and coumarin respectively, effectively inhibited (75-90%) cotinine’s formation from nicotine (Nakajima et al. 1996a; Messina et al. 1997; Yamazaki et al.

1999; Dicke et al. 2005). Collectively, these results demonstrate a significant role for CYP2A6 in the metabolism of nicotine to cotinine. Large interindividual and interethnic variability in

CYP2A6 mRNA, protein, and activity levels have been observed in human liver microsomes

(Shimada et al. 1994; Rodriguez-Antona et al. 2001). Similarly, in vivo, the metabolism of nicotine and coumarin exhibit extensive variability that has largely been attributed to genetic and environmental factors (discussed in later sections) (Benowitz et al. 1982; Peamkrasatam et al. 2006).

1.3.2.2 Cytochrome P450 2A7

The CYP2A7 gene, formerly termed CYP2A4, is 95% identical to CYP2A6 at the nucleotide level (Nelson et al. 1993; Hoffman et al. 2001). Full-length CYP2A7 mRNA transcripts have been identified in human livers at levels similar to CYP2A6 mRNA (Ding et al.

16 1995; Koskela et al. 1999). In addition, cDNA expressed CYP2A7 can generate an immunoreactive protein that is similar in molecular size to CYP2A6 (Ding et al. 1995).

However, unlike CYP2A6, CYP2A7 protein is unable to incorporate heme (i.e. lacks the P450

Soret peak) and exhibits no catalytic activity towards coumarin (Yamano et al. 1990; Ding et al.

1995). Whether or not CYP2A7 mRNA is transcribed and expressed as protein in human liver is currently unknown, but is unlikely since livers heterozygous for the CYP2A6 gene deletion allele seem to have half the amount of CYP2A6 protein estimated among wildtype livers (Al

Koudsi et al. 2009).

1.3.2.3 Cytochrome P450 2A13

The CYP2A13 gene is 85% identical to CYP2A6 at the nucleotide level (Hoffman et al.

2001). The expression of CYP2A13 is highest in the nasal mucosa followed by the lung, and trachea (Su et al. 2000). CYP2A13 mRNA has also been detected, albeit at lower levels, in the liver, brain, mammary gland, prostate, testis, and uterus (Koskela et al. 1999; Su et al. 2000). cDNA expressed CYP2A13 is very efficient in catalyzing nicotine C-oxidation to cotinine and cotinine hydroxylation to trans-3’-hydroxycotinine (Bao et al. 2005; Murphy et al. 2005).

However, due to the negligible expression of CYP2A13 in the liver (Su et al. 2000), CYP2A13 is unlikely to contribute substantially to the systemic pharmacokinetic profiles of either nicotine or cotinine. Consistent with this, individuals lacking CYP2A6 have substantially reduced cotinine formation suggesting a minor role for other CYPs including CYP2A13 and CYP2B6 in the main nicotine-metabolizing pathway (Yamanaka et al. 2004). On the other hand, CYP2A13 mediated metabolism of tobacco specific nitrosamines may be more important (Su et al. 2000;

Zhang et al. 2007) and a number of associations between CYP2A13 genetic variation and the risk for developing lung cancer have been reported (Wang et al. 2003b; Cauffiez et al. 2004;

Cauffiez et al. 2005), although not all studies agree (Timofeeva et al. 2009).

17 1.3.3 Cytochrome P450 2B subfamily

The human CYP2B subfamily consists of one functional gene (CYP2B6) and one pseudogene (CYP2B7P). The two CYP2B genes lie in a block of about 112kb that appears to have been inserted into the middle of the CYP2A18P loci (Figure 5) (Hoffman et al. 2001).

1.3.3.1 Cytochrome P450 2B6

The CYP2B6 gene consists of 9 exons spanning approximately 27kb of coding and noncoding regions. CYP2B6 protein, formerly termed the phenobarbital-inducible cytochrome

P450IIB, contains 492 amino acids and is primarily expressed in the liver where it accounts for approximately 1-4% of total CYP (Mimura et al. 1993; Shimada et al. 1994; Stresser and Kupfer

1999; Hanna et al. 2000). CYP2B6 protein and/or mRNA have also been detected in extra- hepatic tissues such as the brain, kidney, intestine, lung, and nasal mucosa (Gervot et al. 1999;

Ding and Kaminsky 2003; Miksys et al. 2003).

CYP2B6 is thought to play a role in the complete or partial metabolism of numerous

(>50) substrates (Turpeinen et al. 2006) that include the procarcinogen aflatoxin B1 (Aoyama et al. 1990), the antineoplastic agent cyclophosphamide (Roy et al. 1999), the drug of abuse methylenedioxymethamphetamine (MDMA, “ecstasy”) (Kreth et al. 2000), the (Yanagihara et al. 2001) and (Court et al. 2001), the antiretroviral

(Ward et al. 2003), and the and antismoking cessation drug (Faucette et al. 2000). cDNA expressed CYP2B6 is also able to C-oxidize nicotine to cotinine, however it has an approximately 10-fold higher Km (i.e. 10-fold lower affinity) compared to cDNA expressed CYP2A6 (Yamazaki et al. 1999; Dicke et al. 2005). In one study immunochemically determined CYP2B6 levels correlated very weakly with cotinine formation, r2=0.02 (p>0.05) and 0.06 (p>0.05), at low (10µM) and high (500µM) nicotine concentrations respectively

(Yamazaki et al. 1999). In addition, immunochemical inhibition using an anti-CYP2B6 IgG

18 resulted in a modest inhibition that ranged from 0-35% (Dicke et al. 2005). Interestingly, the degree of inhibition was inversely related to hepatic CYP2A6 levels, suggesting that perhaps

CYP2B6 may be more involved in the formation of cotinine among individuals with compromised CYP2A6 levels and/or activity. Nonetheless, together the results of these studies suggest a minor role for CYP2B6 in the metabolism of nicotine.

1.4 Regulation of Cytochromes P450

1.4.1 Overview

In the past several decades it has become evident that substantial interindividual variability in one’s ability to metabolize xenobiotics exists (Kalow 2005). In some cases this has to profound effects on the clinical efficacy and/or of several important therapeutic drugs (Evans and McLeod 2003; Weinshilboum 2003; Zhou et al. 2009). Thus, it is of importance to understand the mechanisms responsible for interindividual variability. The total in vivo activity of a xenobiotic metabolizing enzyme (e.g. CYPs) is primarily determined by three factors: 1) the inherent catalytic activity of the enzyme (i.e. turnover number, which is the amount of substrate transformed per unit of enzyme per unit of time); 2) the level of the enzyme; 3) the presence of compounds that might inhibit the enzyme. Of the three factors, regulation of enzyme level seems to be the most complex.

In humans the level of CYP expressed is governed by multiple mechanisms that include, gene transcription, mRNA processing, mRNA stabilization, translation, and/or protein stabilization and degradation (Figure 6) (Porter and Coon 1991). These mechanisms work in concert to maintain basal expression of CYPs in relevant tissues. Importantly, the expression level of CYPs is not fixed; instead they are highly responsive to endogenous signals and environmental influences that result in either lower (down-regulation) or higher (induction)

19 expression. The down-regulation or induction of CYPs is thought to occur predominantly at the level of transcription, which is often mediated by specific nuclear receptors (Pelkonen et al.

2008). In some cases, the down-regulation or induction of a CYP enzyme can occur at the post- transcriptional stage. For example, induces CYP2E1 by stabilizing the CYP2E1 protein

(Lieber 1999).

Protein Gene mRNA mRNA Translation stabilization or transcription processing stabilization degradation NH2

AAAA COOH

Figure 6. Regulation of cytochromes P450 levels by different mechanisms. The figure is adapted and modified from Porter and Coon (1991).

1.4.2 CYP2A6

In comparison to other CYPs, the regulation of CYP2A6 is understudied and not well understood. Nonetheless, available data suggest the involvement of several nuclear receptors and transcription factors in its regulation (Figure 7). The constitutive expression of CYP2A6 is thought to be governed by an interplay between the following transcription factors, hepatocyte nuclear factor-4α (HNF-4α), CCAAT-box/enhancer binding protein-α (C/EBPα), and octamer transcription factor-1 (Oct-1) (Pitarque et al. 2005). In human primary hepatocytes, transfection with HNF-4α antisense RNA resulted in a dose-dependent reduction in CYP2A6 mRNA levels to approximately 60% of the control values (Jover et al. 2001). In addition, HNF-4α mRNA levels correlate significantly with CYP2A6 mRNA levels in human livers (Wortham et al.

2007). Collectively, these studies suggest an important role for HNF-4α in regulating CYP2A6.

20

-6698 -5478 -4618 -2436 -90 Oct-1

AGATCAcctgAGGTCA cGGATCActtaAGGTCA AGATCAcctgAGGTCA AGGTCAtggCAACCT ATTATGTAATcaGCCAAAGTCCA ATG

CAR, PXR, and RXR ER- HNF-4 CAR, PXR, and RXR CAR, PXR, and RXR α C/EBP α

Figure 7. Transcriptional regulation of CYP2A6. The figure illustrates the DNA sequence of the response elements to which nuclear receptors and transcription factors bind to regulate CYP2A6 expression. Throughout the thesis numbering of the nucleotides is in reference to the ATG start site of CYP2A6 (reference genomic sequence NG_000008.7) in which A is numbered 1 and the base before A is numbered -1. The methylation level of the highlighted CpG dinucleotide was investigated in Chapter 3. The figure is adapted and modified from Pitarque et al. (2005).

21 In primary human hepatocytes the expression of CYP2A6 (mRNA and protein) is induced modestly (2-3 fold) following treatment with the prototypical constitutive androstane receptor (CAR) activator, phenobabrbital, and the prototypical pregnane X receptor (PXR) ligand, (Donato et al. 2000; Rae et al. 2001; Maglich et al. 2002). Itoh et al. (2006) identified three direct repeat separated by four nucleotide (DR4)-like elements on the CYP2A6 gene (Figure 7) to which PXR and CAR could bind after dimerization with the retinoid X receptor (RXR-α) (Itoh et al. 2006). The subsequent transactivation of CYP2A6 was shown to be dependent on two (-5476 and -4618) of the three DR-4 like elements, and the presence of the peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α) (Itoh et al. 2006). In human livers the mRNA levels of CYP2A6 are significantly correlated with CAR mRNA levels, further supporting the role of CAR in regulating CYP2A6 (Wortham et al. 2007).

Recently, it has been shown that estrogen and dexamethasone can both induce CYP2A6 expression (mRNA and protein) in primary human hepatocytes (Higashi et al. 2007; Onica et al.

2008). The effect of estrogen was shown to be mediated by binding of the estrogen receptor-α

(ER-α) to its putative estrogen response element (ERE) on the CYP2A6 gene (-2,436) (Higashi et al. 2007). On the other hand, the induction by dexamethasone was shown to be mediated by the glucocorticoid receptor (GR) via a nonconventional transcriptional mechanism that involved the interaction of HNF-4α with its response element rather than the binding of the GR to a glucocorticoid response element (Onica et al. 2008).

In addition to transcriptional processes, post-transcriptional and post-translational modification could alter CYP2A6 expression. For example, CYP2A6 mRNA is thought to be stabilized by the binding of the RNA-binding protein heterogeneous nuclear ribonucleoprotein

(hnRNPA1) to the 3’-untraslated region (3’-UTR) of CYP2A6 mRNA (Christian et al. 2004).

Thus, it is possible that either variable hnRNPA1 expression or the presence of polymorphisms

22 in the 3’-UTR binding region could affect CYP2A6 mRNA stability and consequently protein expression. One of the most common forms of CYP post-translational modification is protein phosphorylation (Aguiar et al. 2005). There are two phosphorylation sites on CYP2A6, however their functional significance is currently unknown (Redlich et al. 2008).

1.4.3 CYP2B6

CYP2B6 and CYP2A6 share a number of nuclear receptors and transcription factors involved in their regulation which might, in part, account for their correlation at both the mRNA and protein level (Miles et al. 1989; Forrester et al. 1992). For example, the transfection of

C/EBPα into HepG2 cells increases CYP2B6 mRNA levels (Jover et al. 1998), while the reduction of HNF-4α using an HNF-4α antisense RNA reduces CYP2B6 mRNA levels (Jover et al. 2001). In addition, HNF-4α mRNA levels correlate significantly with CYP2B6 and

CYP2A6 mRNA levels in human livers (Wortham et al. 2007) suggesting a role for HNF-4α in regulating both CYPs.

Compared to CYP2A6, CYP2B6 is highly inducible. Following treatment of primary human hepatocytes with phenobarbital, protein levels of CYP2B6 and CYP2A6 are induced approximately 4- to 6-fold and 2- to 3-fold, respectively (Madan et al. 2003). The induction of

CYP2B6 by phenobarbital is mediated by the translocation of CAR to the nucleus, possibly via a dephosphorylation process, where it heterodimerizes with RXR-α (Figure 8) (Kawamoto et al. 1999). In the nucleus the CAR-RXR complex recruits additional transcription factors such as the steroid receptor coactivator-1 (SRC-1) (Muangmoonchai et al. 2001) and binds to the phenobarbital-responsive enhancer module (PBREM) (Sueyoshi et al. 1999) and the xenobiotic responsive enhancer module (XREM) (Wang et al. 2003) to activate the transcription of the

CYP2B6 gene (Figure 8). Car-null mice exhibit lower basal expression of cyp2b10 and no

23 induction following phenobarbital treatment demonstrating that CAR has constitutive activity, at least in mice (Wei et al. 2000; Ueda et al. 2002).

Figure 8 Schematic of the nuclear receptors involved in regulating CYP2B6 expression. Please refer to the text for a detailed description. This figure is adapted and modified from Lee (2007b).

In addition to CAR, PXR has been shown to heterodimerize with RXR and mediate the transactivation of the CYP2B6 PBREM in response to rifampicin (Goodwin et al. 2001). This

“cross-talk” between the different nuclear receptors (i.e. CAR and PXR) at the PBREM site also exists with the vitamin D receptor (VDR) (Figure 8) (Drocourt et al. 2002; Makinen et al.

2002). Together, these studies suggest that the induction of CYP2B6 by xenobiotics can be mediated by multiple nuclear receptors acting on the same site (i.e. PBREM). In human livers, the mRNA levels of CAR and PXR have been correlated with CYP2B6 mRNA suggesting an important role for these nuclear receptors in mediating interindividual variability in CYP2B6 expression (Wortham et al. 2007).

CYP2B6 protein levels can also be regulated by the glucocorticoid receptor (GR), although in an indirect manner that involves the regulation of nuclear receptors. GR agonists, for example dexamethasone, increase CAR, PXR, and RXR-α mRNA levels (Pascussi et al. 2000a;

24 Pascussi et al. 2000b) and further enhance the induction of CYP2B6 by the CAR and PXR activators, phenobarbital and rifampicin, respectively (Wang et al. 2003a).

At the post-translational stage CYP2B6 has been shown to be phosphorylated at the amino acid Ser-128 (Redlich et al. 2008). Interestingly, the rat ortholog CYP2B1 is also reported to be phosphorylated at Ser-128 (Oesch-Bartlomowicz et al. 2001; Oesch-Bartlmowicz and Oesch 2004) resulting in both enzyme deactivation and a reduced mitochondrial import rate

(Anandatheerthavarada et al. 1999; Oesch-Bartlomowicz et al. 2001). The effect of phosphorylation at CYP2B6 Ser-128 is currently unknown.

1.4.4 Genetic Variation

Introduction of variation to the DNA sequence by mutation is essential for biological evolution. Human CYP genes are highly polymorphic, however some are conserved during evolution due to fundamental endogenous functions. A summary of all the CYP alleles identified thus far can be found at the Human CYP allele nomenclature committee home page

(www.cypalleles.ki.se). According to the website (access date 21 December 2009), the highest number of variant alleles described for the CYP2 family occurs in CYP2D6 (75 alleles), followed by CYP2A6 (37 alleles) and CYP2B6 (29 alleles). The various variant alleles are a result of a number of different molecular events that include, SNPs, small (<1 kb) insertions and deletions (indels), gene conversions, gene deletions, and gene duplications and multiplications

(Ingelman-Sundberg et al. 2007). SNPs, the most common type of genetic variation, are a result of the substitution of a single nucleotide (i.e. ) at a particular site in the DNA sequence.

It has been estimated that 10 million SNPs (approximately one variant per 300 bases on average) exist in the (Reich et al. 2003). With the finding that copy number variants (CNVs) are an important class of variation in human genomes (Wain et al. 2009) and

25 the advent of efficient methods to determine whole human genome CNVs (Redon et al. 2006;

Stranger et al. 2007) an interest in identifying CNVs in CYP genes is increasing.

1.4.4.1 CYP2A6 genetic variation

The CYP2A6 gene is characterized by numerous (n=278) SNPs (NCBI dbSNP, http://www.ncbi.nlm.nih.gov/, accessed December 21 2009), multiple gene conversions with

CYP2A7, and several forms of gene deletions (CYP2A6*4A-F) and duplications

(CYP2A6*1x2A&B) (Mwenifumbo and Tyndale 2007; Mwenifumbo et al. 2010). One of the most interesting aspects of CYP2A6 polymorphisms is that many of the variant CYP2A6 alleles identified thus far are due to the presence of the highly homologous CYP2A7 locus nearby.

Evolutionarily the CYP2 gene cluster, of which CYP2A6 is a member, is thought to have originated by multiple duplication processes from a single locus (Hoffman et al. 2001). As a consequence of tandem duplications the CYP2A6 and CYP2A7 loci lie within two matching

25kb blocks that are separated by less than 6kb of unique sequence (Figure 9) (Hoffman et al.

2001). The presence of the highly homologous CYP2A6 and CYP2A7 genes, within a block of very similar sequence, is thought to have promoted mispairing and unequal DNA crossing

(Figure 9). The mispairing is evident as many of the variant CYP2A6 alleles are a result of

SNPs that are the “wildtype” or reference nucleotides found in the CYP2A7 gene (e.g.

CYP2A6*5 (Oscarson et al. 1999), *15 (Kiyotani et al. 2002), *16 (Kiyotani et al. 2002), *18

(Fukami et al. 2005), *22 (Haberl et al. 2005), *28 (Mwenifumbo et al. 2008), and *31

(Mwenifumbo et al. 2008)). Unequal crossover events can also result in gene conversions (e.g.

CYP2A6*1B) (Ariyoshi et al. 2000), gene deletions (e.g. CYP2A6*4A) (Kitagawa et al. 1999), gene duplications (e.g. CYP2A6*1x2A) (Rao et al. 2000), and hybrid genes (e.g. CYP2A6*12)

(Oscarson et al. 2002). A detailed representation of the important haplotypes of the current known CYP2A6 alleles is summarized in Table 1.

26

Figure 9 Schematic of gene conversions thought to occur between CYP2A6 and CYP2A7. The hatched and solid arrows are CYP2A6 and CYP2A7 respectively. The arrow points toward the direction of transcription. The lightly shaded blocks indicate the 25kb blocks of similar sequence in which CYP2A6 and CYP2A7 lie. (A) The unaligned crossover (denoted by X) of CYP2A6 and CYP2A7 genes in the 3-prime end can result in either the complete deletion (e.g. CYP2A6*4B) or duplication (CYP2A6*1x2B) of the CYP2A6 gene (Oscarson et al. 2002; Ariyoshi et al. 2004; Fukami et al. 2007). (B) The unaligned crossover (denoted by X) of CYP2A6 and CYP2A7 genes in intron 2 can result in the formation of a chimeric CYP2A6/CYP2A7 gene (e.g. CYP2A6*12) where the 5’-regulatory region and exons 1–2 are of CYP2A7 origin and exons 3–9 are of CYP2A6 origin (Oscarson et al. 2002; Ariyoshi et al. 2004; Fukami et al. 2007). (C) An illustration of possible gene conversions between CYP2A6 and CYP2A7, where small (<100bp) fragments of CYP2A7 could be found inserted in the CYP2A6 sequence (e.g. CYP2A6*1B) (Ariyoshi et al. 2000). The reverse has also been observed, where fragments of CYP2A6 have been found in CYP2A7 (Fukami et al. 2006). The figure is adapted and modified from Hoffman et al. (2001).

27 Table 1. A detailed illustration of the haplotype structure of currently numbered CYP2A6 alleles. X represents a 58bp CYP2A7 gene conversion. Y represents an insertion of CACTT. Z represents an insertion of GAAAAG. NA represents SNPs with unassigned alleles.

28

29

30

31 CYP2A6*1A is the wildtype or “reference” allele (Yamano et al. 1990). CYP2A6*1B has a 58bp CYP2A7 gene conversion occurring at the 3’-UTR of CYP2A6 (Yamano et al. 1990).

Many haplotypes (CYP2A6*1B1-17) have been identified that contain this 58bp CYP2A7 gene conversion (Oscarson et al. 1999; Ariyoshi et al. 2000; Kiyotani et al. 2002; Pitarque et al. 2004;

Haberl et al. 2005; Nakajima et al. 2006; Mwenifumbo et al. 2008; Mwenifumbo et al. 2008a).

In vitro, CYP2A6*1B results in higher luciferase reporter gene activity and mRNA stabilization

(Wang et al. 2006). In human liver microsomes CYP2A6*1B has been associated with higher

CYP2A6 mRNA, protein, and activity (coumarin 7-hydroxylation) levels (Wang et al. 2006). In vivo, CYP2A6*1B has been associated with higher CYP2A6 activity as measured by the cotinine to nicotine ratio (Yoshida et al. 2002), 3HC/COT ratio (Ho et al. 2009), and nicotine clearance

(Mwenifumbo et al. 2008a). However, the aforementioned results have not been widely replicated (Yoshida et al. 2003; Nakajima et al. 2006; Mwenifumbo et al. 2008; Al Koudsi et al.

2009). The lack of replication is likely due to the different ethnic populations studied which have variable frequencies of coding and noncoding SNPs in linkage disequilibrium (LD) with

CYP2A6*1B. The in vivo impact of CYP2A6*1D, *F, *G, *H, *J, and *K is currently unclear

(Nakajima et al. 2006), however in vitro *1D and *1H result in approximately 50 and 20% lower luciferase activity respectively (Pitarque et al. 2004; von Richter et al. 2004). In human livers, neither CYP2A6*1D nor CYP2A6*1H have been associated with lower CYP2A6 expression (Pitarque et al. 2004; Haberl et al. 2005).

CYP2A6*2 is a null allele that encodes a protein which fails to incorporate heme and is therefore enzymatically unstable and inactive (Yamano et al. 1990). Homozygous CYP2A6*2 individuals do not excrete detectable levels of urine 7-hydroxycoumarin following oral coumarin ingestion (Hadidi et al. 1997; Oscarson et al. 1998). CYP2A6*3 is also thought to be a null allele (Fernandez-Salguero et al. 1995), however it is very rare in all populations (Chen et al. 1999; Oscarson et al. 1999). CYP2A6*4 is a family of gene deletion alleles thought to have 32 occurred by homologous unequal crossover with CYP2A7 (Figure 9) (Nunoya et al. 1999). The various forms of this allele (i.e. CYP2A6*4A-F) are a result of the different positions at which

CYP2A6 and CYP2A7 crossover. Human livers homozygous for CYP2A6*4 do not express any detectable CYP2A6 mRNA or coumarin 7-hydroxylase activity (Kiyotani et al. 2003; Yoshida et al. 2003). In humans, CYP2A6*4 homozygosity is associated with 1) very low (Peamkrasatam et al. 2006) and in some cases undetected 7-hydroxycoumarin excretion from coumarin

(Oscarson et al. 1999a; Xu et al. 2002a) and 2) very low (Xu et al. 2002a) and in some cases undetected COT and 3HC formation from nicotine (Yoshida et al. 2002; Dempsey et al. 2004;

Nakajima et al. 2006).

CYP2A6*5 is a null allele that is the result of an amino acid substitution of the highly conserved Gly 479 with Val (Oscarson et al. 1999). When expressed in yeast CYP2A6.5 is unstable and enzymatically inactive, which is consistent with the poor coumarin 7-hydroxylation observed in humans (Oscarson et al. 1999). CYP2A6*6 encodes a protein that has a disordered holoprotein structure resulting in approximately 85% lower coumarin 7-hydroxylation

(Kitagawa et al. 2001). The in vivo impact of CYP2A6*6 is currently unclear due to its very low allele frequency across many populations (Nakajima et al. 2006). CYP2A6*7 encodes a protein that is thermally unstable (Ariyoshi et al. 2001) and has approximately 80% and 85% lower in vitro coumarin 7-hydroxylation and nicotine C-oxidation activity, respectively, compared to the wildtype (Fukami et al. 2005). In vivo, CYP2A6*7 is associated with lower 7-hydroxycoumarin excretion from coumarin (Peamkrasatam et al. 2006) and lower cotinine formation from nicotine

(Xu et al. 2002a; Nakajima et al. 2006). CYP2A6*8 is thought to have similar properties as the wildtype CYP2A6 (Yoshida et al. 2002; Xu et al. 2002a). CYP2A6*9 contains a SNP (-48T>G) in the TATA box of the CYP2A6 gene. In human livers, CYP2A6*9 is associated with lower

CYP2A6 mRNA expression and activity (coumarin 7-hydroxylation) levels (Kiyotani et al.

2003; Yoshida et al. 2003). Consistent with the lower CYP2A6 expression, CYP2A6*9 has 33 reduced activity towards all tested substrates including coumarin (Peamkrasatam et al. 2006), nicotine (Yoshida et al. 2003), and cotinine (Mwenifumbo et al. 2008). CYP2A6*10 is a haplotype resulting from the SNPs found in CYP2A6*7 and*8. In vivo, this allele is inactive towards coumarin and nicotine (Yoshida et al. 2002; Xu et al. 2002a; Peamkrasatam et al.

2006).

The CYP2A6*11 was first discovered in a Japanese patient that exhibited a very high area under the curve (AUC) for the anticancer agent tegafur (Daigo et al. 2002). In E.coli cells, expressed CYP2A6.11 exhibited approximately 60 and 40% lower activity towards the activation of tegafur and coumarin 7-hydroxylation respectively (Daigo et al. 2002). The in vivo impact of this allele on nicotine metabolism is currently unclear. CYP2A6*12 is the outcome of an unequal crossover between CYP2A6 and CYP2A7 that results in a hybrid allele where the 5- prime regulatory region and exons 1-2 are of CYP2A7 origin and exons 3-9 are of CYP2A6 origin (Oscarson et al. 2002). Expressed CYP2A6.12 is unstable and results in approximately

60% lower coumarin 7-hydroxylation activity (Oscarson et al. 2002). In human liver microsomes CYP2A6*12 is associated with lower CYP2A6 mRNA, protein, and activity

(coumarin 7-hydroxylase) levels compared to wildtype (Haberl et al. 2005; Al Koudsi et al.

2009). In vivo, this allele is associated with reduced coumarin 7-hydroxylation and nicotine clearance (Oscarson et al. 2002; Benowitz et al. 2006). In 2002, Kiyotani et al. discovered four nonsynonymous SNPs that were assigned the alleles CYP2A6*13, *14, *15, and *16 (Kiyotani et al. 2002). CYP2A6*13 and *15 also contain the -48T>G SNP in the TATA box of CYP2A6

(found in CYP2A6*9) and in vivo they are associated with lower cotinine formation from nicotine (Nakajima et al. 2006). On the other hand, CYP2A6*14 and *16 do not appear to affect enzyme activity (Nakajima et al. 2006; Ho et al. 2008). CYP2A6*17 is an excellent example of an allele occurring at relatively high frequencies (7-11%) in one world population (African

Americans) but not found in others (Caucasians, Japanese, and Koreans) (Fukami et al. 2004; 34 Mwenifumbo et al. 2008). In vitro CYP2A6.17 exhibits approximately 60 and 40% lower nicotine C-oxidation and coumarin 7-hydroxylase activities, respectively (Fukami et al. 2004).

Homozygous CYP2A6*17 individuals have approximately 12% of wildtype CYP2A6 activity remaining as measured by the nicotine/cotinine and 3HC/COT metabolite ratios (Fukami et al.

2004; Ho et al. 2009). CYP2A6*18 is a good example of an allele that exhibits a substrate specific effect. Coumarin 7-hydroxylase activity is decreased by 50%, while nicotine C- oxidation activity of this enzyme is unaffected (Fukami et al. 2005). In vivo, the nicotine/cotinine ratio is comparable between individuals with wildtype CYP2A6 and

CYP2A6*18 (Fukami et al. 2005; Nakajima et al. 2006). CYP2A6*19 encodes a protein that has approximately 90% and 70% lower coumarin 7-hydroxylation and nicotine C-oxidation activities in vitro (Fukami et al. 2005). This allele is likely to result in reduced activity in vivo, since it contains the nonsynonymous SNP in the reduced activity alleles CYP2A6*7 and *10.

CYP2A6*20 encodes a truncated protein that is essentially inactive in vitro and in vivo

(Fukami et al. 2005a; Mwenifumbo et al. 2008). To date, only two in vivo studies have assessed the impact of CYP2A6*21; one study showed no effect in a Caucasian population (Al Koudsi et al. 2006) while the other study associated it with lower activity in a population of black African descent (Mwenifumbo et al. 2008). The potential impact of CYP2A6*22 has not yet been assessed, however it is likely to be inactive since it alters the same amino acid (Leu160) present in the inactive allele CYP2A6*2 (Yamano et al. 1990; Haberl et al. 2005). CYP2A6*23 is associated with lower nicotine C-oxidation activity in vitro, and lower in vivo 3HC/COT metabolite ratio in one study (Ho et al. 2008) but not another (Ho et al. 2009). In 2008

Mwenifumbo et al. characterized a number of nonsynonymous SNPs with respect to their haplotype structure and functional impact resulting in the assignment of multiple numbered alleles (CYP2A6*24, *25, *26, *27, *28, and *31). CYP2A6*24 and *28 were associated with lower activity in vivo in one study (Mwenifumbo et al. 2008) but not another (Ho et al. 2009). 35 On the other hand, CYP2A6*25, *26, and *27 were associated with lower activity in vivo by both studies assessing their function (Mwenifumbo et al. 2008; Ho et al. 2009). The functional impact of CYP2A6*31 remains to be tested. The alleles CYP2A6*29, *30, *32, and *33 are yet to be released (http://www.cypalleles.ki.se/). CYP2A6*34 is similar to CYP2A6*12, however the crossover occurs at intron 4 resulting in a hybrid allele where the 5-prime regulatory region and exons 1-4 are of CYP2A7 origin and exons 5-9 are of CYP2A6 origin (di Iulio et al. 2009). The functional impact of this allele is unknown yet, however it is likely to result in lower activity as it contains the same nonsynonymous SNPs present in the reduced activity CYP2A6*12 and *26 alleles. CYP2A6*35 is associated with lower thermal stability and nicotine C-oxidation activity in vitro and individuals with this allele have a lower mean 3HC/COT metabolite ratio (Al

Koudsi et al. 2009a). The functional impact of CYP2A6*36 and *37 remains to be estimated, however they are likely to result in lower activity since they contain the nonsynonymous SNPs present in the reduced activity alleles CYP2A6*7, *10, and *35 (Al Koudsi et al. 2009a).

To date, two CYP2A6 gene duplication alleles have been identified (CYP2A6*1x2A and

CYP2A6*1x2B) (Rao et al. 2000; Fukami et al. 2007). CYP2A6*1x2A and CYP2A6*1x2B are thought to be the reciprocal of the gene deletion alleles CYP2A6*4D and CYP2A6*4B, respectively. Individuals with the duplicated allele CYP2A6*1x2A were shown to smoke with greater intensity as measured by their exposure to carbon monoxide (CO) (Rao et al. 2000).

There are substantial interethnic differences in the frequencies of CYP2A6 alleles (Table

2). In some cases the alleles are more prominent in one world population. For example,

CYP2A6*4, *7, and *10 are more frequent in Asian populations (especially Japanese) compared to Caucasians and individuals of black African descent (Mwenifumbo and Tyndale 2007). On the other hand, CYP2A6*17, *20, and *23 have been predominantly detected in populations of

36 Table 2. CYP2A6 allele frequencies among different world populations. Blank cells represents unavailable data. The table is adapted and modified from Mwenifumbo and Tyndale (2007).

CYP2A6 allele Caucasian African American Chinese Japanese Korean *1B1-1B17 29-35 13-20 43-51 48-55 57 *1D 30-39 48 *1H 3-11 9 5 2 *1J 0-2 0 0 0 *2 1-5 0.3-1 0 0 0 *3 0 0 0 0 0 *4A&D 0-4 2 5-15 17-24 11 *5 0-0.3 0 0.5-1 0 0.5 *6 0 0 0-0.4 0 *7 0 0 6-10 10-13 9-10 *8 0 0 0-2 0-1 0-1 *9A&B 5-8 7-9 16 19-20 20 *10 0 0 2-4 2-3 1-4 *11 0 0 0.5 1 *12A-C 0-3 0-0.4 0 0-1 0 *13 0 0 1 0.2 *14 4 1 0 0 *15 0 0 2 1 *16 0.3 0-2 0 0 *17 0 7-11 0 0 *18A-C 2 0 0 0.5 *19 0 0 0 1 *20 0 2 0 0 *21 0.5-2 1 0 0 *22 0.3 0 0 0 *23 0 2 0 0 *24 1 *25 0.5

37 CYP2A6 allele Caucasian African American Chinese Japanese Korean *26 1 *27 0.2 *28 1 *31 *34 1 out of 5 self-identified Asian individuals was heterozygous *35 0 2-3 0.5 0.8 *36 0 0 0 0 0 *37 0 0 0 0 0 *1X2A 0-1 0 0.4 0 *1x2B 0 2 0 0

References: (Kitagawa et al. 1999; Oscarson et al. 1999; Oscarson et al. 1999a; Kitagawa et al. 2001; Oscarson et al. 2002; Fukami et al. 2004; Pitarque et al. 2004; Schoedel et al. 2004; von Richter et al. 2004; Fukami et al. 2005; Gambier et al. 2005; Gyamfi et al. 2005; Huang et al. 2005; Mwenifumbo et al. 2005; Fukami et al. 2005a; Al Koudsi et al. 2006; Malaiyandi et al. 2006; Minematsu et al. 2006; Nakajima et al. 2006; Audrain-McGovern et al. 2007; Fukami et al. 2007; Mwenifumbo et al. 2008; di Iulio et al. 2009; Al Koudsi et al. 2009a)

38 black African descent (Fukami et al. 2004; Fukami et al. 2005a; Ho et al. 2008). The variable interethnic frequencies of the alleles, in combination with their functional impact, have been predicted to result in decreased (<50%) CYP2A6 activity in approximately 60, 45, 35, 25, and

5% of Japanese, Koreans, Chinese, African Americans, and Caucasians, respectively

(Mwenifumbo and Tyndale 2007). These estimates are likely to increase with time, as more alleles are discovered. Nonetheless, our current knowledge of the interethnic differences in

CYP2A6 allele frequencies is consistent with the interethnic differences seen in CYP2A6 protein expression and nicotine metabolism (Shimada et al. 1996; Parkinson et al. 2004; Nakajima et al.

2006).

1.4.4.2 CYP2B6 genetic variation

CYP2B6 has historically been overlooked as a drug-metabolizing enzyme until its recent characterization as a highly polymorphic enzyme involved in the metabolism of several important therapeutic (e.g. cyclophosphomide, bupropion, and efavirenz) and environmental compounds (e.g. organophosphorous ) (Roy et al. 1999; Ward et al. 2003; Hodgson and Rose 2007). To date, numerous (n=500) SNPs have been identified in CYP2B6 (NCBI dbSNP, http://www.ncbi.nlm.nih.gov/, accessed December 26 2009). However, unlike CYP2A6, only one gene conversion allele (CYP2B6*29) has been identified so far (Rotger et al. 2007a). A detailed representation of the currently known haplotypes of CYP2B6 alleles is summarized in

Table 3. Table 4 lists the different allele frequencies of CYP2B6 in different world populations.

39 Table 3. A detailed illustration of the haplotype structure of currently numbered CYP2B6 alleles.

40

41

42 Table 4. CYP2B6 allele frequencies in different world populations. Empty cells represent unavailable data.

CYP2B6 alleles Caucasian African American African Chinese Japanese Korean *1B 27 17 *1C *1D *1E *1F 1 *1G 48 83 *1H 20-23 19 *1J 8-9 19 *1K *1L *1M 1 *1N 1 *2A-B 3-7 3-4 2-6 3.4 4-10 3 *3 0-0.5 0 0 0 0 0 *4A-D 2-6 0-3 0 7-9 6-9 5 *5A-C 3-12 5,-9 2-4 0.3 1-4 0 *6A-C 22-28 30-35 36-62 18-24 16-23 12-16 *7A&B 0-3 0-2 1-2 0 0 0 *8 1 0 0 0 0 *9 1 2-5 0-11 2 0 0 *10 *11A&B 1 0 in a population described as Black 0 in a population described as Asian *12 *13A&B *14 *15A&B 0.4 0 in a population described as Black 0 in a population described as Asian *16 0 7 *17A&B 0 7 0-6 0 0 0 *18 0 3-8 4-9 0 0 0

43 CYP2B6 alleles Caucasian African American African Chinese Japanese Korean *19 0 1.6 0 0 *20 0 1.6 0 0 *21 0 1.6 0 0 *22 1-3 1 1-3 0 0 0 *23 0.3-0.8 *24 0.5 *25 0.3 *26 1.4 *27 0 0.6 in a population described as Black 0 in a population described as Asian *28 0 0.6 in a population described as Black 0 in a population described as Asian *29 One heterozygote individual was identified among 226 individuals (138 White, 77 Black, 7 Hispanic, and 4 Asians)

References: (Lang et al. 2001; Hiratsuka et al. 2002; Kirchheiner et al. 2003; Lamba et al. 2003; Xie et al. 2003; Cho et al. 2004; Hesse et al. 2004; Hiratsuka et al. 2004; Jacob et al. 2004; Kobayashi et al. 2004; Klein et al. 2005; Zukunft et al. 2005; Guan et al. 2006; Mehlotra et al. 2006; Guan et al. 2006a; Wang et al. 2006a; Gatanaga et al. 2007; Mehlotra et al. 2007; Rotger et al. 2007; Rotger et al. 2007a; Arenaz et al. 2009),

44 As shown in Table 3 many of the SNPs in CYP2B6 show extensive linkage disequilibrium (LD) which give rise to distinct alleles with various functional consequences.

From a technical point of view, the high LD has hampered the accuracy of comparing the allele frequencies and associated activities across different studies. CYP2B6*1A is the wildtype or

“reference” allele (Yamano et al. 1989). Of the many alleles (CYP2B6*1B-N and *22) that contain SNPs in the 5’-flanking region of CYP2B6, CYP2B6*1H, *1J, and *22 are of interest due to their frequencies and potential impact (Zanger et al. 2007). Both *1H and *1J contain two SNPs, -2320T>C and -750T>C, that are predicted to disrupt a putative HNF-4- and HNF-1- site respectively (Lamba et al. 2003). In vitro, the -750T>C SNP did not affect the transcription rate of a luciferase reporter gene in primary human hepatocytes (Zukunft et al. 2005); however it was associated with lower CYP2B6 protein expression (Lamba et al. 2003) and activity

(bupropion hydroxylase) (Hesse et al. 2004) in human livers. This finding in human livers might be due to its occurrence with other functional SNPs; therefore the in vivo effect of the SNPs

-2320T>C and -750T>C is currently inconclusive (Lamba et al. 2003). On the other hand, the

-82T>C SNP present in CYP2B6*22 has been associated with higher luciferase reporter gene transcription in primary human hepatocytes and higher CYP2B6 mRNA expression and activity

(bupropion hydroxylation) in human livers (Zukunft et al. 2005). In vivo CYP2B6*22 has been associated with lower efavirenz plasma AUC, however this did not reach statistical significance likely due to the small numbers (n=4) (Rotger et al. 2007).

Of the many CYP2B6 alleles containing nonsynonymous SNPs, CYP2B6*2, *4, *5, *6,

*9, and *18 are most commonly investigated. When expressed in COS-1 cells, CYP2B6.2

(64C>T; R22C) results in similar activity levels (7-ethoxy-4-trifluoromethylcoumarin (7-EFC)

O-deethylation) and lower protein expression (~50%) that was considered not significant by the authors (Jinno et al. 2003). In human livers, one study found no association between CYP2B6*2 and altered protein expression (Lang et al. 2001), while in another study there was a non- 45 significant trend for lower protein expression (Desta et al. 2007). The in vivo impact of

CYP2B6*2 is unclear, however there is a case report in which one homozygous CYP2B6*2 individual developed adverse CNS symptoms (insomnia and nightmare) after receiving efavirenz therapy (Usami et al. 2007). This is likely a consequence of high efavirenz plasma levels achieved due to its slower metabolism.

The SNP 785A>G (K262R) can either occur on its own to form CYP2B6*4 or with multiple haplotypes that form the alleles CYP2B6*6, *7, *13, *16, *19, *20, and *26. In most cases the variant (785A>G; K262R) is in linkage disequilibrium with 516G>T (Q172H) to form

CYP2B6*6, resulting in a lower frequency of CYP2B6*4 when K262R occurs alone. In vitro, the impact of CYP2B6*4 seems to be cell type and substrate specific. In COS-1 and HEK-293 cells, cDNA expressed CYP2B6.4 resulted in lower (~50%) protein expression but its 7-EFC O- deethylation activity was higher (Jinno et al. 2003; Wang et al. 2006a). By contrast, in yeast cells CYP2B6.4 resulted in higher (~40%) protein expression and lower bupropion hydroxylation activity (Wang et al. 2006a). Using an N-terminally truncated CYP2B6 (Δ2B6), expressed Δ2B6.4 in E.coli resulted in higher bupropion hydroxylase activity; however the Km was also higher resulting in a 50% reduction in the catalytic efficiency (Vmax/Km) of the enzyme (Bumpus et al. 2005). In a subsequent study it was found that unlike the wildtype Δ2B6,

Δ2B6.4 is not inhibited by efavirenz (Bumpus et al. 2006). These interesting findings must be interpreted with caution since the Km value of Δ2B6 for bupropion (8.8 µM) (Bumpus et al.

2005) was substantially different than that reported in human livers (130 µM) (Faucette et al.

2000). In human livers CYP2B6*4 is associated with lower protein expression however the difference did not reach significance (Lang et al. 2001). Individuals with CYP2B6*4 have been associated with a higher 3HC/COT ratio (Johnstone et al. 2006), a higher bupropion clearance

46 rate (Kirchheiner et al. 2003), and a smaller plasma efavirenz AUC (Rotger et al. 2007), suggesting it may be a gain of function allele.

The SNP 1459C>T (R487C) occurs mostly on its own to form the allele CYP2B6*5. In human livers CYP2B6*5 is associated with lower protein expression, bupropion hydroxylation, and S-mephenytoin N-demethylation, however its efavirenz hydroxylation is similar to that of the wildtype genotype (Lang et al. 2001; Desta et al. 2007). In vivo, CYP2B6*5 does not affect efavirenz metabolism or clearance (Burger et al. 2006; Rotger et al. 2007; Wyen et al. 2008).

The apparent discordance between CYP2B6 protein expression and efavirenz hydroxylation for

CYP2B6*5 could be due to an increased specific activity towards efavirenz which would mask the effect of lower protein expression, however this remains to be tested.

The most extensively investigated SNP in CYP2B6 is 516G>T (Q172H). It rarely occurs on its own as CYP2B6*9 and is usually in haplotype with the SNP 785A>G (K262R) to form the most common and important CYP2B6 allele (CYP2B6*6). In COS-1 cells, expressed

CYP2B6.6 results in lower protein expression but higher 7-EFC O-deethylase activity (Jinno et al. 2003). In human livers, CYP2B6*6 has been associated with lower protein expression and lower hydroxylation activity towards efavirenz and bupropion (Hesse et al. 2004; Desta et al.

2007). The lower CYP2B6 protein expression associated with CYP2B6*6 involves a posttranscriptional mechanism whereby the variant 516G>T (Q172H) is thought to cause aberrant splicing resulting in mRNA transcripts missing exons 4 to 6 (Hofmann et al. 2008). In vivo, CYP2B6*6 has been consistently associated with higher plasma levels of efavirenz during treatment (Tsuchiya et al. 2004; Rotger et al. 2007). At least half of the patients that receive efavirenz experience central nervous system (CNS) side effects, thought to reflect higher efavirenz plasma concentrations (Marzolini et al. 2001; Csajka et al. 2003). Interestingly,

Gatanaga et al. (2007) were able to successfully employ CYP2B6*6 genotyping to reduce the therapeutic dose of efavirenz and improve the CNS-related side effects (Gatanaga et al. 2007). 47 An opposite effect of CYP2B6*6 is observed in vivo when the substrate of interest is cyclophosphomide. Specifically, CYP2B6*6 has been associated with a higher clearance rate and a shorter half-life of cyclophosphomide (Nakajima et al. 2007). Together, these studies suggest that CYP2B6*6 might encode an enzyme with an altered active site, such that the metabolism of efavirenz and bupropion are decreased while the metabolism of cyclophosphomide is increased. The functional consequences of CYP2B6*6 described thus far are not easily explained and are likely the result of the alteration of multiple processes that include level of transcription, splicing, protein stability, and/or substrate specificity. The effect of CYP2B6*6 on nicotine metabolism is described in Section 1.5.1.2.

The 983T>C (I328T) SNP can occur on its own as CYP2B6*18 (Klein et al. 2005) or in haplotype with 785A>G (K262R) to form CYP2B6*16 (Wang et al. 2006a). Recent population studies suggest that 983T>C (I328T) usually occurs alone as CYP2B6*18 at higher frequency among populations of African descent (Mehlotra et al. 2007; Rotger et al. 2007). In vitro

CYP2B6.18 results in lower protein expression, however this was associated with a lower bupropion hydroxylation activity in one (Klein et al. 2005) but not another study (Wang et al.

2006a). In vivo CYP2B6*18 is consistently associated with highly elevated plasma concentrations of efavirenz suggesting an important impact on efavirenz metabolism, especially among individuals of African descent (Wang et al. 2006a; Rotger et al. 2007; Wyen et al. 2008).

The remaining alleles (CYP2B6*3, *7, *8, *10, *11, *12, *13, *14, *15, *16, *17, and

*19 to*28) are less common and not very well characterized in vivo. Of interest is the very rare but first example of a CYP2B6 null allele, CYP2B6*28. CYP2B6*28 harbors a premature stop codon (R378X) (Rotger et al. 2007). CYP2B6*29 is the first CYP2B6 example of a hybrid allele, which is likely the result of an unequal crossover event between CYP2B6 and CYP2B7 (Rotger et al. 2007a).

48 1.4.5 Epigenetic regulation

Although interindividual differences in CYP expression and drug metabolism have been largely attributed to genetic heterogeneity, there still remains substantial variability that is unaccounted for by DNA sequence variation. In a study by Haberl et al. (2005), the authors in reference to a Caucasian population stated: “the number of samples and the method used

(sequencing) make it unlikely that there are any frequent variants left undetected within the coding region of CYP2A6” (Haberl et al. 2005). While this statement remains to be validated, there is a growing interest in other mechanisms that might be involved in CYP regulation. These include, 1) genetic variation in CYP regulators (e.g. PXR and CAR) and coenzymes (e.g.

NADPH-cytochrome P450 oxidoreductase (POR)), 2) regulation by noncoding RNAs, and 3) regulation by epigenetic processes (Hart and Zhong 2008; Lamba 2008; Gomez and Ingelman-

Sundberg 2009).

Epigenetic processes are heritable although potentially reversible changes that do not involve variation of the DNA sequence (i.e. SNPs, insertions, etc.), but rather modification of the DNA (e.g. methylation) or its associated proteins such as histones (e.g. histone acetylation)

(Schumacher and Petronis 2006). DNA methylation involves the addition of a methyl group to the 5-carbon position of the cytosine ring, resulting in 5-methylcytosine. This reaction is catalyzed by DNA methyltransferases (DNMTs) and occurs predominantly (in eukaryotes) on cytosines 5’ to guanine, referred to as CpG dinucleotides (Jones and Takai 2001). However, this classical view of DNA methylation has recently been shown to be cell type specific such that only 25% of the methylation observed in stem cells occurred in the context of CpGs compared to 99.98% in fully differentiated fibroblast cells (Kim et al. 2009). Regions with a higher than predicted proportion of CpGs are referred to as CpG islands. These islands occur predominantly within promoter regions, however they may also occur in the coding and 3’-regions of a gene

49 (Bird 1980; Gardiner-Garden and Frommer 1987; Larsen et al. 1992). Many CYP genes

(CYP1A1, 1A2, 1B1, 2A6, 2C19, 2D6, 2E1, 2J2, 2R1, 2S1, and 2W1) contain putative CpG islands in their promoter or coding regions.

In many eukaryotes (i.e. plants and animals) DNA methylation plays a crucial role in multiple processes that include: 1) silencing of parasitic elements (e.g. transposons and retroviruses), 2) X chromosome inactivation, 3) gene imprinting, and 4) tissue-specific gene transcription (reviewed in (Matzke et al. 1999; Attwood et al. 2002)). The regulation of by DNA methylation is thought to occur by two general mechanisms. First, DNA methylation can directly repress transcription by blocking the binding of transcription factors to their binding sites. Alternatively, it can block the binding of repressors thereby facilitating gene expression. Thus, although in general hypermethylation of promoter regions is associated with transcriptional repression this is not always the case. Secondly, DNA methylation can facilitate the binding of methyl-CpG-binding proteins that subsequently recruit co-repressors and histone deacetylases resulting in histone deacetylation and chromatin condensation (remodeling) that prevents gene transcription.

DNA methylation affects the tissue-specific expression of several CYPs. For example, expression of CYP2E1 protein and mRNA in specific tissues (liver, placenta, and lung) has been associated with the level of methylation among specific CpG sites 5-prime of the gene (Botto et al. 1994). In addition, the hypomethylation of a specific CpG dinucleotide in a putative GC box

(-37 to -32bp) of CYP1A2 has recently been suggested to mediate its tissue-specific expression in the liver (Miyajima et al. 2009). Aberrant methylation status, especially among cancer cells, also alters CYP expression (Rodriguez-Antona et al. 2009). CYP1B1 overexpression and

CYP1A1 silencing in prostate cancer cells is regulated by the hypo- and hypermethylation of their promoter/enhancer regions, respectively (Tokizane et al. 2005; Okino et al. 2006).

50 Moreover, the specific expression of CYP2W1 in human colon cancer cells is associated with the hypomethylation of its CpG island in the exon 1-intron 1 junction (Gomez et al. 2007).

CYP1A2 exhibits substantial interindividual variability in its hepatic expression that is currently unaccounted for by genetic polymorphisms (Jiang et al. 2006). Hammons et al. (2001) were the first to demonstrate an association between lower hepatic CYP1A2 mRNA expression and the hypermethylation status of a CpG dinucleotide located adjacent to an activator protein-1

(AP-1) (-2759 ) (Hammons et al. 2001). More recently hepatic CYP1A2 mRNA expression has also been inversely correlated with the methylation status of its CpG island in exon 2 (Ghotbi et al. 2009). Together these studies suggest an important role for DNA methylation at multiple loci in regulating hepatic CYP1A2 expression.

DNA methylation has also been suggested to mediate some environmental influences on

CYP expression. For example, lung CYP1A1 expression is highly induced by smoking. Using lung tissues, it was demonstrated that the complete or partial methylation of multiple CpG sites in the 5’-flanking region (-1.4 kb) of CYP1A1 occurred in 33% of heavy smokers (>15 cigarettes/day), 71% of light smokers (≤15 cigarettes per day), and 98% of nonsmokers (never and ex-smokers). Methylation levels were also found to increase following one to seven days of abstinence suggesting a possible correlation between smoking and DNA demethylation of

CYP1A1 that might result in its enhanced expression.

The role of DNA methylation in regulating human CYP2A gene expression is not fully characterized, but some studies suggest a possible relationship. Recently, the allele-specific expression of CYP2A7 was reported to be dependent on its haplotype sequence and methylation status, suggesting an interesting combinatorial effect between sequence variation and methylation status (Kerkel et al. 2008). In addition, CYP2A13 is induced following the co- treatment of NCI-H441 cells with the demethylating agent 5-Aza-2′-deoxycytidine (5-AzaC) and the histone deacetylase inhibitor trichostatin A (TSA) (Ling et al. 2007). Given that 51 CYP2A6 contains a putative CpG island (Ingelman-Sundberg et al. 2007) and is highly homologous to CYP2A7 (96.5% identity) and CYP2A13 (93.5% identity), it is likely that

CYP2A6 might be epigenetically regulated. These observations supported our rationale to investigate the possible epigenetic regulation of CYP2A6 in Chapter 3. On the other hand,

CYP2B6 does not contain a CpG island nor are there any data on the epigenetic regulation of

CYP2B6.

1.5 Factors influencing nicotine metabolism

A hallmark of all nicotine pharmacokinetic studies is the extensive variability in the observed levels of nicotine and its metabolites, and their respective rates of metabolism

(Hukkanen et al. 2005). This variability has been largely attributed to genetic, physiologic, and environmental factors discussed below.

1.5.1 Genetic factors

Twin studies suggest an important genetic contribution to the total clearance of nicotine and the predominantly CYP2A6-mediated pathways such as the clearance of nicotine via cotinine and the 3HC/COT metabolite ratio (Swan et al. 2005; Swan et al. 2009). The heritability estimates for these parameters range from 50 to 68% with CYP2A6 genotype accounting for only a moderate amount (10-20%) (Swan et al. 2005; Swan et al. 2009). The lower than expected contribution of CYP2A6 genotype is likely due to the small number of

CYP2A6 alleles included since the participants were predominantly from Caucasian populations.

1.5.1.1 CYP2A6

The support for a significant role of CYP2A6 in nicotine metabolism in vivo stems from multiple studies investigating the effect of its complete genetic deletion. Individuals that lack

CYP2A6 (i.e. homozygous for the gene deletion allele CYP2A6*4) excrete only 10-30% of the

52 urinary cotinine (24 hrs collection) found in wildtype individuals following cigarette smoking

(Kitagawa et al. 1999; Yang et al. 2001; Zhang et al. 2002). A more comprehensive study that analyzed urinary nicotine and nine of its metabolites collected for 24 hrs following nicotine gum administration demonstrated that individuals lacking CYP2A6 have a substantially different metabolite profile compared to that of wildtype individuals (Yamanaka et al. 2004). For example, unlike the wildtype controls, individuals lacking CYP2A6 excreted the majority

(~95%) of nicotine in its unchanged form, nicotine N-glucuronide and nicotine N-oxide, while only trace (~5%) amounts were excreted as cotinine, cotinine N-glucuronide, and cotinine N- oxide (Yamanaka et al. 2004). No 3HC or its glucuronide were detected in individuals lacking

CYP2A6 (Yamanaka et al. 2004), whereas in general 3HC is the most abundant (30-40%) metabolite in urine of smokers (Byrd et al. 1992). Another study by Dempsey et al. (2004) did not detect 3HC in plasma and saliva of individuals lacking CYP2A6 following oral nicotine administration (Dempsey et al. 2004). However Zhang et al. (2002) detected 3HC at very low levels in urine of smokers (Zhang et al. 2002). Together these studies suggest that the majority

(~90%) of cotinine and virtually all of 3HC is formed via CYP2A6 activity.

In addition to affecting nicotine and its metabolite profile, genetic variation in CYP2A6 alters several nicotine and cotinine pharmacokinetic parameters (Nakajima et al. 2000; Xu et al.

2002a; Benowitz et al. 2006). Following oral nicotine delivery, the plasma AUC values were

3.6-fold higher for nicotine and 10-fold lower for cotinine among individuals lacking CYP2A6

(Xu et al. 2002a). Since oral administration is subject to first-pass metabolism that might have accentuated the effect, we studied the impact of common CYP2A6 genetic variants following the intravenous delivery of nicotine (Benowitz et al. 2006). This route of administration most closely resembles that of nicotine absorbed from cigarette smoke (Hukkanen et al. 2005). In this study individuals were grouped according to their CYP2A6 genotype into normal, intermediate or slow metabolizers. The most dramatic effect was seen among the slow metabolizers, which 53 had approximately 37%, 34%, and 50% lower total clearance of nicotine, total clearance of cotinine, and clearance of nicotine via the cotinine pathway, respectively (Benowitz et al. 2006).

Moreover, the half-lives of nicotine (169 vs. 113 minutes) and cotinine (1384 vs. 1030 minutes) were significantly prolonged among the slow metabolizers compared to normal metabolizers

(Benowitz et al. 2006). Together these studies demonstrate that reduced CYP2A6 activity, as predicted by genotype, dramatically increases the systemic exposure to nicotine via various routes of administration.

1.5.1.2 CYP2B6

In vitro cDNA expressed CYP2B6 is the second most active CYP in metabolizing nicotine to cotinine, however it has approximately only one-twelfth the catalytic efficiency of

CYP2A6 (Yamazaki et al. 1999). The detection of low cotinine levels among individuals that lack CYP2A6 helped form the hypothesis that perhaps CYP2B6 might play a greater role under circumstances of compromised CYP2A6 activity. In line with this Ring et al. (2007) reported that individuals with CYP2B6*6 had a faster rate of nicotine and cotinine clearance in the total population and especially among individuals genotyped as reduced activity CYP2A6 metabolizers (Ring et al. 2007). However, since cotinine is a substrate of CYP2A6 and not

CYP2B6 (Nakajima et al. 1996) the higher rates of cotinine clearance associated with

CYP2B6*6 raised the question of whether CYP2B6*6 was associated with greater CYP2A6 activity in this particular study. In a subsequent smoking cessation study CYP2B6*6 did not alter the nicotine plasma levels obtained from nicotine patch, even when the population was stratified by CYP2A6 genotype (Lee et al. 2007). Thus the effect of CYP2B6 and its genetic variability on nicotine pharmacokinetics is currently inconclusive, likely due to the contribution of CYP2A6 that is unaccounted for.

1.5.1.3 Additional liver enzymes

54 Aldehyde oxidase is a member of the molybdo-flavoenzyme family that is responsible for the metabolism of a large number of -containing compounds (Garattini et al. 2008).

Cytosolic aldehyde oxidase mediates the second step of nicotine C-oxidation in which the nicotine-Δ1’(5’)-iminium ion is oxidized to cotinine (Brandange and Lindblom 1979).

Considerable interindividual variability (50-fold) in aldehyde oxidase activity has been reported; however this does not seem to be accounted for by gender, age (>1 year), smoking status, nor disease history (Rodrigues 1994; Sugihara et al. 1997; Al-Salmy 2001; Tayama et al. 2007). The enzyme is also not easily induced (Kitamura et al. 2006), but can be inhibited in vitro by a large

(>30) number of therapeutic drugs that include and (Obach et al. 2004).

Many (>200) SNPs have been identified within the gene encoding aldehyde oxidase including nine that result in amino acid changes, however their functional significance is currently unknown (Garattini et al. 2008). It is unlikely that the variability in aldehyde oxidase activity would substantially influence the rate of cotinine formation because its catalytic efficiency in converting nicotine-Δ1’(5’)-iminium ion to cotinine is at least 25-fold greater than that of

CYP2A6 converting nicotine to nicotine-Δ1’(5’)-iminium ion (Mwenifumbo and Tyndale 2009).

As described previously, nicotine, cotinine, and 3HC all undergo phase II metabolic reactions catalyzed by UDP-glucuronosyltransferases (UGTs) to form the more polar products nicotine N-glucuronide, cotinine N-glucuronide, and 3HC O-glucuronide respectively. In smokers approximately 5, 17, and 9% of a dose of nicotine can be recovered as urinary nicotine

N-glucuronide, cotinine N-glucuronide, and 3HC O-glucuronide, respectively (Byrd et al. 1992;

Benowitz et al. 1994a). However among individuals lacking CYP2A6 up to 47% of nicotine can be recovered as nicotine N-glucuronide in urine, suggestive of metabolic rerouting (Yamanaka et al. 2004). In vitro and in vivo the glucuronidation activities of nicotine and cotinine correlate, while the glucuronidation activity of 3HC does not correlate with either nicotine or cotinine

55 (Benowitz et al. 1994a; Ghosheh and Hawes 2002; Nakajima et al. 2002; Ghosheh and Hawes

2002a; Kuehl and Murphy 2003; Kuehl and Murphy 2003a). This suggests the presence of a common UGT enzyme(s) catalyzing nicotine’s and cotinine’s glucuronidation that is different from the enzyme(s) catalyzing 3HC’s glucuronidation. Originally UGT1A4 and to a minor extent UGT1A9 were thought to mediate the N-glucuronidation of nicotine and cotinine

(Nakajima et al. 2002; Kuehl and Murphy 2003a). However, using improved analytical methods, UGT2B10 was later identified as the main catalyst for the N-glucuronidation of nicotine and cotinine (Kaivosaari et al. 2007). Many (>100) SNPs have been identified within

UGT2B10 including five that result in amino acid changes. To date, only one allele

(UGT2B10*2) has been characterized (UDP-glucuronosyltransferase allele nomenclature home page, http://www.pharmacogenomics.pha.ulaval.ca, accessed January, 5, 2010). The allele frequency of UGT2B10*2 ranges from 7-21% and 7-14% among Caucasians and African

Americans, respectively (Chen et al. 2007; Chen et al. 2008; Berg et al. 2010). These frequencies are tentative due to the small number of subjects in some studies. Human livers homozygous for the allele UGT2B10*2 exhibit a 5- and 16- fold lower glucuronidation activity towards nicotine and cotinine, respectively (Chen et al. 2007). This has also been observed in vivo, were smokers heterozygous for UGT2B10*2 had significantly (10-20%) lower glucuronide conjugation of nicotine and cotinine (Berg et al. 2010). Together, these studies suggest an important role for UGT2B10 in nicotine and cotinine glucuronide conjugation, however its effect on their pharmacokinetics (e.g. AUC and half-life) is still not clear.

The O-glucuronidation of 3HC is thought to be mediated mainly by the enzyme

UGT2B7, with a minor contribution by UGT1A9 (Yamanaka et al. 2005). A multitude (>100) of SNPs have been identified within UGT2B7, including ten that result in amino acid changes and one that encodes a premature stop codon (UDP-glucuronosyltransferase allele nomenclature home page, http://www.pharmacogenomics.pha.ulaval.ca, accessed January, 5, 2010). The 56 functional consequence of these SNPs on the glucuronide conjugation of 3HC is currently unknown.

Nicotine N-oxide is a minor (4-7%) urinary metabolite of nicotine that could be detected at higher levels (~31%) in situations where nicotine C-oxidation activity by CYP2A6 is lacking

(i.e. in CYP2A6*4 homozygous individuals) (Benowitz et al. 1994a; Yamanaka et al. 2004). The production of nicotine N-oxide is catalyzed by FMO3 (Cashman et al. 1992) which is the predominant isoform in human hepatocytes (Krueger and Williams 2005). Immunodetected

FMO3 protein levels and nicotine N-oxidation activity levels have been shown to vary by approximately 6- and 9-fold, respectively in human livers (Cashman et al. 1992; Overby et al.

1997). This variability is largely thought to be due to genetic variation in FMO3 since the human enzyme is not readily induced and its activity is not dependent on gender, age, or smoking history (Cashman et al. 1992; Cashman et al. 2001; Krueger and Williams 2005).

However, recently the mouse FMO3 has been shown to be induced by 3-methylcholanthrene and 2,3,7,8-tetrachlorodibenzo-p-dioxin via the activation of the aryl hydrocarbon receptor

(Celius et al. 2008; Celius et al. 2010). The human FMO3 gene is highly polymorphic with currently 40 alleles identified (http://human-fmo3.biochem.ucl.ac.uk/Human_FMO3/ accessed

January, 5, 2010). The effect of these variants on nicotine N-oxidation and nicotine disposition kinetics is currently unknown, however there is a case report of a rare genetic defect in FMO3 that is associated with a loss of function and deficient nicotine N-oxidation (Ayesh et al. 1988).

1.5.1.4 Interethnic differences

Substantial interethnic variability in nicotine and cotinine metabolic clearance, nicotine and cotinine glucuronidation activity, and nicotine C-oxidation activity has been observed. For example compared to Caucasians, smokers of Asian descent have approximately 18% and 31% lower metabolic clearance rates of nicotine and cotinine respectively (Benowitz et al. 2002a). In

57 addition, CYP2A6 mediated pathways such as the fractional clearance of nicotine via the cotinine pathway and the metabolite ratio of cotinine/nicotine has also been found to be significantly lower among Asians compared to Caucasians (Benowitz et al. 2002a; Nakajima et al. 2006). The lower nicotine clearance, cotinine clearance, and CYP2A6 activity among Asian individuals compared to Caucasians is largely attributed to genetic factors (Mwenifumbo and

Tyndale 2007). As described earlier individuals of Asian descent have significantly higher frequencies of CYP2A6 null alleles compared to Caucasians (Table 2). The glucuronidation of nicotine and its metabolites does not differ between Asians and Caucasians (Benowitz et al.

2002a).

Individuals of black African descent have also been shown to have approximately 31% lower metabolic clearance of cotinine compared to Caucasians (Perez-Stable et al. 1998;

Benowitz et al. 1999). The metabolic clearance of nicotine is also lower (12%) but this did not reach statistical significance (Perez-Stable et al. 1998; Benowitz et al. 1999). Nonetheless, more specific CYP2A6 mediated pathways such as the fractional clearance of nicotine via the cotinine pathway and the metabolite ratio of 3HC/COT are significantly lower among individuals of black African descent compared to Caucasians (Benowitz et al. 1999; Moolchan et al. 2006;

Kandel et al. 2007). The total glucuronidation of nicotine and cotinine is lower and exhibits a biphasic distribution among individuals of black African descent, while the excretion of 3HC glucuronide is similar to that seen in Caucasians (Benowitz et al. 1999). Both the lower

CYP2A6 and glucuronide activities towards cotinine might explain, at least in part, the slower cotinine metabolism observed in individuals of black African descent. Both environmental and genetic factors might be playing a role. A considerable proportion of African Americans smoke mentholated cigarettes compared to Caucasians (69% vs. 22% in the United States) (Giovino et al. 2004). In vitro inhibits CYP2A6 activity (coumarin 7-hydroxylase) (MacDougall et al. 2003) and smoking mentholated cigarettes results in lower metabolic clearance of nicotine 58 and its glucuronidation, but it does not affect the clearance or glucuronidation of cotinine

(Benowitz et al. 2004). Of relevance to genetics, we and others have been identifying and characterizing loss/reduction of function CYP2A6 alleles that are more prevalent in individuals of black African descent (Fukami et al. 2004; Mwenifumbo et al. 2008; Al Koudsi et al. 2009a).

This may in part explain the lower CYP2A6 activity observed among individuals of black

African descent compared to Caucasians. The biphasic distribution of nicotine and cotinine glucuronidation unique to individuals of black African descent suggests the potential presence of specific polymorphisms within UGT2B10. To date, only one allele (UGT2B10*2) has been characterized, but it is unlikely to mediate this effect as the frequency of this allele is fairly similar between individuals of African descent and Caucasians (Berg et al. 2010).

1.5.2 Physiological influences

1.5.2.1 Gender

Women have approximately 18 and 24% higher rates of nicotine and cotinine clearance, respectively compared to men (Benowitz et al. 2006a). The use of oral contraceptives further increases the rates of nicotine and cotinine clearance among women by 28 and 30%, respectively (Benowitz et al. 2006a). Consistent with this, large epidemiological studies have found a higher mean 3HC/COT metabolite ratio present among women and women taking oral contraceptives compared to men (Johnstone et al. 2006; Berlin et al. 2007; Ho et al. 2009).

Moreover, the clearance of nicotine and cotinine during pregnancy is increased by 60 and 140%, respectively compared to postpartum (Dempsey et al. 2002). The significantly (140%) higher rate of cotinine clearance during pregnancy is suggestive of CYP2A6 induction since cotinine is a low extraction drug, such that its clearance is primarily determined by the activity of its metabolizing enzyme rather than hepatic blood flow (Hukkanen et al. 2005).

59 Together these studies suggest that CYP2A6 activity is influenced by female sex hormones. Indeed, estrogen has been shown to induce CYP2A6 in an estrogen receptor-α (ER-

α) dependent manner (Higashi et al. 2007). Another possible important regulator of sexually dimorphic CYP expression is growth hormone (GH). In general, the sexually dimorphic effects of GH on human CYPs (e.g. CYP1A2 and CYP3A4) (Liddle et al. 1998; Jaffe et al. 2002; Dhir et al. 2006) are not as dramatic as those observed in rats (Waxman and O'Connor 2006). The treatment of GH-deficient children with daily subcutaneous injections of GH for 4 weeks does not seem to alter CYP2A6 activity, suggesting the lack of CYP2A6 regulation by GH (Sinues et al. 2008a).

Several in vitro studies utilizing human liver have not detected significant differences in the amount or activity of CYP2A6 in women compared to men (Pearce et al. 1992; Shimada et al. 1994; Baker et al. 2001; Parkinson et al. 2004). This is likely due to a combination of a small number of livers in some studies and in each study the inclusion of livers with CYP2A6 variant alleles. Thus, in Chapter 3 we investigated the influence of gender on CYP2A6 mRNA and protein levels while accounting for genetic variability in CYP2A6. This is important given the highly polymorphic nature of CYP2A6.

1.5.2.2 Age

The total, metabolic, and renal clearances of nicotine are reduced by 23, 21, and 49%, respectively among elderly individuals (65-75 years) compared to younger adults (22-34 years)

(Molander et al. 2001). Although the difference in nicotine total clearance was statistically significant the authors described it as clinically insignificant (Molander et al. 2001). In another study, the steady-state nicotine plasma levels, or estimated plasma clearance values, attained from nicotine patch were similar between several age groups (18-39, 40-59, and 60-69 years)

(Gourlay and Benowitz 1996). Together these studies suggest, at most, a modest effect of age on

60 nicotine metabolism. In human liver microsomes, CYP2A6 protein and activity levels do not correlate with age (Parkinson et al. 2004). In Chapter 3 we analyzed the influence of age on

CYP2A6 expression and activity while accounting for CYP2A6 genetic variation, a limitation of previous studies. If there are no differences in CYP2A6 expression and activity with increasing age, this would suggest that the changes in nicotine disposition kinetics described by Molander et al. (2001) are likely mediated by physiological changes associated with aging such as decreased liver mass, liver blood flow, glomerular filtration rate, and tubular function, to name a few (Rowe et al. 1976; Wynne et al. 1989; Klotz 2009).

1.5.3 Pathological conditions

Kidney failure reduces the renal and metabolic clearance of nicotine in patients with severe renal impairment compared to healthy subjects (Molander et al. 2000). Other studies have tested the effect of pathological conditions on CYP2A6 using the probe drug coumarin.

For example, Hepatitis A and alcoholic liver disease are both associated with reduced coumarin metabolism in vivo (Pelkonen et al. 1985; Sotaniemi et al. 1995; Pasanen et al. 1997). In addition, patients infected with a liver fluke parasite, Opistorchiasa viverrini, have higher rates of coumarin metabolism that is reduced following treatment (Satarug et al. 1996).

1.5.4 Medications

1.5.4.1 Inducers

A number of drugs have been shown to induce CYP2A6 in human primary hepatocytes, however only a few studies have assessed their effect on nicotine metabolism in vivo. As mentioned earlier oral contraceptive use induces CYP2A6 activity and the clearance of nicotine and cotinine in vivo (Benowitz et al. 2006a; Sinues et al. 2008). More recently, five days of repeated oral administration of the antimalarial drug was shown to increase the

61 excretion of 7-hydroxycoumarin and reduce the mean plasma AUC values of nicotine and cotinine suggesting CYP2A6 induction (Asimus et al. 2008). Finally, using coumarin as a substrate, epileptic patients treated with , , phenobarbital, and/or exhibit higher rates of metabolism suggestive of CYP2A6 induction (Sotaniemi et al.

1995).

In human primary hepatocytes CYP2A6 is induced by rifampicin, dexamethasone, phenobarbital, and pyrazole (Maurice et al. 1991; Donato et al. 2000; Meunier et al. 2000;

Robertson et al. 2000; Rae et al. 2001; Edwards et al. 2003; Madan et al. 2003). The extent of induction is variable across individuals (2- to 3-fold). Human livers from phenobarbital treated patients have higher in vitro CYP2A6 protein expression and activity as assessed by cotinine formation from nicotine (Kyerematen et al. 1990; Yamano et al. 1990; Cashman et al. 1992).

1.5.4.2 Inhibitors

To date, only two drugs, methoxsalen (an antipsoriatic agent) and (a monoamine oxidase inhibitor), have been demonstrated to inhibit nicotine metabolism in vivo

(Tyndale and Sellers 2001). These compounds are described as potent and moderately selective inhibitors of CYP2A6 since methoxsalen also inhibits CYP1A2 and CYP2A13 (Zhang et al.

2001; von Weymarn et al. 2005), whereas tranylcypromine can also inhibit CYP1A2, CYP2B6, and CYP2E1 (Taavitsainen et al. 2001; Zhang et al. 2001). Oral methoxsalen inhibits the first- pass metabolism of orally delivered nicotine (Sellers et al. 2000). In addition, oral methoxsalen can decrease the clearance of nicotine by 39% following nicotine’s subcutaneous injection

(Sellers et al. 2003). Moreover, oral methoxsalen decreases the number of cigarettes smoked and increases the routing of the procarcinogen NNK to its inactive metabolite NNAL-glucuronide

(Sellers et al. 2003). Together, these studies suggest a potential role for the use of CYP2A6 inhibitors in or cessation (Sellers et al. 2003a). Tranylcypromine also reduces

62 the first-pass metabolism of orally administered nicotine, however its effect on smoking was not studied (Tyndale and Sellers 2001).

In vitro, nicotine C-oxidation is inhibited by menthol, methoxsalen, tranylcypromine, tryptamine, coumarin, and (Nakajima et al. 1996a; Messina et al. 1997; Zhang et al.

2001; Le Gal et al. 2003; MacDougall et al. 2003; Siu and Tyndale 2008). Other therapeutic compounds that inhibit coumarin 7-hydroxylation in vitro are also likely to inhibit nicotine C- oxidation, including pilocarpine, , ketoconazole, and rifampicin (Draper et al. 1997;

Pelkonen et al. 2000; Xia et al. 2002). Examples of mechanism-based (suicide) inhibitors of

CYP2A6 include selegiline, valproic acid, and methoxsalen (Koenigs et al. 1997; Wen et al. 2001; Wen et al. 2002; Siu and Tyndale 2008).

1.5.5 Meals and diet

Nicotine is characterized as a drug with a high degree of hepatic extraction and as a consequence its metabolic clearance is likely to be affected by events that alter liver blood flow such as, meals, posture, and exercise (Hukkanen et al. 2005). Following a meal, hepatic blood flow increases by 30%, nicotine clearance increases by 40%, and nicotine concentrations decrease by approximately 18% (within 30-60 minutes) (Lee et al. 1989; Gries et al. 1996). In addition to meals, specific diets have the potential to alter rates of nicotine clearance. For example, juice reduces the formation of cotinine from orally delivered nicotine, however the total clearance of nicotine is unaffected due to a concurrent substantial increase

(47%) in nicotine renal clearance (Hukkanen et al. 2006). Menthol a widely used flavorant in foods, mouthwash, and cigarettes inhibits nicotine metabolism in vivo (Benowitz et al. 2004).

Supplementation with soy isoflavone for 5 days reduces cotinine formation from nicotine delivered by gum in Japanese nonsmokers (Nakajima et al. 2006a). Ascorbic acid supplementation is associated with lower urinary cotinine levels suggesting a reduction in

63 nicotine C-oxidation activity (Dawson et al. 1999). However, the results from this study should be interpreted with caution since the amount and depth of cigarette inhalation nor the urinary pH was controlled for (Dawson et al. 1999). Chronic watercress consumption does not seem to affect the excretion of nicotine, cotinine, and 3HC, however the excretion of their respective glucuronide conjugates are increased, suggesting induction of the UGT enzymes involved and not CYP2A6 (Hecht et al. 1999). Using coumarin as a substrate Murphy et al. (2001) did not find an effect of chronic watercress consumption on CYP2A6 activity in vivo (Murphy et al.

2001).

Wheatgrass juice consumption and dietary/environmental exposure to cadmium induces

CYP2A6 activity in vivo using coumarin as a substrate (Rauma et al. 1996; Satarug et al. 2004;

Satarug et al. 2004a). In addition, broccoli consumption induces CYP2A6 activity assessed by a specific metabolic pathway of metabolism in vivo (Hakooz and Hamdan 2007). It is not clear whether most of these dietary influences described above are sufficient to alter rates of nicotine clearance that might modify smoking behaviors.

1.5.6 Smoking

Due to the presence of polycyclic aromatic in cigarette smoke, smoking increases the metabolism of CYP1A2 substrates as a result of CYP1A2 induction via the activation of the aryl hydrocarbon receptor (AHR) (Zevin and Benowitz 1999). On the other hand smokers have approximately 25% lower rates of nicotine clearance compared to nonsmokers, arguing against the presence of metabolic tolerance (Benowitz and Jacob 1993).

Consistent with this, two within-subject crossover design studies showed that nicotine clearance is increased by 14 and 36% following 4 and 7 days of abstinence, respectively, when compared to overnight abstinence from cigarettes (Lee et al. 1987; Benowitz and Jacob 2000). The slower nicotine clearance due to smoking is thought to be mediated, at least in part, by the inhibition

64 and/or downregulation of CYP2A6 since the metabolism of its probe substrate (coumarin) is also reduced by smoking (Iscan et al. 1994; Poland et al. 2000). The compound(s) in cigarette smoke responsible for the slower nicotine metabolism is/are currently unknown, but the involvement of carbon monoxide and cotinine has been ruled out (Zevin et al. 1997; Benowitz and Jacob 2000). β-Nicotyrine is a minor tobacco alkaloid that is a potent CYP2A6 mechanism- based (suicide) inhibitor, however it is currently unknown whether sufficient concentrations of it accumulate following cigarette smoking to effectively inhibit CYP2A6 and modulate nicotine metabolism (Denton et al. 2004). It is also possible that nicotine inhibits its own metabolism since it has been described to be a mechanism-based inhibitor of CYP2A6 (Denton et al. 2004; von Weymarn et al. 2006). In addition, nicotine might mediate the downregulation of CYP2A6.

Chronic (21 days) treatment of nonhuman primates with nicotine reduces the hepatic levels of

CYP2A mRNA, protein, and activity (assessed by nicotine C-oxidation) (Schoedel et al. 2003).

Interestingly, CYP2A6 mRNA expression in bronchial epithelial cells of smokers is reduced compared to nonsmokers (Crawford et al. 1998). Nonetheless, a very recent well-designed pharmacokinetic study observed no influence of chronic (10 days) transdermal nicotine treatment on nicotine metabolism, suggesting that nicotine is likely not the compound responsible for reduced CYP2A6 activity/nicotine metabolism among smokers (Hukkanen et al.

2010).

With respect to glucuronidation, smoking induces the excretion of 3HC glucuronide, but not that of nicotine or cotinine glucuronide conjugates (Benowitz and Jacob 2000). The increase in 3HC-glucuronide secretion in smokers is likely mediated by the induction of UGT1A9 via

AHR activation (Munzel et al. 1999).

65 1.6 Importance of understanding variability in nicotine metabolism and

CYP2A6 activity

Dependent smokers are thought to smoke, at least in part, to maintain desired levels of nicotine, a process that has been termed “nicotine-titration” and/or “nicotine-regulation”

(Russell 1980). Numerous studies show that when smokers switch to low-nicotine yield cigarettes they tend to compensate by smoking more cigarettes, or by increasing the frequency and depth of inhalation (Russell et al. 1975; Schachter et al. 1977; Ashton et al. 1979). Perhaps one of the most compelling examples of regulating nicotine intake is the observation of “vent blocking” (Kozlowski et al. 1989). Low-nicotine yield cigarettes contain a filter at the end that is ventilated with holes to dilute each puff with air. A significant proportion (40%) of smokers switching to low yield cigarettes subconsciously block these air holes with their lips or fingers, negating the benefit of this ventilation and resulting in similar nicotine and carcinogenic exposure observed in smokers of regular yield cigarettes (Kozlowski et al. 1989). Conversely, when switching to cigarettes with high nicotine yield smokers smoke fewer cigarettes and inhale less (Russell et al. 1975; Ashton et al. 1979). Pre-loading smokers with nicotine, either by intravenous infusion or nicotine products (e.g. patch and gum), results in fewer cigarettes smoked and reduces the frequency and depth of inhalation (Lucchesi et al. 1967; Russell et al.

1976; Benowitz et al. 1998). Moreover, manipulating nicotine’s pharmacokinetics by either reducing its metabolism via CYP2A6 inhibition (Sellers et al. 2000) or enhancing its excretion by urine acidification (Schachter et al. 1977a; Benowitz and Jacob 1985) results in the reducing or increasing of smoking, respectively.

Together, these examples highlight the key role of nicotine in regulating smoking behaviors. Because nicotine is a high extraction drug that is only 5% eliminated unchanged, its extensive metabolism is a major player in regulating plasma nicotine levels (Hukkanen et al.

66 2005). Individuals with genetically determined slow or absent CYP2A6 activity have been associated with an altered risk of becoming nicotine dependent (O'Loughlin et al. 2004;

Audrain-McGovern et al. 2007), reduced risk of being a smoker (Iwahashi et al. 2004; Schoedel et al. 2004), greater likelihood of cessation (Gu et al. 2000), lower cigarette consumption (Rao et al. 2000; Fujieda et al. 2004; Schoedel et al. 2004; Malaiyandi et al. 2006; Minematsu et al.

2006; Mwenifumbo et al. 2007a), and reduced inhalation (Malaiyandi et al. 2006a; Strasser et al.

2007). In addition, individuals with the 25% lowest CYP2A6 activity, assessed by the

3HC/COT ratio, are more likely to be abstinent (i.e. nonsmoking) at the end of placebo or nicotine patch treatment (Lerman et al. 2006; Patterson et al. 2008; Schnoll et al. 2009). These studies underscore the importance of CYP2A6 mediated nicotine metabolism in altering an individual’s duration and level of exposure to cigarettes, two of the leading risk factors for smoking related illnesses such as lung cancer and cardiovascular diseases. Thus, studying the variability in nicotine metabolism and CYP2A6 activity provides the potential of identifying individuals at greatest risk for smoking and also assists in personalizing treatment to improve efficacy. Not all studies have found an association between CYP2A6 genetic variation and smoking behaviors (London et al. 1999; Loriot et al. 2001; Carter et al. 2004). The lack of replication may be due to a number of factors (Al Koudsi and Tyndale 2005), one of which is the lack of a very strong correlation between current CYP2A6 genotypes and CYP2A6 phenotype. This is evident by the presence of large unaccounted variation in nicotine metabolism and CYP2A6 activity among “wildtype” CYP2A6 individuals (Benowitz et al. 2006;

Mwenifumbo et al. 2008). One aim of this thesis is to enhance the genotype-phenotype correlation by identifying new CYP2A6 variants and other sources of variation.

Nicotine is also being considered for the treatment of diseases/disorders such as attention deficit disorder (Levin et al. 1996; Potter and Newhouse 2004), Alzheimer’s disease (Engeland et al. 2002; White and Levin 2004), Parkinson’s disease (Vieregge et al. 2001; Lemay et al. 67 2004), Tourette’s syndrome (Howson et al. 2004; Orth et al. 2005), and ulcerative colitis

(McGrath et al. 2004; Ingram et al. 2005). If nicotine is prescribed for these disorders our knowledge of sources of variability in nicotine metabolism will provide an opportunity to personalize therapy and improve efficacy while reducing potential adverse effects (e.g. nausea).

In addition to its metabolism of nicotine CYP2A6 metabolically activates the structurally related tobacco specific nitrosamine procarcinogens NNK and NNN (Patten et al. 1996; Patten et al. 1997; Dicke et al. 2005). Therefore, individuals that possess reduced CYP2A6 activity are thought to be protected against tobacco related cancers by a combination of reduced activation and lower cigarette consumption. A meta-analysis of studies in Asian populations suggests that the CYP2A6 deletion allele (CYP2A6*4) is protective against lung and head and neck cancers (OR = 0.25, 95%C.I. 0.16-0.39) (Rodriguez-Antona et al. 2009). Nonetheless, these studies are described as preliminary (Rossini et al. 2008) as other studies have either found an increased (Tan et al. 2001) or similar (London et al. 1999; Loriot et al. 2001; Wang et al.

2003) risk of lung cancer among genetically determined slow CYP2A6 metabolizers.

Methodological issues, statistical power, racial differences, and polymorphisms in other genes involved in the activation or detoxification of may underlie the different results observed between studies. Further research is required to clarify the relationship between

CYP2A6 genotype and tobacco related cancers. Interestingly, a recent study by Kong et al.

(2009) suggests a potential role for CYP2A6 in the therapy of metastatic gastric cancer using the therapeutic S-1, which contains the CYP2A6 substrate tegafur (Kong et al. 2009).

Statement of research hypotheses

In Chapter 1 “Novel and established CYP2A6 alleles impair in vivo nicotine metabolism in a population of black African descent”, we hypothesize that: (1) many of the

68 nonsynonymous SNPs will be associated with slower nicotine metabolism and lower CYP2A6 activity, and (2) through sequencing new genetic variants and CYP2A6 alleles will be identified.

In Chapter 2 “A novel CYP2A6 allele (CYP2A6*35) resulting in an amino-acid substitution (Asn438Tyr) is associated with lower CYP2A6 activity in vivo”, we hypothesize that: (1) the nonsynonymous SNP (N438Y) will be associated with lower CYP2A6 activity in vivo, (2) the nonsynonymous SNPs V110L and N438Y, alone or in combination, will be associated with a lower nicotine catalytic efficiency (Vmax/Km) and lower thermal stability in vitro, and (3) through sequencing new genetic variants and CYP2A6 alleles will be identified.

In Chapter 3 “Hepatic CYP2A6 levels and nicotine metabolism: impact of genetic, physiological, environmental, and epigenetic factors”, we hypothesize that: (1) CYP2A6 mRNA, protein, and activity (nicotine C-oxidation) levels will significantly correlate, (2) some variant CYP2A6 alleles (CYP2A6*4, *9, and *12) will be associated with lower CYP2A6 mRNA expression, (3) variant CYP2A6 alleles will be associated with lower CYP2A6 protein expression, nicotine C-oxidation activity (Vmax), and affinity for nicotine (i.e. higher Km), (4) livers from female donors will have higher CYP2A6 mRNA, protein, and activity levels, but similar affinity for nicotine compared to livers from male donors, (5) CYP2A6 expression

(mRNA and protein), activity, and affinity for nicotine will not be influenced by age, (6) livers exposed to CYP2A6 inducers (phenobarbital and dexamethasone) will have higher CYP2A6 expression (mRNA and protein) and nicotine C-oxidation activity, and (7) DNA hypermethylation of the CpG island and a CpG dinucleotide within a transcription factor binding site will be associated with lower CYP2A6 (mRNA and protein) expression in human livers, cryopreserved hepatocytes, and HepG2 cells.

In Chapter 4 “CYP2B6 is altered by genetic, physiologic, and environmental factors but plays little role in nicotine metabolism”, we hypothesize that: (1) CYP2B6 will not significantly contribute to nicotine C-oxidation, (2) CYP2B6*6 will not be associated with 69 higher nicotine C-oxidation activity, (3) variant CYP2B6 alleles (CYP2B6*5 and *6) will be associated with lower CYP2B6 protein expression, (4) livers from female donors will be associated with higher CYP2B6 protein expression compared to livers from male donors, (5)

CYP2B6 protein expression will not be influenced by age, and (6) livers exposed to known

CYP2B6 inducers (phenobarbital and dexamethasone) will have higher CYP2B6 protein expression.

70 Section 2 Thesis Chapters

Chapter 1: Novel and established CYP2A6 alleles impair in vivo nicotine metabolism in a population of black African descent

Jill C. MwenifumboA, Nael Al KoudsiA, Man Ki Ho, Qian Zhou, Ewa B. Hoffmann, Edward M. Sellers, and Rachel F. Tyndale. A-the authors contributed equally to this work

Reprinted with the permission from John Wiley & Sons Inc. This chapter appears as published in Human Mutation, 29(5): 679-688, 2008 with modifications to the figure and table numbers.

Supplementary tables are denoted with an S following the table number (e.g. Table 5S)

Dr. Edward M. Sellers and Dr. Rachel F. Tyndale designed the original pharmacogenetic- pharmacokinetic study. Nael Al Koudsi and Jill C. Mwenifumbo identified SNPs from the literature that were previously uncharacterized. Nael Al Koudsi (1) designed primers and developed assay conditions to genotype the samples for CYP2A6*24, (2) assisted in genotyping for CYP2A6*21 and *25, (3) developed an assay in which 9.2 kb of the CYP2A6 gene can be specifically amplified (i.e. long-PCR) and cloned for sequencing, and (4) did long-PCR cloning and sequencing for some of the samples. Jill C. Mwenifumbo (1) genotyped samples for

CYP2A6*1B, *2, *4, *17, *20, *25, *26, and *28, (2) did long-PCR cloning and sequencing for some of the samples, (3) expressed and assessed the ability of cDNA expressed variants to metabolize nicotine to cotinine in E.coli cells, and (4) performed data analyses and wrote the manuscript. Man Ki Ho genotyped samples for CYP2A6*16 and *23. Frankie Lee genotyped samples for CYP2A6*27. Ewa B. Hoffmann, Abbas Assadzadeh, and Qian Zhou assisted with

CYP2A6 genotyping and genomic DNA extractions. Helma Nolte determined the nicotine, cotinine, and 3’-hydroxycotinine plasma levels by high pressure liquid chromatography

(HPLC).

71 Abstract

CYP2A6 is a human enzyme best known for metabolizing tobacco-related compounds, such as nicotine, cotinine, and nitrosamine precarcinogens. CYP2A6 genetic variants have been associated with smoking status, cigarette consumption, and tobacco-related cancers. Our objective was to functionally characterize four nonsynonymous CYP2A6 sequence variants with respect to their haplotype, allele frequency, and association with in vivo CYP2A6 activity. In vivo, nicotine was administered orally to 281 volunteers of black African descent. Blood samples were collected for kinetic phenotyping and CYP2A6 genotyping. In vitro, nicotine C- oxidation catalytic efficiencies of heterologously expressed variant enzymes were assessed. The four uncharacterized sequence variants were found in seven novel alleles CYP2A6*24A&B,

*25, *26, *27, and *28A&B; most were associated with impaired in vivo CYP2A6 activity.

Nicotine metabolism groupings, based on the in vivo data of variant alleles, were created. Mean

3HC/COT differed (P<0.001) between normal (100%), intermediate (64%), and slow (40%) groups. Systemic exposure to nicotine following oral administration also differed (P<0.001) between normal (100%), intermediate (139%), and slow (162%) metabolism groups. In addition, alleles of individuals with unusual phenotype-genotype relationships were sequenced, resulting in the discovery of five novel uncharacterized alleles and at least one novel duplication allele. A total of 7% of this population of black African descent had at least one of the eight novel characterized alleles and 29% had at least one previously established allele. These findings are important for increasing the accuracy of association studies between CYP2A6 genotype and behavioural, disease, or pharmacological phenotypes

72 Introduction

Nicotine is the primary psychoactive substance implicated in tobacco dependence

(Henningfield et al. 1985) and CYP2A6 mediates the majority of its metabolic inactivation to cotinine (Nakajima et al. 1996a; Messina et al. 1997). Cotinine is often used as a biomarker of tobacco exposure and its hydroxylation to 3’-hydroxycotinine is also mediated, almost exclusively, by CYP2A6 (Nakajima et al. 1996; Dempsey et al. 2004).

The gene for CYP2A6 is polymorphic (www.cypalleles.ki.se/.htm) (MIM:

122720). Although advances in the identification and characterization of CYP2A6 genetic variants have been made, gaps in this knowledge still exist, particularly in non-Caucasian and non-Asian populations. It has become increasingly apparent that genetic variation in CYP2A6 contributes to the complexity of nicotine pharmacology (Malaiyandi et al. 2005), smoking aetiology (Schoedel et al. 2004), and tobacco-related cancer risk (Fujieda et al. 2004). For example, CYP2A6 variants that result in a decrease or a loss of enzyme function have been associated with smoking-related behaviours such as a decreased likelihood of being a current smoker (Iwahashi et al. 2004; Schoedel et al. 2004), lower cigarette consumption (Rao et al.

2000; Malaiyandi et al. 2006; Minematsu et al. 2006), and increased success in smoking cessation (Gu et al. 2000). Of note, earlier studies did not uniformly find that CYP2A6 genotype associates with smoking behaviours (Carter et al. 2004). Multiple factors could have contributed to the disparate findings between studies, such as variable phenotypes used, different outcome measures, and/or the fact that the earlier studies were done before many variant alleles were identified. As more alleles are discovered, we expect that CYP2A6 genotype and smoking behavioural phenotype associations will become stronger and concordance between studies will improve.

73 Investigating CYP2A6 genetic variation is particularly important in populations of black

African descent given that African Americans have significantly higher rates of most smoking- related diseases (U.S. Department of Health and Human Services 1998) including lung cancer

(Haiman et al. 2006). In addition, this ethnoracial group has been understudied with respect to smoking aetiology, nicotine kinetics, and the CYP2A6 gene. We conducted a CYP2A6 genetic and in vivo nicotine kinetic association study in a population of black African descent. Our primary objective was to characterize four CYP2A6 nonsynonymous sequence variants, namely

594G>C (p.V110L) (Haberl et al. 2005), 1672T>C (p.F118L) (Solus et al. 2004),

2162_2163GC>A (R203frameshift) (Solus et al. 2004), and 5750G>C (p.E419D) (Saito et al.

2003), with respect to their haplotype, allele frequency, and in vivo functional impact. A secondary objective was to confirm whether established CYP2A6 alleles that have been predominately discovered and characterized in Caucasian and Asian populations have similar pharmacokinetic effects in a population of black African descent.

74 Materials and Methods

Overview We conducted a CYP2A6 genetic and nicotine kinetic study in current smokers (smoked at least 5 of 7 days) and nonsmokers (1–99 cigarettes during their lifetime but not during the study period), from a community-based population of black African descent (N=281). The details of study recruitment, procedures, and pharmacokinetic analyses can be found in a previous report (Mwenifumbo et al. 2007a). Briefly, participants were interviewed regarding their smoking history and administered 4 mg of oral nicotine. The resulting nicotine, cotinine, and 3’-hydroxycotinine plasma levels were assessed by high performance liquid chromatography (HPLC) (Siu et al. 2006). Participants received 100 Canadian dollars in compensation for their time. Both the University of Toronto ethics board and the Institutional

Review Board Services (Canada) approved the study, which was conducted in accordance with the Declaration of Helsinki.

In Vivo CYP2A6 Phenotyping

The hydroxylation of cotinine to 3’-hydroxycotinine is mediated solely by CYP2A6 as determined in vivo (Dempsey et al. 2004; Yamanaka et al. 2004) and in vitro (Nakajima et al.

1996). In vivo, in persons homozygous for the CYP2A6 deletion, no 3’-hydroxycotinine is detected in the urine (Yamanaka et al. 2004). The 3HC/COT ratio correlates well with the fractional conversion of nicotine to cotinine (Dempsey et al. 2004) and is associated with the

CYP2A6 genotype in European Americans (Benowitz et al. 2006; Malaiyandi et al. 2006).

Among the participants in this study the plasma 3HC/COT ratio derived from ad libitum smoking was highly correlated with the ratio at 270 min (Spearman’s rho=0.89, P<0.001) after 4 mg of oral nicotine (Mwenifumbo et al. 2007a). The 3HC/COT ratio (measured at 270 min) and the estimated nicotine AUC (over 360 min) were used as proxy measures of CYP2A6 activity and nicotine disposition kinetics, respectively (Dempsey et al. 2004).

75 Participant Description

Recruitment was stratified to achieve nearly equal numbers with respect to gender (47% male) and smoking status (49% smoker). Participants had a median age of 33 years (range, 20–

59 years). Smokers reported smoking at least 5 of 7 days and a median of eight CPD (range, 0–

35 cigarettes). Nonsmokers had baseline breath CO levels below 10 ppm. Genetic analysis was conducted on 280 of the 281 because one DNA sample failed to amplify in the genotyping assays. Associative genetic/kinetic analyses were conducted on 270 participants with reliable genetic and pharmacokinetic data (Mwenifumbo et al. 2007a).

CYP2A6 Genotyping

Assays were developed for four uncharacterized nonsynonymous sequence variants found in the literature, namely 594G>C (p.V110L), 1672T>C (p.F118L), 2162_2163GC>A

(p.R203frameshift), and 5750G>C (p.E419D). Alleles that would likely occur at a frequency of

>1% and alter nicotine metabolism, CYP2A6*1B, *1B15, *2, *7, *8, *9, *10, *12, *16, *21, and

*23, were genotyped as previously described (Schoedel et al. 2004; Mwenifumbo et al. 2005; Al

Koudsi et al. 2006; Ho et al. 2008; Mwenifumbo et al. 2008a). We modified the assay for

CYP2A6*4A&D and developed the first and second steps of PCR-based genotyping assays for

CYP2A6*14, *15, *20, and 2162_2163GC>A (p.R203frameshift). The first step for CYP2A6*2 was used for 594G>C (p.V110L) and 1672T>C (p.F118L) genotyping assays (Goodz and

Tyndale 2002). The first step for CYP2A6*1B was used for CYP2A6*17 and 5750G>C

(p.E419D) genotyping assays (Mwenifumbo et al. 2008a). Primers and conditions utilized for all modified or novel genotyping assays are shown in Supplementary Tables 5S, 6S, 7S, and 8S

(available online at http://www.interscience.wiley.com/jpages/1059-7794/suppmat). All new

PCR-based genotyping assays were developed using cosmids containing genomic clones of

CYP2A6 as a positive control and CYP2A7 and CYP2A13 as negative controls (Rao et al. 2000;

Hoffman et al. 2001). 76 Supplementary Table 5S Polymerase chain reaction first-step amplification conditions First-step amplification I II III DNA (ng) 50 50 75 Buffer (µL) 2.5† 2.5† 2.5‡ Primers (each) (nM) 62.5 62.5 250 DMSO - - 2 % dNTPs (each) (mM) 0.2 0.2 0.5 MgCl2 (mM) 1.7 1.5 - Taq (units) 1.25 1.25 1.25∫ First denaturation (°C:sec) 95:60 95:60 98:30 Denaturation (°C:sec) 95:15 95:15 94:20 Annealing (°C:sec) 52:20 58:20 58:20 Extension (°C:sec) 72:120 72:120 68:600 Number of cycles 35 30 30 Last extension (°C:min) 72:7 72:7 68:10 I : First amplification for CYP2A6*4A&D II: First amplification for CYP2A6*15,*16, *20, and 2162_2163GC>A III: Long PCR amplification and first amplification for CYP2A6*14 and *21 † 10X Buffer (10 mM Tris pH8.8, 50 mM KCl) (Fermentas, Canada) ‡ AccuTaqTM LA 10X Buffer (Sigma-Alderich, Canada) ∫ AccuTaqTM LA polymerase (Sigma-Alderich, Canada) Supplementary Table 6S Polymerase chain reaction second-step amplification conditions for CYP2A6*4A&D, *14, *15, *17, *20, 594G>C, 1672T>C, 2162_2163GC>A, and 5750G>C Second-step established alleles Second-step uncharacterized variants *4 *14 *15 *17 *20 594G>C 1672T>C 2162_2163GC>A 5750G>C First-step product (µl) 0.8 0.6 0.8 0.8 0.8 0.8 0.8 0.8 0.8 Primers (each) (nM) 125 125 62.5 125 125 62.5 125 62.5 125 dNTPs (each) (mM) 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 MgCl2 (mM) 1.5 1.2 1 1 1.5 1.3 0.75 1.3 1.5 Taq (units) 0.3 0.3 0.3 0.4 0.3 0.3 0.3 0.3 0.3 First denaturation (°C:sec) 95:60 95:60 95:60 95:60 95:60 95:60 95:60 95:60 95:60 Denaturation (°C:sec) 95:15 95:15 95:15 95:15 95:15 95:15 95:15 95:15 95:15 Annealing (°C:sec) 50:20 60:20 56:10 58:20 56:20 58:20 65:20 57:20 58:20 Extension (°C:sec) 72:120 72:30 72:60 72:90 72:30 72:130 72:45 72:120 72:60 Number of cycles 25 18 18 22 18 20 16 18 20

77 Supplementary Table 7S Primers Reference for published Primer name Sequence (5’ to 3’) primers 2A65Pr1F ACC TAG ACT TAA TCT TCC CGT ATA C (Pitarque et al. 2001) 2A75Pr1F ACC TAG ACT TAA TCT TCC CAT ATA T 2A6*14FMW GTT TGG CAG CAG AGG AAT AG 2A6*14FMV GTT TGG CAG CAG AGG AAT AA 2A6ex2F594W GCC ACC TTC GAC TGG G 2A6ex2F594V GCC ACC TTC GAC TGG C 2A6in2ex3F1672W CAC CTC CCC AGG CGT GGT AT 2A6in2ex3F1672V CAC CTC CCC AGG CGT GGT AC E3R-1 AAC GCG CGC GGG TTC CTC GT (Gullsten et al. 1997) 2A6exin3F GGC ACT GGC GGT GAG CAG (Ho et al. 2008) 2A6in3F CTG CCT CCT GGA ATT CTG AC (Ho et al. 2008) 2A6ex4F2134W TGG GGA CCG CTT TGA CTA TA 2A6ex4F2134V TGG GGA CCG CTT TGA CTA TG 2A6ex4R2144W ACA GTG ACA GGA ACT CTT 2A6ex4R2144V ACA GTG ACA GGA ACT CTG 2A6ex4R2161W GGA AGA TTC CTA GCA TCC TGC G (Ho et al. 2008) 2A6*16RV GGA AGA TTC CTA GCA TCC TGC A 2A6ex4R2162W GGA AGA TTC CTA GCA TCC TG 2A6ex4R2162V GGA AGA TTC CTA GCA TCC TT 2A6ex5RA CAG CCC TTG CAG CAA CTG 2A6in5R GGC CTG TGT CAT CTG CCT (Ho et al. 2008) 2A6in6F ATT TCC TGC TCT GAG ACC (Al Koudsi et al. 2006) 2Aex7F GGC CAA GAT GCC CTA CAT G (Oscarson et al. 1999a) 2A6*17FMW GAG ATC CAA AGA TTT GGA GCC G 2A6*17FMV GAG ATC CAA AGA TTT GGA GCC A 2A6in7AS CTG AGA TTT CTG TCC CTA T (Fukami et al. 2004) 2A6in7F ACC CAC ATT AGA AGC TTT CTA GA (Mwenifumbo et al. 2008a) 2A7in7F CCC CAT TAG AAG CTT TCT ACT CA 2A6ex8R5750W TCT TCT TAA ACT GCC CCT TC 2A6ex8R5750V TCT TCT TAA ACT GCC CCT TG 2A6ex9WF TGA CGT GTC CCC CAA (Al Koudsi et al. 2006) 2A6ex9VF TGA CGT GTC CCC CAG (Al Koudsi et al. 2006) 2A6*1Bwt ACT GGG GGC AGG ATG GC (Oscarson et al. 1999) 2A6*1Bmut AAT GGG GGG AAG ATG CG (Oscarson et al. 1999) 2A6R4 GCT TTT TAA GAA TCT GTC TAG AA (Al Koudsi et al. 2006) 2A6R0 AGG TCA TCT AGA TTT TCT CCT ACA (Mwenifumbo et al. 2008a) Underlined nucleotides indicate where a mismatch has been introduced to a primer

78

Supplementary Table 8S Primer combinations Product Start End Forward Reverse size (bp) position† position First amplification I 2Aex7F 2A6R0 2773 5010 7783 II 2A6exin3F 2A6in5R 1834 1804 3638 III 2A65Pr1F 2A6R0 9201 -1417 7783 Second amplification CYP2A6*4A&D 2A6in7F 2A6R4 1895 5200 7095 2A7in7F 2A6R4 1906 - - CYP2A6*14 2A6*14FMW E3R-1 1815 67 1882 2A6*14FMV E3R-1 1815 67 1882 CYP2A6*15 2A6ex4F2134W 2A6ex5RA 1338 2115 3453 2A6ex4F2134V 2A6ex5RA 1338 2115 3453 CYP2A6*16 2A6in3F 2A6ex4R2161W 213 1969 2182 2A6in3F 2A6*16RV 213 1969 2182 CYP2A6*17 2A6*17FMW 2A6in7AS 382 5044 5426 2A6*17FMV 2A6in7AS 382 5044 5426 CYP2A6*20 2A6in3F 2A6ex4R2144W 189 1969 2158 2A6in3F 2A6ex4R2144V 187 1969 2156 CYP2A6*21 2A6ex9WF 2A6R4 536 6559 7095 2A6ex9WF 2A6R4 536 6559 7095 594G>C 2A6ex2F594W E3R-1 1303 579 1882 2A6ex2F594V E3R-1 1303 579 1882 1672T>C 2A6in2ex3F1672W E3R-1 229 1653 1882 2A6in2ex3F1672V E3R-1 229 1653 1882 2162_2163GC>A 2A6in3F 2A6ex4R2162W 213 1969 2182 2A6in3F 2A6ex4R2162V 212 1969 2181 5750G>C 2A6in7F 2A6ex8R5750W 569 5200 5769 2A6in7F 2A6ex8R5750V 569 5200 5769 † Relative to +1 ATG start site on the reference genomic sequence NG_000008.7

Sequencing

Genomic DNA was amplified (9.2 kb) according to a long PCR protocol (Supplementary

Tables 5S, 7S, and 8S). This region spanned from ~1.4 kb upstream to 7.8 kb downstream of the ATG start site, on reference genomic sequence NG_000008.7. The long PCR product was subcloned into a pCR-XL-TOPO vector (TOPO XL PCR Cloning Kit; Invitrogen, Burlington,

Canada) and sequenced utilizing a walk-on strategy (The Centre for Applied Genomics,

Toronto, Canada). Sequencing of the long PCR products was used to validate the genotyping

79 assays and to determine the haplotype of alleles containing the uncharacterized sequence variants. The variant CYP2A6 genes, from seven individuals with the uncharacterized sequence variants, were sequenced: two 594G>C (p.V110L), two 1672T>C (p.F118L), the one

2162_2163GC>A (p.R203frameshift), and two 5750G>C (p.E419D). In addition, the variant gene was fully sequenced from a person with the 6960_6961insGAAAAG, which is found in the 3’-flanking region and is the variant detected by the CYP2A6*1B15 assay. Both alleles from nine individuals with established CYP2A6 alleles were also sequenced as their genotype to phenotype relationships were discordant: two with the genotype CYP2A6*1/*2, four

CYP2A6*1/*4, one CYP2A6*4/*9, one CYP2A6*1/*9, and one CYP2A6*1/*17. Data from these nine individuals was not included in subsequent analyses.

CYP2A6 Activity Grouping Strategy

Our grouping strategy for the normal, intermediate, or slow nicotine metabolism groups was based on the in vivo 3HC/COT ratio as follows. Individuals with one or two CYP2A6*14 alleles were included in the normal group. Alleles were considered decrease-of-function (D) if they were associated with modestly reduced 3HC/COT in vivo. Alleles were considered loss-of- function (L) if they had dramatically reduced 3HC/COT in vivo. Specifically, decrease-of- function alleles were CYP2A6*9, *21, and the novel *25 allele. Loss-of-function alleles were

CYP2A6*17, *20, *23, and the novel *24, *26, *27, and *28 alleles. Individuals in the intermediate group had one D allele and the slow group had two D alleles, one or two L alleles, or the combination of one L and one D allele.

CYP2A6 Construct Heterologous Expression and Nicotine C-Oxidation Assay

Bicistronic constructs encoding hepatic NADPH-cytochrome P450 reductase (hNPR) and CYP2A6.1, .17, .24, .25, .26, .27 or .28 were expressed in Escherichia coli and membrane fractions were prepared as previously described (Pritchard et al. 1997; Ho et al. 2008). The amount of CYP2A6 protein in the membrane fractions was determined by immunoblotting with 80 a monoclonal CYP2A6 antibody (BD Biosciences, Mississauga, ON, Canada) using supersome expressed CYP2A6 (BD Biosciences) as a positive control for quantification as previously described (Schoedel et al. 2003; Ho et al. 2008). The function of the expression construct was confirmed by immunoblotting for hNPR. Nicotine C-oxidation was determined as previously described (Siu et al. 2006; Ho et al. 2008). Briefly, concentrations of 30 µM (~Km) and 300 µM

(~Vmax) of nicotine were used to screen for catalytic activity and the amount of cotinine formed was detected by HPLC.

Statistics Chi-square was used to assess Hardy-Weinberg equilibrium and Fisher’s exact test was used when expected cell size was less than or equal to five for the genotype data. The Mann-

Whitney U test was used to assess the difference between two groups for the in vitro kinetic data. For in vivo kinetic data, the 3HC/COT ratio and the nicotine AUC were log-transformed for statistical analyses. Univariate ANOVA was used to assess the equality of multiple group means and Bonferroni’s test was used for multiple-comparison corrections. All p-values are one-tailed. Because smoking status and gender alter CYP2A6 activity, they were included as covariates in statistical analyses (Dempsey et al. 2004; Benowitz et al. 2006a; Mwenifumbo et al. 2007a). In all figures the 3HC/COT ratio was adjusted by smoking status and gender means.

The nicotine AUC was adjusted by the smoking status mean. All analyses were done on SPSS software version 11.0.4 for Macintosh (SPSS Inc., Chicago, IL).

81 Results

Novel CYP2A6 Alleles Genotype assays were developed for four uncharacterized polymorphisms: 594G>C

(p.V110L), 1672T>C (p.F118L), 2162_2163GC>A (p.R203frameshift), and 5750G>C

(p.E419D). Eight novel alleles, CYP2A6*24A&B, *25, *26, *27, *28A&B, and *1B17, were found through the sequencing of individuals with at least one of the above-mentioned uncharacterized sequence variants (see Table 9A). Sequencing alleles with 594G>C (p.V110L) led to the discovery of two novel alleles, CYP2A6*24A&B. Both contain 594G>C (p.V110L),

6458A>T (p.N438Y), and the 58 bp CYP2A7 3’-UTR gene conversion, which is found in all

CYP2A6*1B alleles (see Table 9A). Of interest, one individual with the CYP2A6*24 allele had at least three copies of the CYP2A6 gene (CYP2A6*1/*24/*25; see Table 10). Sequencing alleles containing 1672T>C (p.F118L) resulted in the discovery of two novel alleles,

CYP2A6*25 and CYP2A6*26. CYP2A6*25 contains one nonsynonymous sequence variant

1672T>C (p.F118L) while CYP2A6*26 contains three nonsynonymous sequence variants

1672T>C (p.F118L), 1703G>T (p.R128L), and 1711T>G (p.S131A) (see Table 9A).

Sequencing an allele with 2162_2163GC>A (p.R203frameshift) lead to the discovery of the novel allele CYP2A6*27 that contains both 1672T>C (p.F118L) and 2162_2163GC>A

(p.R203frameshift) (see Table 9A). Alleles positive for 5750G>C (p.E419D) were sequenced and resulted in the finding of two novel alleles CYP2A6*28A&B. CYP2A6*28A consists of two nonsynonymous sequence variants, 5745A>G (p.N418D), 5750G>C (p.E419D), and the 58 bp

CYP2A7 gene conversion in the 3’-UTR. CYP2A6*28B is similar but also contains the

6960_6961insGAAAAG in the 3’-flanking region (see Table 9A). Sequencing an allele positive for the 6960_6961insGAAAAG lead to the discovery of another novel allele

CYP2A6*1B17 (see Table 9A). The frequencies of these novel alleles are reported in Table 11.

82 Table 9A Novel characterized alleles Amino Acid Novel Alleles Genomic Nucleotide Changes Change

51G>A; 1620T>C; 1646C>T; 2296C>T; 2483G>A; 2994T>C; 3788C>T; 4071delA; 4636A>C; 4850C>T; 4977C>T; 5668A; 6143C>A; 58 bp gene CYP2A6*1B17 conversion in the 3’-UTR; 6782C>G; 6832A>G; 6835C>A; 6935_6936insCACTT; 6960_6961insGAAAAG; 6989A>G; 7160A>G

-1301A>C; -1289G>A; -1013A>G; 51G; 578A>G; 594G>C; 720G>A; 1137C>G; 1620T>C; 2483G>A; 3225A>G; 5668A; 6218A>G; 6282A>G; V110L; CYP2A6*24A 6293T>C; 6354T>C; 6458A>T; 58 bp gene conversion in the 3’-UTR; N438Y 6782C>G; 7160A>G

-1301A>C; -1289G>A; -1013A>G; 51G; 578A>G; 594G>C; 720G>A; 1137C>G; 1381_1382CT>TC; 1481_1486delCTCTCT; 1620T>C; V110L; CYP2A6*24B 2483G>A; 3225A>G; 5668A; 6218A>G; 6282A>G; 6293T>C; 6354T>C; N438Y 6458A>T; 58 bp gene conversion in the 3’-UTR; 6782C>G; 7160A>G

-1301A>C; -1289G>A; -745A>G; 22C>T; 51G; 768A>T; 1620T>C; CYP2A6*25 1672T>C; 2296C>T; 2483G>A; 2605G>A; 2921G>A; 2994T>C; F118L 4636A>C; 5668A; 6586T>C; 6692C>G; 7160A>G -1301A>C; -1289G>A; -745A>G; 22C>T; 51G; 1165G>A; 1620T>C; F118L; 1672T>C; 1703G>T; 1710C>T; 1711T>G; 2296C>T; 2483G>A; CYP2A6*26 R128L; 2994T>C; 4071delA; 4636A>C; 5668A; 6115C>T; 6586T>C; 6692C>G; S131A 7160A>G

-1301A>C; -1289G>A; -745A>G; 22C>T; 51G; 1620T>C; 1672T>C; F118L; CYP2A6*27 2162_2163GC>A; 2296C>T; 2483G>A; 2994T>C; 3872G>A; 4071delA; R203FS 4636A>C; 5668A; 5857T>A; 6586T>C; 6692C>G; 7160A>G

-1269T>C; 51G>A; 656G>T; 1620T>C; 4681T>G; 5668A; 5738C>T; N418D; CYP2A6*28A 5745A>G; 5750G>C; 6354T>C; 6361C>A; 6385G>T; 6389C>G; E419D 6390T>C; 58 bp gene conversion in the 3’-UTR; 6782C>G; 7160A>G -1269T>C; 51G>A; 656G>T; 1381_1382CT>TC; 1481_1486delCTCTCT; 1620T>C; 4681T>G; 5668A; 5738C>T; 5745A>G; 5750G>C; 6354T>C; N418D; CYP2A6*28B 6361C>A; 58 bp gene conversion in the 3’-UTR; 6960_6961insGAAAAG; E419D 7160A>G DNA numbering is relative to +1 ATG start site on the reference genomic sequence NG_000008.7 as is common for this gene (51G and 5668A are considered wildtype). Protein numbering is relative to the NP_000753.2 reference protein sequence. All that are bolded are found in exons. Sequencing gaps in noncoding regions may exist.

83 Table 9B Novel uncharacterized alleles Amino Acid Novel Alleles Genomic Nucleotide Changes Change -1301A>G; -1289G>A; -1199_-1198ins316bpAluYa5; 51G; 431T>G; CYP2A6*1K 1620T>C; 1779G>A; 2483G>A; 2994T>C; 4074delA; 5668A; 6354T>C; 6952G>A; 7160A>G

-1399A>G; -1394T>C; -933G>A; -911C>T; -908_- F61I; C64R; 907insGTTCTTTGTAAATTA; -834C>A; 447T>C; 460T>A; 469T>C; A117V; 496T>C; 600C>T; 810G>C; 816C>T; 875C>T; 1494G>A; 1532C>A; L128R; 1569C>G; 1572G>C; 1589C>T; 1622T>G; 1629_1630TG>CT; 1696T>G; A131S; 1834C>G; 1907T>C; 1980T>G; 2282G>C; 2316A>C; 2342T>C; 2348A>G; CYP2A6*4E† S153A; 2382G>A; 2489G>A; 2583C>A; 2586C>T; 2682A>C; 2781G>C; 2984C>T; D169E; 3001C>T; 3441A>G; 3479G>A; 3730_3731GA>CC; 4014C>A; 4032A>G; H274R; 4317C>T; 4599C>G; 4615T>C; 4738G>T; 4772A>G; 5451C>T; 5461A>C; R311C; 5463T>C; 5471C>T; 5499delC; 5504G>T; 5506C>A; 5509_5510insG; V479G 6276C>G; 6469T>G; 6668C>G; 7160A>G

-1399A>G; -1394T>C; -1242G>C; -933G>A; -908_- 907insGTTCTTTGTAAATTA; -911C>T; -834C>A; -823C>T; 447T>C; 726delA; 2342T>C; 2348A>G; 2619T>C; 3166A>C; 3441A>G; 3538G>C; H274R CYP2A6*4F† 3730_3731GA>CC; 4032A>G; 4599C>G; 4738G>T; 4772A>G; 5451C>T; V479G 5461A>C; 5463T>C; 5471C>T; 5499delC; 5504G>T; 5506C>A; 5509_5510insG; 6469T>G; 6668C>G; 7160A>G; 7658T>C

-1289G>A; -1013A>G; -492_- 470delCCCCTTCCTGAGACCCTTAACCCinsAATCCATATGTGGAATCT CYP2A6*31A M6L G; 16A>C; 51G; 1339C>G; 1620T>C; 2721G>A; 2994T>C; 3315C>T; 3255A>G; 3315C>T; 5668A; 6692C>G; 7160A>G; 7568C>T

-1289G>A; -1013A>G; -492_- 470delCCCCTTCCTGAGACCCTTAACCCinsAATCCATATGTGGAATCT CYP2A6*31B M6L G; -975T>C; 16A>C; 51G; 467C>T; 1339C>G; 1620T>C; 2721G>A; 2994T>C; 3225A>G; 3315C>T; 4074delA; 5668A; 6692C>G; 7160A>G DNA numbering is relative to +1 ATG start site on the reference genomic sequence NG_000008.7 as is common for this gene (51G and 5668A are considered wildtype). Protein numbering is relative to the CYP2A6 NP_000753.2 reference protein sequence. † Numbering is relative to +1 ATG start site on the CYP2A7 reference genomic sequence NG_000008.7 and protein numbering is relative to the NP_000755.2 reference protein sequence. All that are bolded are found in exons. All that are italicized have numbering based on the CYP2A6 reference genomic sequence in the CYP2A7/CYP2A6 hybrid allele. Sequencing gaps in noncoding regions may exist.

84 Table 10 CYP2A6 genotypes and their mean in vivo adjusted 3HC/COT Allele Genotype N Mean 3HC/COT SD % Mean F (P) CYP2A6*1 *1A/*1A 96 1.17 0.63 100 *1A/*1B 50 1.22 0.64 104 0.6 (0.29) *1B/*1B 15 1.15 0.71 98 CYP2A6*9 *1/*1 161 1.19 0.64 100 *1/*9 29 0.76 0.44 65 8.6 (<0.001) *9/*9 2 0.75 0.32 64 CYP2A6*14 *1/*1 161 1.19 0.64 100 *1/*14 3 0.89 0.30 75 0.1 (0.36) *14/*14† 1 2.11 - 177 CYP2A6*17 *1/*1 161 1.19 0.64 100 *1/*17 23 0.57 0.44 48 19.1 (<0.001) *17/*17 4 0.15 0.11 13 CYP2A6*20 *1/*1 161 1.19 0.64 100 6.5 (0.01) *1/*20 3 0.35 0.35 29 CYP2A6*21 *1/*1 161 1.19 0.64 100 *1/*21 2 0.76 0.15 64 3.8 (0.03) *21/*21† 1 0.26 - 22 CYP2A6*23 *1/*1 161 1.19 0.64 100 *1/*23 4 0.76 0.54 64 3.0 (0.04) *23/*23† 1 0.00 - 0 CYP2A6*24 *1/*1 161 1.19 0.64 100 3.7 (0.03) *1/*24 5 0.60 0.23 50 CYP2A6*25 161 1.19 0.64 100 *1/*1 0.7 (0.20) *1/*25 2 0.89 0.41 75 CYP2A6*26 161 1.19 0.64 100 *1/*1 4.1 (0.02) *1/*26 4 0.66 0.60 55 CYP2A6*28 *1/*1 161 1.19 0.64 100 2.7 (0.05) *1/*28 4 0.55 0.06 46 Two or more *9/*17 3 0.25 0.43 21 different variant *9/*23 1 1.04 - 87 alleles *17/*23 2 0.26 0.37 22 *17/*27 1 0.11 - 9 *20/*23 2 0.55 0.51 46 *20/*28 1 0.22 - 18 *1/*24/*25‡ 1 0.79 - 66 Genotypes for each allele were found to be in Hardy-Weinberg equilibrium. F represents the F statistic, which is a value used in ANOVA. † If the homozygote group contained N = 1 then it was combined with the heterozygote group for statistical analyses. ‡ This individual was found, through sequencing, to have at least three CYP2A6 genes.

85

Table 11 CYP2A6 allele frequencies Alleles N = Total alleles N = (frequency %) CYP2A6*1B† 110 (19.8) 556 CYP2A6*9 40 (7.2) 558 CYP2A6*12 0 (0) 560 CYP2A6*14 5 (0.9) 560 CYP2A6*15 0 (0) 560 CYP2A6*16 0 (0) 560 CYP2A6*17 41 (7.3) 560 CYP2A6*20 6 (1.1) 560 CYP2A6*21 4 (0.7) 560 CYP2A6*23 11 (2.0) 560 CYP2A6*24 7 (1.3) 560 CYP2A6*25 3 (0.5) 560 CYP2A6*26 4 (0.7) 560 CYP2A6*27 1 (0.0) 560 CYP2A6*28 5 (0.9) 560 † Four CYP2A6*1B17 were found and are included in the CYP2A6*1B group

86 There was no difference in the in vivo mean 3HC/COT between the three major wild type genotypes CYP2A6*1A/*1A, *1A/*1B, and *1B/*1B, so they were combined as a

CYP2A6*1/*1 reference group (see Table 10). CYP2A6*1/*24 heterozygotes had 50% activity, as assessed by the 3HC/COT ratio, CYP2A6*1/*25 heterozygotes had 75% activity, and

CYP2A6*1/*26 heterozygotes had 55% activity (see Table 10; Figure 10). The individual with

CYP2A6*27 also had CYP2A6*17 and consequently had very poor activity. CYP2A6*1/*28 heterozygotes had 46% activity, while the CYP2A6*20/*28 individual had only 18% in vivo activity.

87

Figure 10. 3HC/COT varies with CYP2A6 genotype. Open circles represent individual data and lines denote group means

88 Established CYP2A6 Alleles Frequencies of established CYP2A6 alleles are reported in Table 11. CYP2A6*9, *17,

*20, *21, and *23 were all associated with lower in vivo CYP2A6 activity (see Table 10;

Figure 10). CYP2A6*17 exhibited the predicted gene-dose effect (see Table 10).

CYP2A6*1/*14 heterozygotes did not have significantly different CYP2A6 activity compared to the CYP2A6*1/*1 group (see Table 10; Figure 10). CYP2A6*12, *15, and *16 were not detected in this population (see Table 11); however, this was not unexpected given their low frequencies in other populations of black African descent (Schoedel et al. 2004; Nakajima et al.

2006). Some individuals with established CYP2A6 alleles, particularly those that have been discovered and characterized in Caucasian and Asian populations, had unexpected in vivo pharmacokinetics; thus, sequencing of both alleles from these individuals (two with the genotype CYP2A6*1/*2, four CYP2A6*1/*4, one CYP2A6*4/*9, one CYP2A6*1/*9, and one

CYP2A6*1/*17) was performed. Several novel alleles were identified through this sequencing approach including, CYP2A6*1K, CYP2A6*31A&B (see Table 9B). Substantial sequence variation was found in the five individuals with CYP2A6*4 deletion alleles; we identified the novel CYP2A6*4E and CYP2A6*4F alleles that had extensive sequence variation in both the noncoding and coding sequence (Table 9B). What is more, two CYP2A6*1/*4 heterozygotes with higher activity both had at least three alleles representing novel duplication alleles. These novel alleles will be characterized in future studies.

CYP2A6 Genotype as a Predictor of CYP2A6 Activity and Nicotine Metabolism

Due to the large number of low prevalence alleles, the utility of grouping genotypes for gene association studies was examined, as we have done previously for CYP2A6 in European

Americans (Benowitz et al. 2006) and as is common for CYP genes (Gaedigk et al. 2008). Based on CYP2A6 genotype and in vivo 3HC/COT phenotype relationships, individuals were grouped in normal, intermediate, or slow nicotine metabolism groups. The three groups were associated

89 with altered CYP2A6 activity, as assessed by 3HC/COT (F=38.1, degrees of freedom [df]=2,

P<0.001) (see Figure 11A). The normal, intermediate, and slow metabolism groups had mean in vivo CYP2A6 activities of 0.28±0.18 (100%), 0.18±0.09 (64%), and 0.11±0.09 (40%), respectively. Moreover, using a difference measure of nicotine metabolism capacity, the three groups were associated with the extent of the body’s exposure to nicotine, as assessed by the nicotine AUC (F=15.6, df=2, P<0.001) (see Figure 11B). The normal, intermediate, and slow metabolism groups had mean nicotine AUCs of 1,381±1,115 (100%), 1,921±1,169 (139%), and

2,231±1,421 (162%) ng/ml/min, respectively.

90

Figure 11. Normal, intermediate, and slow metabolism groups were associated with in vivo CYP2A6 activity and the extent of the systemic exposure to oral nicotine. Individuals were grouped in either the normal (white bars) (N=165), intermediate (grey bars) (N=33), or slow (black bars) (N=61) metabolism groups. A In a univariate ANOVA model, in which smoking status and gender were covariates, the normal metabolism group had higher mean 3HC/COT compared to the intermediate (P<0.001) and slow (P<0.001) metabolism groups. B The normal metabolism group had lower nicotine AUC compared to the intermediate (P<0.01) and slow (P<0.001) groups in a univariate ANOVA model in which smoking status was a covariate. **P<0.01 and ***P<0.001 compared to normal metabolism group.

91 Novel Variants In Vitro Nicotine Catalytic Efficiency

To determine the functional impact of the nonsynonymous sequence variants found in the novel characterized alleles, the levels of expressed CYP2A6 and resulting nicotine C- oxidation catalytic efficiency of the variant proteins were assessed. Neither CYP2A6.26

(p.F118L, p.R128L, and p.S131A) nor CYP2A6.27 (p.F118L and p.R203frameshift) made appreciable levels of cotinine, while clearly expressing hNPR from the bicistronic construct (see

Figure 12). Compared to wild type (CYP2A6.1) the rate of cotinine formation per picomole of expressed CYP2A6 at ~Vmax was 92% for CYP2A6.24 (p.V110L and p.N438Y), 90% for

CYP2A6.25 (p.F118L), and 91% for CYP2A6.28 (p.E419D and p.N418D) (see Figure 12).

92

Figure 12 Levels of CYP2A6 and in vitro nicotine C-oxidation catalytic activity for heterologously expressed CYP2A6.24, .25, .26, .27, and .28 compared to wild type CYP2A6.1. A Immunoblots showing both hNPR and CYP2A6 expression from the bicistronic expression constructs. CYP2A6.27 (the predicted truncated protein) was not detected while the other constructs clearly expressed CYP2A6 protein1. B Per picomole of expressed immunodetected CYP2A6, neither CYP2A6.26 nor .27 formed appreciable amounts of cotinine at either 30 µM (~Km) or 300 µM (~Vmax), while CYP2A6.24, .25, and .28 were not significantly different from CYP2A6.1. *P<0.05.

1 Different lanes represent amounts of total membrane protein loaded: 4.5, 3.0, 1.5, & 1.0 µg for CYP2A6.1 to CYP2A6. 27 and 2.25, 1.5, 0.75, & 0.5 µg for CYP2A6.28. 93 CYP2A6*28 Confounds Some Traditional Genotyping Assays

CYP2A6*28 confounds some traditional genotyping assays. Specifically, the

CYP2A6*28A allele can result in the overestimation of CYP2A6*4A and *4D, which are

CYP2A7/CYP2A6 hybrid deletion alleles. Figure 13 illustrates how this can occur. Figure 13A is a schematic of CYP2A6*4D, in which the entire 5’ region is CYP2A7 until midway through exon 9, where CYP2A6 sequence begins. In the original PCR-based CYP2A6*4A&D assay

(Oscarson et al. 1999a), the first step amplifies both the wild type CYP2A6 and the CYP2A6*4A or *4D deletion alleles (see Figure 13A). The original second step primer 2A7ex8F was intended to specifically identify the CYP2A6*4A or *4D deletion alleles by annealing to

CYP2A7 sequence in exon 8 (see Figure 13A). However, the CYP2A6*28 variant nucleotides

5738T, 5745G, and 5750C are the wild type nucleotides in CYP2A7 and are located where the

2A7ex8F primer anneals (see Figure 13B). Therefore, CYP2A6*28A results in a false-positive amplification for the CYP2A6*4A and *4D alleles (see Figure 13B). In this population, using the original CYP2A6*4A&D assay, eight individuals with the CYP2A6*1/*4A&D genotype were detected. Two of the eight were actually CYP2A6*1/*28A, meaning initially there was a 33% overestimation of the CYP2A6*4A and *4D alleles. We have redesigned the CYP2A6*4A&D assay as seen in Tables 5S, 6S, 7S, and 8S.

As illustrated in Figure 13C, the CYP2A6*28B allele also confounds any original genotyping assays, such as CYP2A6*1B, *5, *7, *8, and *10, that use the 2A6R2 primer. A

CYP2A7 gene conversion in the 3’-flanking region of CYP2A6, found in CYP2A6*28B, results in DNA variation where the 2A6R2 primer is designed to selectively anneal to CYP2A6 sequence. This DNA variation results in amplification failure when using this particular reverse primer. Of note, CYP2A6*1B17 also has this 3’-flanking region variation (i.e.

6960_6961insGAAAAG). Consequently, the frequency of one allele is underestimated while the allele on the opposing chromosome is overestimated. 94

Figure 13 CYP2A6*28A confounds the traditional CYP2A6*4A&D assay. A This schematic shows the CYP2A6*4D deletion, CYP2A7/CYP2A6 hybrid allele; the 50 region is composed of the CYP2A7 sequence until midway through exon 9, where it becomes the CYP2A6 sequence. B CYP2A6*28A has the 5738C>T, 5745A>G, 5750G>C sequence variants in exon 8 (black bars). C CYP2A6*28B also has the 6960_6961insGAAAAG CYP2A7 gene conversion in the noncoding 3’-flanking region (gray box). Arrows represent primers. The dashed line between two primers represents the region of the gene that is PCR amplified if both primers successfully anneal to the genomic DNA.

95 Discussion

In this population of black African descent, 7% had at least one of the novel alleles characterized here and 31% had at least one established allele (excluding CYP2A6*1B). Our data demonstrate the importance of determining haplotypes for uncharacterized CYP2A6 nonsynonymous sequence variants and highlights the usefulness of reassessing both the genotyping assays and the in vivo functional impact of variant alleles in distinct ethnoracial groups. Moreover, through sequencing, 26 new CYP2A6 DNA sequence variations (see

Supplementary Table 12S) and 13 novel alleles, including two novel deletion hybrid

CYP2A7/CYP2A6 alleles, were described (see Table 9A and B).

The novel CYP2A6*24 allele had two nonsynonymous sequence variants, Val110, which is located within substrate recognition site one and may alter the active site cavity, and Asn438, which is adjacent to the heme binding amino acids Arg437 and Cys439 (Lewis 2003). The novel

CYP2A6*25 allele had one nonsynonymous sequence variant, Phe118, which is predicted to be involved in stabilizing the compact structure of the enzyme and formation of the active site

(Yano et al. 2005). The novel CYP2A6*26 allele had three nonsynonymous sequence variants; of these, Arg128 has been demonstrated to be important in maintaining the holoenzyme’s ordered structure (Kitagawa et al. 2001). The novel CYP2A6*27 allele had two nonsynonymous sequence variants, one of which, 2162_2163GC>A, causes a frameshift and introduces a stop codon in exon five. The novel CYP2A6*28 allele had two nonsynonymous sequence variants.

96 Supplementary Table 12S DNA sequence variations in the CYP2A6 gene detected through sequencing Predicted Nucleotide change Location Flanking Sequence dbSNP ID Reference effect -1301A>C 5’-flanking TTCTCCACTC*CACCCACCCC rs7260629 (Haberl et al. 2005) -1289G>A 5’-flanking ACCCACCCCA*GATCTGCCTC rs7259706 (Haberl et al. 2005) -1269T>C 5’-flanking CTGTGCCTAA*ACTGGAGTTT -1199_-1198ins316bpAluYa5 5’-flanking CCTGACAAAG*CAGGAATCAT -1126C>A 5’-flanking TCCCTTTCCA*TGGCAGGAAA (Haberl et al. 2005) -1013A>G 5’-flanking TGTCCTCTGT*GATCTTTATA (Pitarque et al. 2004) -975T>C 5’-flanking ACAACAATCA*AATATTAATA -745A>G 5’-flanking CCCCTTCCCA*TCAGAGATGG rs2892625 (von Richter et al. 2004) -686A>G 5’-flanking TGAGGTTCCA*TGAGGATTCT rs3822484 -492_-470del CCCCTTCCTGAGACCCTTA GAATC* ACCCinsAATCCATATGTGG AATCCATATGTGGAATCTG AATCTG 5’-flanking *TGCAT -48T>G 5’-flanking TCAGGCAGTA*AAAGGCAAAC rs28399433 (Pitarque et al. 2001) 16A>C Exon 1 GGCCTCAGGG*TGCTTCTGGT M6L (Haberl et al. 2005) 22C>T Exon 1 AGGGATGCTT*TGGTGGCCTT rs8192720 (Kiyotani et al. 2002) 51G>A Exon 1 GCCTGACTGT*ATGGTCTTGA rs1137115 (Kiyotani et al. 2002) 209C>T Intron 1 ATGGGTGGCA*GGGGTGGGGG rs8192722 (Saito et al. 2003) 413T>G Intron 1 AGTGGAGGCT*CTCCCTCTAA rs28399438 (Solus et al. 2004) 431C>T Intron 1 TAACCACTCC*ACCCACCTCC 467C>T Exon 2 GCTATGGCCC*GTGTTCACCA (Haberl et al. 2005) 578A>G Exon 2 GAGGCGAGCA*GCCACCTTCG (Haberl et al. 2005) 594G>C Exon 2 CTTCGACTGG*TCTTCAAAGG V110L (Haberl et al. 2005) (Solus et al. 2004; Haberl et 618delG Intron 2 TGGTGAGGGG*TGCCCAAGAG rs28399439 al. 2005) 656G>T Intron 2 GGACACGAAG*TCTCAGTGTT 720G>A Intron 2 AGAGTCCCCA*TCTGGTCTTC rs56145769 768A>T Intron 2 CTCCATGTGT*TCCCTCACCT 1137C>G Intron 2 GCTTAAGAAT*TTTCACCATT rs7250713 1165G>A Intron 2 CCTCCTCCCA*ATCTCCCCAT

97 Supplementary Table 12S DNA sequence variations in the CYP2A6 gene detected through sequencing 1339C>G Intron 2 TAAACCTGGT*TCTCTCTCTC 1381_1382CT>TC Intron 2 CTCTCTCTC**CTCTCTCTTC (Mwenifumbo et al. 2008a) 1481_1486delCTCTCT Intron 2 TCTCTCT******ACCTCGAC (Mwenifumbo et al. 2008a) 1620T>C Intron 2 CTCCTGCCCC*GCCGCCCCCT rs8192725 (Kiyotani et al. 2002) 1646C>T Intron 2 GTCTCCATTC*CGCGTTCACC 1672T>C Exon 3 AGGCGTGGTA*TCAGCAACGG F118L rs28399440 (Solus et al. 2004) 1703G>T Exon 3 AAGCAGCTCC*GCGCTTCTCC R128L rs58116892 (Haberl et al. 2005) 1710C>T Exon 3 TCCGGCGCTT*TCCATCGCCA rs3815709 (Haberl et al. 2005) 1711T>G Exon 3 CCGGCGCTTC*CCATCGCCAC S131A rs3815708 (Haberl et al. 2005) 1726C>A Exon 3 CGCCACCCTG*GGGACTTCGG rs4986892 (Haberl et al. 2005) 1779G>A Exon 3 AGGAGGAGGC*GGCTTCCTCA rs28399441 (Solus et al. 2004) 1799T>A Exon 3 ATCGACGCCC*CCGGGGCACT L160H rs1801272 (Yamano et al. 1990) 1836G>T Intron 3 CCCCGAGTGC*GGGGCAGGAG rs8192726 (Kiyotani et al. 2002) 2162_2163GC>A Exon 4 CACTGTTGC**ATGATGCTAG 203FS rs28399445 (Solus et al. 2004) 2296C>T Intron 4 AAACTCCCAC*GCCCTCCAGA (Haberl et al. 2005) 2483G>A Intron 4 ACCTGGGCAC*TGTTCCCATC rs2388868 2605G>A Intron 4 AACACCTGGT*CTGCAAAATC 2721G>A Intron 4 CTGAAAACAT*GACAGCTGCC 2921G>A Intron 4 TAATACCTGA*CACCTGAACA rs56100930 2994T>C Intron 4 CCTACTTGAA*GCCTAAATAC rs59796550 3225A>G Intron 4 ACTTAGATAT*AGTTCCTATC rs60845437 3315C>T Intron 4 AATCCCATTC*CATCAGCTCC (Haberl et al. 2005) 3378C>T Exon 5 AACCCCAGCT*TATGAGATGT 3788C>T Intron 5 GGTCAAACCC*GTCTCTACTA 3872G>A Intron 5 TCACGGCATT*CACTCCAGTC 3904G>A Intron 5 ATGAGGCCCT*TGTCAAAAAA 4071delA Intron 5 AAGAAAAAAA*AAACAAAAAA 4074delA Intron 5 AAAAAAAAAA*CAAAAAAAAA 4489C>T Intron 6 TGGAGGGGGA*GGAAGTGGAG (Kiyotani et al. 2002) 4636A>C Intron 6 GTCTCATTAG*TATTAAAATA (Solus et al. 2004) 4681T>G Intron 6 CAGTCACTTC*GTCCCAAGCC rs56048590

98 Supplementary Table 12S DNA sequence variations in the CYP2A6 gene detected through sequencing 4718C>T Intron 6 CCCGTTCCAC*GGGTCATCCC rs55971367 4850C>T Intron 6 GTCTCATAGG*GGAGCCATAT rs28399450 4977C>T Exon 7 ACAGAGTGAT*GGCAAGAACC 5065G>A Exon 7 ATTTGGAGAC*TGATCCCCAT V365M rs28399454 (Solus et al. 2004) 5668A>T Exon 8 ACCGAAGTGT*CCCTATGCTG Y392F rs1809810 5684T>C Exon 8 TGCTGGGCTC*GTGCTGAGAG rs28399461 (Solus et al. 2004) 5738C>T Exon 8 ATCCCCAGCA*TTCCTGAATG rs2002977 (Solus et al. 2004) 5745A>G Exon 8 GCACTTCCTG*ATGAGAAGGG N418D rs28399463 (Solus et al. 2004) 5750G>C Exon 8 TCCTGAATGA*AAGGGGCAGT E419D rs8192730 (Saito et al. 2003) 5825A>G Intron 8 CTGCCAGGCC*CGGCTCACAC rs58416032 (Nakajima et al. 2004) 5857T>G Intron 8 TCCCTCACCC*CCTCCCCTCT 6115C>T Intron 8 ACATCTAAAC*TCCCATTGCT 6143C>A Intron 8 GCATGGATCA*CCCATCTATG 6218A>G Intron 8 GTAAACCCTA*TGTAAACTAT 6282A>G Intron 8 TACATCTCTT*TAGAAAGAAA (Solus et al. 2004) 6293T>C Intron 8 TAGAAAGAAA*TGAGGCTCAG rs2431414 (Haberl et al. 2005) 6354T>C Intron 8 TATTCCACCC*TCCTCCCTGG rs2431413 (Kiyotani et al. 2002) 6361C>A Intron 8 CCCTTCCTCC*TGGGAGAGCC rs28399464 (Solus et al. 2004) 6385G>T Intron 8 GCTGGAGGTC*GTACTGGGGC 6389C>G Intron 8 GAGGTCGGTA*TGGGGCGAGG rs28399465 (Solus et al. 2004) 6390T>C Intron 8 AGGTCGGTAC*GGGGCGAGGC (Haberl et al. 2005) 6458A>T Exon 9 AGGAAAGCGG*ACTGTTTCGG N438Y 6586T>C Exon 9 ACGTGGGCTT*GCCACGATCC (Haberl et al. 2005) 6692C>G 3’-UTR GGGCCAAGAC*GGGCTTGGGA rs7248240 (Kiyotani et al. 2002) 6782C>G 3’-UTR GAAACAGAAG*GGCTCAGTTC rs8192733 (Kiyotani et al. 2002) 6832A>G 3’-UTR GAGAGGAAGG*AACCCTTACA rs8192734 (Kiyotani et al. 2002) 6835C>A 3’-UTR AGGAAGGAAA*CCTTACATTA rs8192735 (Kiyotani et al. 2002) 6935_6936insCACTT 3’-flanking GCTCACCT*****AATTGCCA (Mwenifumbo et al. 2008a) 6960_6961insGAAAAG 3’-flanking CGAAGGG******CGTTCATG (Mwenifumbo et al. 2008a) 6989A>G 3’-flanking ACGTGACAAA*CTGAGGCTTA (Mwenifumbo et al. 2008a) 7073T>C 3’-flanking GCCCATAGCC*TCTAGACAGA rs2431412

99 Supplementary Table 12S DNA sequence variations in the CYP2A6 gene detected through sequencing 7082G>C 3’-flanking CTTCTAGACA*ATTCTTAAAA rs2259219 7094G>C 3’-flanking TTCTTAAAAA*CACCTATTCC rs2259218 7099T>C 3’-flanking AAAAAGCACC*ATTCCTCACG rs2259217 7160A>G 3’-flanking TGTCCTGGGG*GTTTTCCAGA rs28742185 7568C>T 3’-flanking ACTATGGGTT*TGAAGGCTGG Does not include the 58 bp gene conversion in the 3’-UTR (i.e. CYP2A6*1B) DNA numbering is relative to +1 ATG start site on the reference genomic sequence NG_000008.7 as is common for this gene (51G and 5668A are considered wildtype). Protein numbering is relative to the CYP2A6 NP_000753.2 reference protein sequence. Those bolded have not, to the authors’ knowledge, previously been reported in the literature. Data is from CYP2A6 alleles from 17 individuals. Sequencing gaps in noncoding regions may exist.

100 In vivo 3HC/COT data suggest that the novel alleles, CYP2A6*24, *26, *27, and *28, are likely to be loss-of-enzyme function alleles, while CYP2A6*25 results in a decrease-of- enzyme function. In vitro data suggests that CYP2A6.26 and CYP2A6.27 result in a complete loss of catalytic activity, while CYP2A6.24, .25, and .28 appear to have nicotine C-oxidation catalytic activity similar to the wildtype. CYP2A6.27 is predicted to be a truncated form of the enzyme. It was not detected by Western blotting, while hNPR from the same bicistronic construct was. This indicates that the expression system was functional and suggests that the truncated protein was either degraded or that no CYP2A6 antibody epitopes were present.

There are several possible explanations for the apparent discordance between in vivo and in vitro data for CYP2A6*24, *25, and *28. First, our in vitro assay is designed to test differences in the catalytic efficiency of nicotine C-oxidation per unit of CYP2A6 detected and not the stability of the holoenzyme. Second, some of the novel alleles contain 5’-flanking region sequence variants that have been demonstrated to result in decreased expression in a reporter construct system. Specifically, -1013A>G disrupts a putative enhancer region (Pitarque et al.

2004), and -745A>G disrupts a CCAAT box (von Richter et al. 2004). Both of these sequence variants may result in lower in vivo protein level and thus lower in vivo 3HC/COT ratio.

Additionally, it is possible that these amino acid changes in these three alleles may alter substrate metabolism selectively, such that COT hydroxylation to 3’-hydroxycotinine metabolism is impaired while nicotine C-oxidation is unchanged. The possibility that some of the amino acid changes might not alter enzymatic function should also be considereda. For example, our data suggest that the amino acid substitution F118L (CYP2A6.25) does not alter the rate of cotinine formation in vitroa. Nonetheless, in vivo, individuals heterozygous for

F118L (i.e. CYP2A6*1/*25) trend towards having lower (~50%) CYP2A6 activity (Ho et al.

a Sentences added following publication during thesis correction. 101 2009)a. Factors that might contribute to the discrepancies between the in vitro and in vivo results is discussed abovea.

CYP2A6*1B contains a 58 bp CYP2A7 gene conversion in the 3’-UTR and is associated with greater CYP2A6 activity and nicotine clearance in European Americans (Mwenifumbo et al. 2008a). However, in this population of black African descent, the CYP2A6*1B allele was not associated with higher activity. One likely possibility is that unidentified decrease-of-function polymorphisms exist in linkage with the CYP2A6*1B allele confounding interpretation; this is currently under investigation. The established alleles CYP2A6*9, *17, *20, *21, and *23 all resulted in decrease- or loss-of-function CYP2A6 activity. CYP2A6*9 contains a T>G sequence variant in the TATA box that decreases transcription by approximately 50% in vitro (Pitarque et al. 2001). CYP2A6*17 results in an amino acid change p.V365M in exon 7. In vitro (Fukami et al. 2004) and in vivo (Nakajima et al. 2006), this allele causes impaired nicotine metabolism and our study confirms these findings. CYP2A6*20 results in a p.K196frameshift that to a stop codon, and thus no functional protein is produced (Fukami et al. 2005a). The current study identified six individuals with the CYP2A6*20 allele and demonstrated substantially lower

CYP2A6 activity, supporting the predicted loss-of-function. CYP2A6*23 was first discovered in this population of black African descent (Ho et al. 2008); it results in an amino acid change p.R203C in exon 4 that impairs both in vitro and in vivo CYP2A6 activity.

Among individuals with established CYP2A6 alleles but unexpected in vivo 3HC/COT, we report five novel alleles that remain to be characterized with respect to frequency and functional impact. The CYP2A6*1K allele contains a 316 bp AluYa5 repeat inserted into the 5’- flanking region of the gene. Alu elements can potentially function in regulation of gene expression (Hasler et al. 2007). The CYP2A6*31 allele contains a nonsynonymous 16A>C

(p.M6L) sequence variant in the membrane-anchoring region in exon 1 (Haberl et al. 2005); this a Sentences added following publication during thesis correction. 102 specific nucleotide change has a frequency of 1.2% in a population of black African descent

(Haberl et al. 2005). CYP2A6*4 is a CYP2A7/CYP2A6 hybrid allele in which most of the

CYP2A6 gene is deleted (see Figure 13A) and it is thought to result in a complete loss-of- function due to a lack of functional enzyme production (Yoshida et al. 2002). CYP2A6 and

CYP2A7 are homologous genes and the proteins produced share 94% amino acid identity.

CYP2A7 does not incorporate a heme molecule (Ding et al. 1995) but there is considerable genetic variation in CYP2A7 and consequently it is possible that polymorphisms could result in an active CYP2A7 gene or an active CYP2A6*4 (CYP2A7/CYP2A6 hybrid) allele.

CYP2A6*4E&F are examples of the heterogeneity found among the sequenced deletion alleles.

In addition, two of the five individuals with deletion alleles sequenced had at least three copies of the CYP2A6 gene. In total, three individuals with more than two copies of the CYP2A6 gene were discovered; these alleles were not the known duplication alleles CYP2A6*1X2A or

CYP2A6*1X2B (Rao et al. 2000; Fukami et al. 2007), suggesting that novel duplication alleles exist.

As illustrated in Figure 10, interesting kinetic outliers, with both exceptionally slow and fast CYP2A6 activity, were observed in this population. This is most likely due to a combination of several genetic and nongenetic factors. As convention, the wild type CYP2A6*1 allele is assigned based on the absence of genotyped variants, but it may contain as yet unidentified variants. We believe that much of the variation in the CYP2A6*1/*1 group is due to additional genetic variation in this population, some of which has been identified here, but has not yet been characterized. In addition, several other factors contribute to the interindividual differences in CYP2A6 levels. For instance, levels of the constitutive androstane receptor and

HNF-4α have been correlated with CYP2A6 mRNA levels (Wortham et al. 2007). Differential exposure to CYP2A6 inducers (e.g. rifampicin and cadmium) (Pichard-Garcia et al. 2000;

Satarug et al. 2004) or inhibitors (e.g. dietary constituents such as broccoli) (Hakooz and 103 Hamdan 2007) may also modulate CYP2A6 levels. Women have faster nicotine clearance

(Benowitz et al. 2006a). Smoking results in slower nicotine metabolic clearance (Benowitz and

Jacob 2000). All of the above-mentioned factors likely contribute to the large interindividual variation in nicotine metabolic activity within CYP2A6 genotype groups.

Given the large contribution of CYP2A6 to the metabolism of nicotine, higher CYP2A6 activity (3HC/COT) results in increased first-pass nicotine metabolism, as well as faster nicotine clearance, and is reflected in a smaller systemic exposure to oral nicotine (nicotine

AUC) (Mwenifumbo et al. 2007a). Here, we confirmed that normal, intermediate, and slow nicotine metabolism groups are associated with both CYP2A6 activity and the systemic exposure to oral nicotine. Confirmation of the genotypic impact on phenotypic nicotine metabolism is valuable because many studies examine the association between CYP2A6 genotypes and behavioural, disease, or pharmacological phenotypes. Clearly, these groupings require further refinement by identifying and characterizing additional novel CYP2A6 genetic variants.

Gene conversions, particularly between CYP2A6 and CYP2A7 (which are 96% identical in genomic nucleotide sequence) are proving to be a major challenge to gene- and allele- specific genotyping. In this report, a 33% overestimation of the frequency of CYP2A6*4A&D due to CYP2A6*28A was found. This may be of particular importance in Asian populations because the frequency of CYP2A6*4 is much higher (Schoedel et al. 2004), the frequency of

CYP2A6*28A is unknown, and the 5750G>C sequence variant (found in CYP2A6*28A) was first identified in a Japanese population (Saito et al. 2003).

Discovering novel polymorphisms in populations of black African descent may aid in explaining their paradoxical light smoking and high disease risk. More generally, with the advent of nicotine-replacement therapies for long-term maintenance against tobacco dependence and for treatment of other disorders such as Alzheimer disease, Parkinson’s 104 disease, Tourette’s syndrome, and ulcerative colitis, CYP2A6 genotype may become particularly important, as it would significantly affect nicotine plasma levels from these sources.

105 Acknowledgments

J.C.M. receives funding from CIHR-funded SPICE and TUSP awards. N.K. receives funding from CHIR-funded STPTR and TUSP awards. M.K.H. receives funding from a Natural

Sciences and Research Council of Canada–Canadian Graduate Scholorship–

Doctoral (NSERC-CGS-D) award. R.F.T. holds a Canada Research Chair in Pharmacogenetics.

We also thank Dr. Howard Kaplan for valuable guidance in the statistical analyses, Frankie Lee and Abbas Assadzadeh for their technical excellence in genotyping, Elizabeth Gillam for her generous gift of the CYP2A6 construct, and Eric Siu for his careful review of the manuscript.

Drs. R.F. Tyndale and E.M. Sellers hold shares in Nicogen Research Inc., a company focused on novel smoking cessation treatment approaches. None of the data contained in this manuscript alters or improves any commercial aspect of Nicogen, the manuscript was not reviewed by others involved with Nicogen, and no Nicogen funds were used in this work. This study involved two primary research centers. Data collection was conducted at DecisionLine

(formerly Ventana) Clinical Research Corporation, 720 King Street West, Toronto, Ontario,

Canada M5V 2T3. Data analyses were conducted at Rm 4326 Medical Science Building, 1

King’s College Circle, University of Toronto, Ontario, Canada, M5S 1A8.

106 Significance of chapter

This chapter contributes to the literature through its characterization and identification of

CYP2A6 genetic variants in a population that had been largely understudied thus far. My contribution to the paper and the significance of this chapter to the thesis are as follows. We were able to account for some of the observed variability in CYP2A6 activity by characterizing several CYP2A6 alleles (e.g. CYP2A6*24) that result in either a decrease- or loss-of-function. In addition, through sequencing we were able to identify novel CYP2A6 variants that include SNPs

(e.g. 6458A>T; N438Y), indels (e.g. CYP2A6*1K), and copy number variants (e.g. an uncharacterized duplication and distinct deletion alleles). This suggests that much genetic variation in CYP2A6 still remains to be identified and characterized. With our current knowledge of CYP2A6 genetic variation in this population of black African descent we developed genotype groups with predicted clusters of nicotine metabolism. This grouping adds to the field by enhancing the ability to effectively perform genetic association studies between

CYP2A6, nicotine metabolism, smoking behavior, and disease.

107 Chapter 2: A Novel CYP2A6 allele (CYP2A6*35) resulting in an amino acid substitution (Asn438Tyr) is associated with lower CYP2A6 activity in vivo

Nael Al Koudsi, Jasjit S. Ahluwalia, Shih-Ku Lin, Edward M. Sellers, and Rachel F. Tyndale

Reprinted with the permission from Nature publishing group. This chapter appears as published in The Pharmacogenomics Journal, 9(24): 274-282, 2009 with modifications to the figure and table numbers.

Dr. Rachel F. Tyndale and Nael Al Koudsi designed the experimental study. Nael Al Koudsi (1) designed primers and developed assay conditions to genotype the samples for CYP2A6*35, (2) did long-PCR cloning for all the specified samples, (3) expressed all the CYP2A6 variant constructs in E.coli cells and assessed their thermal stability and ability to metabolize nicotine to cotinine and (4) performed data analyses and wrote the manuscript. Evan Dorey assisted with genotyping for CYP2A6*35. Helma Nolte and Dr. Benowitz’s lab determined the metabolite ratio of 3HC/COT in the African Canadian and African American populations, respectively. Bin

Zhao determined the in vitro nicotine and cotinine levels. DNA of the African Canadian,

Caucasian, Japanese, and Chinese populations were obtained from previous studies designed by

Dr. Rachel F. Tyndale and Dr. Edward M. Sellers. DNA of the African American and

Taiwanese populations were obtained from Dr. Jasjit S. Ahluwalia and Shih-Ku Lin, respectively.

108 Abstract Cytochrome P450 2A6 (CYP2A6) is the primary human enzyme involved in nicotine metabolism. The objective of this study was to characterize two nonsynonymous single nucleotide polymorphisms in CYP2A6*24, 594G>C (Val110Leu) and 6458A>T (Asn438Tyr).

We determined their haplotype, allele frequencies, effect on CYP2A6 activity in vivo, as well as their stability and ability to metabolize nicotine in vitro. CYP2A6*35 (6458A>T) occurred at a frequency of 2.5–2.9% among individuals of black African descent, 0.5–0.8% among Asians and was not found in Caucasians. In addition, we identified two novel alleles, CYP2A6*36

(6458A>T and 6558T>C (Ile471Thr)) and CYP2A6*37 (6458A>T, 6558T>C and 6600G>T

(Arg485Leu)). In vivo, CYP2A6*35 was associated with lower CYP2A6 activity as measured by the 3HC/COT ratio. In vitro, CYP2A6.35 had decreased nicotine C-oxidation activity and thermal stability. In conclusion, we identified three novel CYP2A6 alleles (CYP2A6*35, *36 and

*37); the higher allele frequency variant CYP2A6*35 was associated with lower CYP2A6 activity.

109 Introduction

Cigarette’s addictive properties are largely due to the psychoactive effects of nicotine

(Benowitz 2008). Owing to the relatively short half-life of nicotine, approximately 2 h

(Benowitz and Jacob 1994), dependent smokers smoke at regular intervals to maintain nicotine levels. Therefore, factors that influence the intake or removal (for example, metabolism) of nicotine from the body can affect smoking behaviors such as the number of cigarettes smoked or the likelihood of cessation (Schoedel et al. 2004; Lerman et al. 2006; Minematsu et al. 2006;

Patterson et al. 2008).

In vivo, 70–80% of nicotine is metabolized to cotinine (COT) (Benowitz and Jacob

1994), and cytochrome P450 2A6 (CYP2A6) mediates approximately 90% of this reaction

(Nakajima et al. 1996a; Messina et al. 1997). CYP2A6 also exclusively mediates cotinine’s hydroxylation to trans-3’-hydroxycotinine (3HC), making the metabolic ratio of 3HC/COT a specific and reliable marker of CYP2A6 activity (Nakajima et al. 1996; Dempsey et al. 2004;

Yamanaka et al. 2004). The 3HC/COT ratio correlates with the rate of nicotine clearance

(Dempsey et al. 2004) and is stable over time (Lea et al. 2006), thereby facilitating its use as a proxy measure for the rate of CYP2A6-mediated nicotine metabolism in population studies. A common feature of nicotine pharmacokinetic studies is the large interindividual variability in the rate of metabolism (Benowitz et al. 1982; Benowitz et al. 2006; Mwenifumbo et al. 2008).

Sources of this large variability may include environmental (for example, smoking), physiological (for example, systemic diseases) and genetic (for example, CYP2A6 polymorphisms) factors (Mwenifumbo and Tyndale 2007).

Recently, we characterized the effect of a number of established and novel CYP2A6 alleles on CYP2A6 activity and nicotine disposition kinetics in a population of black African descent (Mwenifumbo et al. 2008). Interestingly, most of the novel alleles were associated with

110 lower in vivo CYP2A6 activity, accounting for some of the previously uncharacterized variability and slower nicotine metabolism in this population. Nonetheless, even after controlling for gender, smoking and known CYP2A6 alleles, large variability in CYP2A6 activity still remains. One possible source of the remaining variation is the presence of additional genetic variants. Through sequencing, we found that CYP2A6*24 contains two nonsynonymous single nucleotide polymorphisms (SNPs): 594G>C (Val110Leu) and 6458A>T

(Asn438Tyr) (Mwenifumbo et al. 2008). A question that remained is whether 594G>C or

6458A>T can occur on their own, and if so, what are their allele frequencies and functional impact on CYP2A6 activity.

111 Results

Allele frequency of CYP2A6*35 and identification of novel alleles CYP2A6*36 and CYP2A6*37

Because CYP2A6*24 is a haplotype containing 594G>C and 6458A>T (Mwenifumbo et al. 2008), individuals positively genotyped for these two variants were considered to have

CYP2A6*24, whereas individuals positively genotyped for 6458A>T alone formed the

CYP2A6*35 group. We did not find any individuals with the 594G>C SNP alone in any of the ethnic groups. CYP2A6*24 occurred at an allele frequency of 0.7% among African-American smokers (n = 1234 alleles) and was not found among Caucasians (n = 304 alleles), Japanese (n =

120 alleles), Chinese (n = 196 alleles) or Taiwanese (n = 334 alleles). Alleles here are inferred by assuming that each individual has two copies of chromosome 19 containing one allele on each chromosome. Previously we found CYP2A6*24 at a frequency of 1.3% among an African

Canadian population (1.4% in smokers and 1.1% in nonsmokers) and characterized it as a low- activity variant (Mwenifumbo et al. 2008). CYP2A6*35 was found among individuals of African descent with a frequency of 2.9 and 2.5% in African-American smokers and African Canadians

(1.8% in smokers and 3.3% in nonsmokers), respectively. The allele was also found in Japanese

(0.8%), Taiwanese (0.6%) and Chinese (0.5%) but not among Caucasians. The genotype frequencies of CYP2A6*24 and CYP2A6*35 did not significantly deviate from Hardy–Weinberg equilibrium in any population.

The variant CYP2A6*35 alleles from one African Canadian, one African American and one Taiwanese individual were fully sequenced. The variant allele CYP2A6*35A was identical among the African-American and African Canadian individuals, whereas the allele

CYP2A6*35B in the Taiwanese individual differed modestly (Table 13). The allele frequencies of CYP2A6*35 presented herein include both CYP2A6*35A and CYP2A6*35B. In addition, we identified two novel alleles, CYP2A6*36 and CYP2A6*37. CYP2A6*36 contains 6458A>T in

112 haplotype with 6558T>C present in CYP2A6*7, whereas CYP2A6*37 contains 6458A>T and

two additional SNPs present in CYP2A6*10 (6558T>C and 6600G>T; Table 13). CYP2A6*36

and CYP2A6*37 were only found among Taiwanese individuals, both at a frequency of 0.3%.

These alleles were not included in estimating the allele frequency of CYP2A6*35.

Table 13. CYP2A6*35(A and B), CYP2A6*36, and CYP2A6*37 alleles

CYP2A6*35A CYP2A6*35B CYP2A6*36 CYP2A6*37 -1301A>C -1301A>C -1301A>C -1301A>C -1289G>A -1289G>A -1289G>A -1289G>A -745A>G -745A>G -745A>G -1013A>G 22C>T (L8L, Exon 1) 22C>T (L8L, Exon 1) 22C>T (L8L, Exon 1) 720G>A, Intron 2 1137C>G, Intron 2 1620T>C, Intron 2 1620T>C, Intron 2 1620T>C, Intron 2 1620T>C, Intron 2 2483G>A, Intron 4 3225A>G, Intron 4 4084delA, Intron 5 4084delA, Intron 5 4084delA, Intron 5 6218A>G, Intron 8 6282A>G, Intron 8 6293T>C, Intron 8 6354T>C, Intron 8 6354T>C, Intron 8 6458A>T (N438Y, Exon 9) 6458A>T (N438Y, Exon 9) 6458A>T (N438Y, Exon 9) 6458A>T (N438Y, Exon 9) 6558T>C (I471T, Exon 9) 6558T>C (I471T, Exon 9) 6600G>T (R485L, Exon 9) 2A6*1B 2A6*1B 2A6*1B 2A6*1B 6782C>G 6782C>G 6782C>G 6782C>G 6835C>A 6835C>A 6835C>A 6936_6937insCACTT 6961_6962insGAAAAG 6989A>G 6999T>C 6999T>C 6999T>C 7160A>G 7160A>G 7160A>G 7160A>G

Numbering of the nucleotides is in reference to the ATG start site of the CYP2A6 gene (reference genomic sequence NG_000008.7) in which A is numbered 1 and the base before A is numbered -1. Bolded nucleotides are synonymous SNPs. Bolded and italicized nucleotides are nonsynonymous SNPs. 2A6*1B refers to the 58 bp gene conversion in the 3’ untranslated region that was present in all alleles.

113 CYP2A6*35 is associated with lower CYP2A6 activity in vivo

In the African Canadian population, individuals with CYP2A6*1/*35 genotype had significantly lower mean 3HC/COT compared to those in the wild-type group (CYP2A6*1/*1;

Table 14). Four individuals were compound heterozygotes, one having CYP2A6*9/*35 and three having CYP2A6*17/*35. All four individuals had substantially lower 3HC/COT ratios compared to wild-type and heterozygote CYP2A6*1/*35 individuals (Table 14).

Among the African Americans, those with CYP2A6*1/*35 genotype also had significantly lower mean 3HC/COT compared to individuals in the wild-type group

(CYP2A6*1/*1; Table 14). In addition, an individual homozygous for CYP2A6*35 had lower

3HC/COT compared to the heterozygous individuals suggestive of a gene–dose effect. As seen in the African Canadian population, this population had compound heterozygote individuals

(CYP2A6*9/*35, n = 2; CYP2A6*17/*35, n = 3) with lower 3HC/COT values compared to wild type.

114 Table 14. CYP2A6*35 is associated with lower in vivo CYP2A6 activity.

Population Genotype n Mean adjusted s.d. P-value Wild-type 3HC/COT activity (%) African *1/*1 151 1.20 0.63 100 Canadian *1/*35 10 0.84 0.50 0.02 70 *9/*35a 1 0.47 - 39 *17/*35a 3 0.08 0.14 7

African *1/*1 258 1.21 0.80 100 American *1/*35 20 0.75 0.26 0.003b 62 *35/*35a 1 0.51 - 42 *9/*35a 2 0.78 0.51 62 *17/*35a 3 0.27 0.09 22

Abbreviations: 3HC/COT, trans-3’-hydroxycotinine/cotinine ratio; n, number of samples; s.d., standard deviation. a-The number of individuals with greater than one variant was too small for statistical analyses. b-The P-value obtained does not include the homozygote CYP2A6*35 individual. The ratio was derived from plasma metabolite levels following oral nicotine intake or ad libitum smoking among the African Canadian and African American population, respectively. The effect of CYP2A6*35 remained significant in both populations when the wild-type group was restricted to individuals with at least one CYP2A6*1B allele (i.e., CYP2A6*1A/*1B and CYP2A6*1B/*1B). The 3H/COT ratios among individuals with CYP2A6*35 compared to individuals with only CYP2A6*1B were as follows: 0.84±0.50 vs 1.25±0.66, P = 0.03 among African Canadians and 0.75±0.26 vs 1.38±0.75, P<0.001 among African Americans. The P-values are from the regression model as described in the Materials and methods section with the dependent variable being log 3HC/COT, controlling for gender and smoking. When the data were analyzed with a Mann–Whitney test where the dependent variable was 3HC/COT (i.e., not log- normalized), and where the effects of gender and smoking were not considered, the effect of CYP2A6*35 remained significant in both populations (African Canadian P = 0.016 and the African American P = 0.001).

115 In vitro enzymatic activities of the CYP2A6 variants

The apparent Km and Vmax values for nicotine C-oxidation by CYP2A6.1 and the variant constructs are presented in Table 15. The Vmax value of CYP2A6.17 tended toward being lower compared to CYP2A6.1 (P = 0.1), whereas the other variants (CYP2A6.V110L,

CYP2A6.35 and CYP2A6.24) had similar values compared to CYP2A6.1. The apparent Km value of CYP2A6.17 was significantly higher than that of CYP2A6.1. The Km values of

CYP2A6.V110L, and CYP2A6.35 tended to be higher compared to CYP2A6.1, whereas

CYP2A6.24 tended to have a lower value. The resulting Vmax/Km (catalytic efficiency) values of CYP2A6.17, CYP2A6.V110L and CYP2A6.35 were significantly lower compared to that of

CYP2A6.1. In contrast, the Vmax/Km value of CYP2A6.24 was similar to that of CYP2A6.1.

Table 15. Kinetic parameters of the wildtype and variant CYP2A6 constructs for the metabolism of nicotine.

Km (µM) Vmax Vmax/Km (pmol/min/pmol (nl/min/pmol CYP2A6) CYP2A6) CYP2A6.1 7.0±0.7 0.53±0.04 8.0±0.6 CYP2A6.17 33.4±3.3* 0.36±0.03 1.1±0.1* CYP2A6.V110L 11.3±1.4 0.52±0.06 4.7±0.3* CYP2A6.35 9.9±0.8 0.49±0.05 5.0±0.2* CYP2A6.24 5.4±1.8 0.53±0.09 10.5±1.9

Data are presented as mean±s.e.m. of 2–7 independent experiments. *P<0.05 when compared to CYP2A6.1.

To determine whether the amino-acid changes could affect protein stability we tested the thermal stability of the constructs by incubating them at 37°C and collecting aliquots at various times to measure nicotine C-oxidation activity and CYP2A6 protein levels. Generally, the activity of the variant constructs decreased to a greater extent compared to the wild-type

116 construct (Figure 14a). The time it took to decrease CYP2A6 activity by 50% (t1/2) was significantly shorter for CYP2A6.17 (t1/2 = 2.2 h), CYP2A6.V110L (t1/2 = 2.9 h) and

CYP2A6.35 (t1/2 = 2.6 h) compared to CYP2A6.1 (t1/2 = 4.3 h, P<0.05), whereas the t1/2 of

CYP2A6.24 was similar to CYP2A6.1 (4.6 h compared with 4.3 h, respectively). With incubation time the CYP2A6.1 protein levels did not change, whereas the protein levels of

CYP2A6.17, CYP2A6.V110L, CYP2A6.35 and CYP2A6.24 decreased (Figure 14b). Of note,

CYP2A6 protein expression following E. coli culture was lower among the variant constructs compared to wild type, CYP2A6.1 > CYP2A6.V110L and CYP2A6.24 > CYP2A6.35 >

CYP2A6.17 (0.44 > 0.38 > 0.20 > 0.10 pmol of CYP2A6 per µg of membrane protein).

117 CYP2A6*35 reduces enzyme activity N Al Koudsi et al 277

Table 3 Kinetic parameters of the wild-type and variant a 100 CYP2A6 constructs for the metabolism of nicotine

80 Km (mM) Vmax (pmol/min/ Vmax/Km (nl/min/ pmol CYP2A6) pmol CYP2A6)

60 CYP2A6.1 7.0±0.7 0.53±0.04 8.0±0.6 CYP2A6.17 33.4±3.3* 0.36±0.03 1.1±0.1* CYP2A6.V110L 11.3±1.4 0.52±0.06 4.7±0.3* 40 CYP2A6.35 9.9±0.8 0.49±0.05 5.0±0.2* CYP2A6.24 5.4±1.8 0.53±0.09 10.5±1.9 20 % Remaining activity (Mean -S.E.M.) (Mean activity Remaining % Data are presented as mean±s.e.m. of 2–7 independent experiments. *Po0.05 when compared to CYP2A6.1. 0 048 12 16 20 24 Time (hrs)

CYP2A6.1 (t1/2 4.3 h, Po0.05), whereas the t1/2 of ¼ b 120 Time (hrs) 0812 CYP2A6.24 was similar to CYP2A6.1 (4.6 h compared with 100 2A6.1 4.3 h, respectively). With incubation time the CYP2A6.1 80 protein levels did not change, whereas the protein levels of 2A6.17 60 CYP2A6.17, CYP2A6.V110L, CYP2A6.35 and CYP2A6.24 2A6.V110L decreased (Figure 1b). Of note, CYP2A6 protein expression 40 Relative protein

remaining (% + SEM) 20 following E. coli culture was lower among the variant 2A6.35 constructs compared to wild type, CYP2A6.14CY- 0 04812 2A6.24 P2A6.V110L and CYP2A6.244CYP2A6.354CYP2A6.17 Time (hrs) (0.4440.3840.2040.10 pmol of CYP2A6 per mg of mem- Figure 14. Thermal stability of the CYP2A6 wild-type and variant constructs at 37 C. (a) Figure 1 Thermal stability of the CYP2A6 wild-type and variant° brane protein). Activities are expressed as the percentage of remaining enzymatic activity for each construct constructswith the activity at 37 at 1 0C. h ( recordeda) Activities as 100%. are Aliquots expressed (10pmol as of the CYP2A6 percentage enzyme of for each remainingconstruct) were enzymatic taken at 0,activity 4, 8, 12 for and each 24h toconstruct measure cotinine with the form activityation using at 0 30 h µM of recordednicotine. The as activity 100%. of each Aliquots construct (10 at pmol0 h is displayed of CYP2A6 as 100%, enzyme and over fortime, each the activity Discussion construct)was measured were for each taken construct at 0, and 4, expressed 8, 12 as and a percent 24 h remaining. to measure We used cotinine this approach to minimize potential impact of modestly different amounts of estimated CYP2A6 in the initial formationreaction. The using lines 30 throughmM of the nicotine. points are The stimulated activity using ofeach the plateau construct and rate at 0 of h decay is (k) The focus of the present study was two SNPs found in displayedvalues obtained as 100%, for each and CYP2A6 over construct time, thefrom activity the one-phase was decay measured equation for as eachdescribed in CYP2A6, 594G4C and 6458A4T. Here, we did not find constructthe Materials and and expressed methods section. as a percent(b) Immunoreactive remaining. CYP2A6 We used protein this at the approach indicated times toof minimizeincubation. With potential time, the impact protein levels of modestly of the CYP2A6 different variant amounts constructs decreased of individuals with 594G4C alone, suggesting that 594G4C is whereas CYP2A6.1 protein levels remained the same. At 24 h, no CYP2A6 protein was in obligate haplotype with 6458A4T forming CYP2A6*24. estimateddetected with CYP2A6 any of the in CYP2A6 the initial constructs reaction. (data not The shown). lines CYP2A6 through (1.2 the pmol) points was loaded Conversely, individuals with only 6458A4T were found. arein each stimulated well. using the plateau and rate of decay (k) values obtained The allele containing 6458A4T was sequenced and assigned for each CYP2A6 construct from the118 one-phase decay equation as described in the Materials and methods section. (b) Immunoreactive the name CYP2A6*35 by the CYP allele nomenclature CYP2A6 protein at the indicated times of incubation. With time, the committee (http://www.cypalleles.ki.se/). CYP2A6*35 was protein levels of the CYP2A6 variant constructs decreased whereas predominantly found among individuals of African descent CYP2A6.1 protein levels remained the same. At 24 h, no CYP2A6 protein and to a lesser extent among individuals of Asian descent. was detected with any of the CYP2A6 constructs (data not shown). The CYP2A6*35 allele sequence differed slightly between CYP2A6 (1.2 pmol) was loaded in each well. African and Asian descent distinguishing it as CYP2A6*35A and CYP2A6*35B, respectively. We also discovered two novel alleles among Taiwanese individuals where the 6458A4T B70% of the activity of the wild-type CYP2A6 group, a SNP was in haplotype with 6558T4C in CYP2A6*7 or reduction in activity that is similar to well-characterized 6558T4C 6600G4T in CYP2A6*10, forming CYP2A6*36 decreased function alleles CYP2A6*9, CYP2A6*12 and þ and CYP2A6*37, respectively. We could not test the func- CYP2A6*17.14,15,22 Several case–control studies have asso- tional in vivo consequence of CYP2A6*36 and CYP2A6*37, ciated CYP2A6 genetic variants that result in lower CYP2A6 but because both of these rare alleles contain the amino-acid activity with a lower likelihood of smoking;5,23 however, substitution (Ile471Thr) found in CYP2A6*7 and negative findings have also been reported.24,25 Among the CYP2A6*10, it is likely they will result in a loss of function. African Canadians, a trend was observed in which indivi- The substitution Ile471Thr has been associated with lower duals with CYP2A6*35 were less likely to be current adult CYP2A6 activity and stability in vitro (E. coli system)17 in smokers (OR 0.51; 95% CI 0.17–1.60, P 0.19). The ¼ ¼ ¼ addition to lower in vivo CYP2A6 activity in multiple allele frequency of CYP2A6*35 was 3.3% (n 9, 269 alleles) ¼ studies.18–21 in smokers compared to 1.8% (n 5, 275 alleles) in ¼ Our in vivo data strongly suggest that CYP2A6*35 is nonsmokers. The magnitude of the effect was the same associated with lower CYP2A6 activity. In both African when analysis was restricted to individuals with populations, individuals heterozygous for CYP2A6*35 had only CYP2A6*1/*1 and CYP2A6*1/*35 (OR 0.39, 95% ¼

The Pharmacogenomics Journal Discussion

The focus of the present study was two SNPs found in CYP2A6, 594G>C and 6458A>T.

Here, we did not find individuals with 594G>C alone, suggesting that 594G>C is in obligate haplotype with 6458A>T forming CYP2A6*24. Conversely, individuals with only 6458A>T were found. The allele containing 6458A>T was sequenced and assigned the name CYP2A6*35 by the CYP allele nomenclature committee (http://www.cypalleles.ki.se/). CYP2A6*35 was predominantly found among individuals of African descent and to a lesser extent among individuals of Asian descent. The CYP2A6*35 allele sequence differed slightly between African and Asian descent distinguishing it as CYP2A6*35A and CYP2A6*35B, respectively. We also discovered two novel alleles among Taiwanese individuals where the 6458A>T SNP was in haplotype with 6558T>C in CYP2A6*7 or 6558T>C + 6600G>T in CYP2A6*10, forming

CYP2A6*36 and CYP2A6*37, respectively. We could not test the functional in vivo consequence of CYP2A6*36 and CYP2A6*37, but because both of these rare alleles contain the amino-acid substitution (Ile471Thr) found in CYP2A6*7 and CYP2A6*10, it is likely they will result in a loss of function. The substitution Ile471Thr has been associated with lower CYP2A6 activity and stability in vitro (E. coli system) (Ariyoshi et al. 2001) in addition to lower in vivo

CYP2A6 activity in multiple studies (Yoshida et al. 2002; Xu et al. 2002a; Nakajima et al. 2006;

Peamkrasatam et al. 2006).

Our in vivo data strongly suggest that CYP2A6*35 is associated with lower CYP2A6 activity. In both African populations, individuals heterozygous for CYP2A6*35 had ~70% of the activity of the wild-type CYP2A6 group, a reduction in activity that is similar to well- characterized decreased function alleles CYP2A6*9, CYP2A6*12 and CYP2A6*17 (Benowitz et al. 2006; Mwenifumbo et al. 2008; Ho et al. 2009). Several case–control studies have associated

CYP2A6 genetic variants that result in lower CYP2A6 activity with a lower likelihood of

119 smoking (Iwahashi et al. 2004; Schoedel et al. 2004); however, negative findings have also been reported (Tan et al. 2001; Zhang et al. 2001). Among the African Canadians, a trend was observed in which individuals with CYP2A6*35 were less likely to be current adult smokers

(OR=0.51; 95% CI=0.17–1.60, P=0.19). The allele frequency of CYP2A6*35 was 3.3% (n=9,

269 alleles) in smokers compared to 1.8% (n=5, 275 alleles) in nonsmokers. The magnitude of the effect was the same when analysis was restricted to individuals with only CYP2A6*1/*1 and

CYP2A6*1/*35 (OR=0.39, 95% CI=0.01–1.60, P=0.19). All the African-American individuals were participants in a smoking cessation trial (that is, smokers) limiting our ability to further test this hypothesis.

Our in vitro enzymatic function data (Table 15) suggest that the variants

CYP2A6.V110L (Val110Leu) and CYP2A6.35 (Asn438Tyr) result in lower nicotine C- oxidation activity compared to the wild type. This was driven by the higher apparent Km values observed for CYP2A6.V110L and CYP2A6.35. In addition, CYP2A6.V110L and CYP2A6.35 had reduced thermal stability compared to the wild type. The most dramatic reduction in stability was seen following 4h of incubation, in which the constructs CYP2A6.17,

CYP2A6.V110L and CYP2A6.35 lost more than 50% of their initial activity, whereas the wild- type (CYP2A6.1) lost only about 30% of its initial activity. It is interesting to note that the amino-acid substitutions Val110Leu and Asn438Tyr are not highly conserved among the

CYP2A family. For instance, the human CYP2A13 and the mouse CYP2A5, both of which metabolize nicotine efficiently, have leucine (Leu) at position 110 and tyrosine (Tyr) at position

438. In addition, the physical and chemical properties of valine (Val) and leucine (Leu) and asparginine (Asn) and tyrosine (Tyr) are similar to each other (for example, Val and Leu are similarly aliphatic, whereas Asp and Tyr are similarly polar). It is possible though that the positions of the following amino-acid changes are important. Valine 110 resides within the substrate recognition site 1 (residues 101–120) (Lewis et al. 1999) and is in very close proximity 120 to the active site of CYP2A6 (Johnson and Stout 2005; Yano et al. 2005). Thus, this substitution

(Val110Leu) might alter the active site cavity (Johnson and Stout 2005). On the other hand, aspargine 438 resides on the surface of CYP2A6 and is situated next to the conserved heme- binding cysteine 439 (Juvonen et al. 1991; Lewis et al. 1999). Thus, it is possible that the amino-acid substitution Asn438Tyr might affect the binding of the heme. In our study the reduction in protein levels did not fully complement the reduction in CYP activity suggesting that the heat treatment might have also facilitated the loss of heme. We were unable to test this

(due to the low sensitivity of our assay); however it would be valuable in the future to determine the stability of the holoenzyme by obtaining the CO difference spectrum for the constructs.

It was surprising for us to find that individually Val110Leu and Asn438Tyr reduced

CYP2A6 activity and stability, whereas in CYP2A6.24 the occurrence of both amino-acid changes appeared to have stabilized the enzyme. This is in contrast to our previously published study (Mwenifumbo et al. 2008), in which individuals with CYP2A6*24 were associated with lower CYP2A6 activity measured by the 3HC/COT ratio in vivo. There are several possible reasons for the discordance between the in vivo and in vitro data for CYP2A6*24. It is possible that the coding SNPs might be affecting protein degradation (discussed below). Alternatively, the coding SNPs might be in linkage disequilibrium with polymorphisms in the gene promoter or within introns that affect transcription and mRNA splicing, and as a result might influence function in vivo. For example, CYP2A6*35A and CYP2A6*24 contain -1013A>G, which is thought to disrupt a putative enhancer region (Pitarque et al. 2004), whereas CYP2A6*35B contains -745A>G that is thought to disrupt a CCAAT box to which the regulatory element NF-

Y binds (von Richter et al. 2004). Both of these promoter SNPs have been shown to result in decreased expression in a reporter construct system (Pitarque et al. 2004; von Richter et al.

2004).

121 The protein levels of CYP2A6.1 did not appear to decrease during the first 12 h of incubation, although its activity did. This is likely due to the fact that immunoblotting not only detected functional holoenzyme, but also apoenzyme and misfolded CYP2A6. However, it is interesting to note that unlike CYP2A6.1, the CYP2A6 protein levels of the variant constructs all decreased with incubation time suggesting that these constructs are more susceptible to degradation. It was also noted that the CYP2A6 variant constructs had lower CYP2A6 protein expression (pmol of CYP2A6 per µg of membrane protein) compared to the wild-type construct following culturing of the E. coli. There is a growing body of evidence for nonsynonymous coding SNPs influencing enzymatic function by decreasing the amount of protein through accelerated degradation (Wang et al. 2003c; Weinshilboum and Wang 2004; Bandiera et al.

2005; Salavaggione et al. 2005). Much of this evidence comes from cytoplasmic proteins such as thiopurine S-methyltransferase (Salavaggione et al. 2005), and has also been shown for cytochrome P450. For example, Bandiera et al. (2005) demonstrated that CYP1B1.4 is expressed at lower levels compared to CYP1B1.1 in COS-1 cells largely due to its increased rate of degradation.

The present study suggests that CYP2A6*35 is associated with lower in vivo CYP2A6 activity, likely due to its lower nicotine C-oxidation activity, stability and greater susceptibility to protein degradation. Because of its lower stability, CYP2A6*35 is likely to result in lower in vivo activities toward all CYP2A6 substrates; however this remains to be tested. In conclusion, we have identified several novel alleles (CYP2A6*35, CYP2A6*36 and CYP2A6*37), one of which (CYP2A6*35) is found at a relatively high frequency among individuals of African descent and is associated with lower CYP2A6 activity in vivo. Identifying and characterizing novel CYP2A6 variants among different ethnic groups will help refine the relationship between

CYP2A6 genotype and nicotine metabolism phenotype, thus increasing the utility of genotyping

122 in epidemiological and clinical treatment studies (Fujieda et al. 2004; Malaiyandi et al. 2006;

Mwenifumbo et al. 2008).

123 Materials and Methods

CYP2A6 6458A>T genotype assay

A two-step PCR assay was developed to detect the 6458A>T SNP based on the reference genomic sequence NG_000008.7. The first PCR reaction amplified a gene-specific region from intron 6 to the 3’-flanking region using the primers 2A6in6F1 and 2A6R0 as described previously (Mwenifumbo et al. 2008a). In the second PCR reaction, an allele-specific region (1327 bp) from exon 9 to the 3’–flanking region of CYP2A6 was amplified using one of the two forward primers 2A6in8ex9F6458W or 2A6in8ex9F6458V in combination with one reverse primer 2A6R0 (Table 16). The 25 µl reaction consisted of 1X PCR buffer (10mM Tris

(pH 8.8), 50mM KCl), 0.1 mM of each dNTP, 1.25 mM MgCl2, 0.15 µM of each primer, 0.25 U of Taq polymerase (MBI Fermentas, Burlington, Canada), 0.8µl of undiluted first amplification

PCR product and H2O. Initial denaturation at 95°C for 1 min was followed by 20 cycles, each consisting of denaturation at 95°C for 15 s, annealing at 55°C for 20 s and extension at 72 °C for

1 min, followed by a final extension at 72°C for 5 min. 594G>C was detected as described previously (Mwenifumbo et al. 2008). The PCR products were analyzed by electrophoresis on a

1.2% agarose gel (OnBio, Richmond Hill, Canada) stained with ethidium . The assay was developed using multiple cosmids containing genomic clones of CYP2A6 as a positive control and CYP2A7 and CYP2A13 as negative controls (provided by Dr Linda Ashworth,

Human Genome Centre, Liverpool) (Hoffman et al. 1995).

124 Table 16. Primers used.

Primer Sequence Location 2A65Pr1Fa 5’-ACC TAG ACT TAA TCT TCC CGT ATA C-3’ 5’-flanking region 2A6in8ex9F6458W 5’-TCC TCA GGA AAG CGG A-3’ Intron8/Exon9 2A6in8ex9F6458V 5’-TCC TCA GGA AAG CGG T-3’ Intron8/Exon9 2A6R0b 5’-AGG TCA TCT AGA TTT TCT CCT ACA-3’ 3’-flanking region 2A6_594G>C 5’-GCC ACC TTC GAC TGG CTC TTC AAA GGC TAT G-3’ Exon 2 2A6_594G>C Anti 5’-CAT AGC CTT TGA AGA GCC AGT CGA AGG TGG C-3’ Exon 2 2A6_6458A>T 5’-TTC CAT CGG AAA GCG GTA CTG TTT CGG AGA AGG-3’ Exon 8/9 2A6_6458A>T Anti 5’-CCT TCT CCG AAA CAG TAC CGC TTT CCG ATG GAA-3’ Exon 8/9

a- (Pitarque et al. 2004) b- (Mwenifumbo et al. 2008a)

CYP2A6 sequencing

To confirm our genotyping assay and determine the haplotype of the allele, we amplified

(long PCR) and sequenced a 9.2kb fragment containing the CYP2A6 gene from three individuals

heterozygous for 6458A>T (one African Canadian, one African American and one Taiwanese).

The 25 µl PCR reaction used the primers 2A65Pr1F and 2A6R0 (Table 16) and was performed

following the manufacturer’s instructions (long PCR enzyme mix; MBI Fermentas) with the

following minor modifications. The annealing temperature was 57°C and the reaction contained

75 ng of DNA, 0.3 mM of each dNTP and 0.25 µM of each primer. The PCR product was

subcloned and sequenced as described previously (Mwenifumbo et al. 2008). In addition, the

CYP2A6 gene (9.2kb) of two individuals was amplified. One was genotyped as having

6458A>T as well as 6558T>C present in CYP2A6*7, whereas the other was genotyped as

having 6458A>T and 6558T>C + 6600G>T present in CYP2A6*10. Following CYP2A6 gene

(9.2kb) amplification the products were cloned and then genotyped or sequenced to determine

whether 6458A>T is in haplotype with the SNPs in CYP2A6*7 and CYP2A6*10.

125 CYP2A6 genotyping and phenotyping

Individuals from different ethnic backgrounds, African Canadian (n=279), African

American (n=617), Japanese (n=60), Taiwanese (n=67), Chinese (n=98) and Caucasian (n=152) were genotyped for the 6458A>T variant. In addition, the African Americans and African

Canadians were genotyped for other CYP2A6 variants including CYP2A6*1B, *2, *4, *9, *12,

*14, *17, *20, *21, *23, *24, *25, *26, *27 and *28, whereas the Taiwanese, Chinese, Japanese and Caucasians were genotyped for CYP2A6*1B, *2, *4, *7, *8, *9, *10 and *12. The demographics of the African Canadian, African-American, Japanese, Chinese, Taiwanese and

Caucasian populations have been previously described (Schoedel et al. 2004; Ahluwalia et al.

2006; Mwenifumbo et al. 2007a).

Phenotypic measures were available for the African Canadian and African-American populations. The plasma 3HC/COT ratio collected at 270 min following oral nicotine (4mg capsule) was used as a proxy measure for CYP2A6 activity among the African Canadians

(Mwenifumbo et al. 2007a), whereas the plasma 3HC/COT ratio collected from ad libitum smoking (Dempsey et al. 2004) was used in the African-American population (Ho et al. 2009).

In both populations CYP2A6 genotype has been shown to be significantly associated with the ratio (3HC/COT), supporting its use as a measure of CYP2A6 activity (Mwenifumbo et al.

2008; Ho et al. 2009). All studies were approved by ethics board of the University of Toronto.

Construction and expression of CYP2A6 constructs in E. coli

A bicistronic construct containing the full-length cDNA of CYP2A6 followed by the human nicotinamide adenine dinucleotide phosphate oxidase (NADPH)-cytochrome P450 reductase (hNPR) inserted into a pCW expression vector was generously provided by Dr E

Gillam (Gillam et al. 1999; Ho et al. 2008). Two amino-acid substitutions, Val110Leu and

Asn438Tyr, were introduced either alone or in combination to create CYP2A6.V110L,

CYP2A6.35 (Asn438Tyr) and CYP2A6.24 (Val110Leu + Asn438Tyr), respectively. The 126 commercial kits QuickChange II XL Site-Directed Mutagenesis and Quick-Change Multi Site-

Directed Mutagenesis were used to introduce the single and multiple amino-acid substitutions, respectively, according to the manufacturer’s directions (Stratagene, La Jolla, CA, USA).

Primers used to introduce Val110Leu (CYP2A6.V110L) were 2A6_594G>C and 2A6_594G>C

Anti (Table 16). Asn438Tyr (CYP2A6.35) was introduced using the primers 2A6_6458A>T and 2A6_6458A>T Anti (Table 16). The combination of 2A6_594G>C Anti and

2A6_6458A>T Anti was used to construct CYP2A6.24 (Val110Leu + Asn438Tyr). Introduction of the correct changes to the variant constructs was confirmed by sequencing. In addition, a construct with 802 bp deleted from the CYP2A6 cDNA (CYP2A6.NEG) was used as a negative control, whereas a CYP2A6.17 construct served as a positive control (Ho et al. 2008). All five variant constructs (CYP2A6.NEG, CYP2A6.17, CYP2A6.V110L, CYP2A6.35 and

CYP2A6.24) in addition to the wild-type CYP2A6 (CYP2A6.1) were expressed in E. coli and membrane fractions were prepared as described previously (Ho et al. 2008). Membrane protein content was determined by the Bradford protein assay (Bio-Rad Labs, Mississauga, Ontario,

Canada) and the amount of CYP2A6 protein was determined by immunoblotting with a monoclonal CYP2A6 antibody (BD Biosciences, Mississauga, Ontario, Canada) as described previously (Ho et al. 2008).

In vitro nicotine assay

Nicotine C-oxidation activity was assessed as previously described with minor modifications (Ho et al. 2008). The reaction mixture (final volume of 0.5ml) contained E. coli membrane preparations (10 pmol CYP2A6), 20 nmol of expressed cytochrome b5 (Invitrogen,

Burlington, Canada), 50 mmol Tris-HCl buffer (pH 7.4), 1 mmol NADPH (Sigma-Aldrich,

Oakville, Canada), nicotine substrate (ranging from 2 to 100 µM), and 1.0 mg protein per ml human liver cytosol (added in excess so that CYP-mediated oxidation would be rate limiting).

The reaction was initiated by the addition of 1 mmol NADPH following a 2 min preincubation 127 at 37°C. After 15 min incubation, the reaction was stopped by the addition of 100µM of

Na2CO3. Cotinine formation was determined by high-pressure liquid chromatography as described previously (Siu et al. 2006). Cotinine formation increased linearly with CYP2A6 protein concentrations (5–70 pmol) and time (5–30 min) at 100 and 500 µM of nicotine (data not shown). The minimal amounts of cotinine detected in the reaction with CYP2A6.NEG were used as baseline correction to control for any contamination of the substrate or cotinine formation by the E. coli system. The kinetic parameters (Vmax and Km) were estimated from the high-affinity site using the Michaelis–Menten equation in the computer program GraphPad

Prism (GraphPad Software Inc., La Jolla, CA, USA version 5.0 for Windows).

In vitro thermal stability assay

To test the thermal stability of the different CYP2A6 constructs, E. coli membrane preparations were incubated at 37°C for 24 h. The experiment was carried out twice independently. On one day, single aliquots were collected, whereas on the other day duplicate aliquots were collected. Data were consistent within and between days. Aliquots were taken at

0, 4, 8, 12 and 24h to measure nicotine C-oxidation activity. Activity assays were carried out with 30µM of nicotine as described above. In addition, aliquots were taken at 0, 8, 12 and 24 h to measure CYP2A6 protein levels by immunoblotting.

Statistical analyses

Hardy–Weinberg equilibrium was tested using χ2 - or Fisher’s exact test if five or fewer individuals were in one genotype group. Because gender and smoking status significantly influence the 3HC/COT ratio in the African Canadian population (Mwenifumbo et al. 2007a) and other populations (Johnstone et al. 2006; Nakajima et al. 2006; Benowitz et al. 2006a), we controlled for gender and smoking. The African-American population were all smokers, thus we only controlled for gender. Throughout the paper the 3HC/COT ratio is presented as the mean

128 adjusted value. The adjustment for the impact of gender and smoking status was made as previously described (Mwenifumbo et al. 2008). Briefly, the metabolic ratio for each individual was adjusted by dividing the 3HC/COT value with the overall mean from their respective group.

For example, the 3HC/COT of each female nonsmoker was divided by the overall mean of their respective group (that is, female nonsmokers). The mean adjusted ratios were not used in any statistical analyses. Instead we used the raw 3HC/COT values and controlled for the effects of gender and smoking using a regression model. We adopted a forward step-wise regression strategy in which gender and/or smoking status were in the first block and CYP2A6 genotype in the second block. The reference for CYP2A6 genotype in the regression analyses was that with wild-type genotype (CYP2A6*1/*1) and the dependent variable was log 3HC/COT. The wild- type genotype group (CYP2A6*1/*1) included individuals with undetected CYP2A6 variants and the wild-type variant CYP2A6*1B. The CYP2A6*1B allele was included in the wild-type group

(CYP2A6*1/*1) for two main reasons. First, the SNPs in CYP2A6*35 are found on the

CYP2A6*1B background, meaning they are in linkage with the 58 bp gene conversion found in

CYP2A6*1B (Table 13). Second, CYP2A6*1B is a wild-type allele usually included in the wild- type reference group in the literature (Nakajima et al. 2006; Mwenifumbo et al. 2008) along with other variants of the CYP2A6*1 allele (CYP2A6*1B–*1K, listed at http://www. cypalleles.ki.se/cyp2a6.htm). To compare the impact CYP2A6*35 to other variants in the literature, we have included the individuals with CYP2A6*1B in the wild-type (CYP2A6*1/*1) group. The 3HC/COT values were log-normalized as the 3HC/COT values were not normally distributed.

In vitro kinetic parameters (Vmax, Km, Vmax/Km) of the variant CYP2A6 constructs were compared to the wild-type CYP2A6 construct using one-way analysis of variance followed by the Bonferroni post hoc test. The half-life (t1/2) for the thermally induced loss of CYP2A6 activity was determined according to the one-phase decay equation (GraphPad Prism, version 129 5). The t1/2 values reported herein are a measure of the thermal activity of the enzyme as previously determined (Ariyoshi et al. 2001; Nakamura et al. 2002) and not the t1/2 of the

CYP2A6 protein. Time courses of multiple experiments performed were analyzed in concert to obtain the value of the decay rate constant k; which was subsequently used to determine the half-life (t1/2 = ln2/k). By fixing the value of k obtained for the wild type (CYP2A6.1), the changes in the absolute sum of squares of the variant constructs were compared to their baseline fits to determine the level of significance. All analyses were performed on the collected data, whereas the data were normalized as the percentage remaining activity for presentation. All P- values are one tailed with P<0.05 considered statistically significant. Statistical analyses were performed using SPSS (SPSS Inc., Chicago, IL, USA version 15.0 for Windows) and GraphPad

Prism (GraphPad Software Inc., version 5.0 for Windows).

130 Acknowledgments

This study was supported by the Centre for Addiction and Mental Health, Canadian

Institutes for Health Research (CIHR) MOP86471 and CA91912 (JSA). We thank Bin Zhao,

Ewa B Hoffmann, Qian Zhou, Jill C Mwenifumbo, Man Ki Ho and Evan Dorey for technical assistance; Rabindra Shivnaraine for valuable input in data analyses and Sandy Faheim for careful review of the paper. We acknowledge the generous contribution of Dr Benowitz’s Lab in measuring the 3HC and COT levels among the African–American population. We also thank Dr

Elizabeth Gillam for generous gift of the CYP2A6 construct and Dr Linda Ashworth for generously providing us with cosmid DNA clones 19296, 19019 and 27292 that contain

CYP2A6, CYP2A7 and CYP2A13. NK receives funding from CIHR-Strategic Training Program in Tobacco Use in Special Populations (TUSP) and Ontario Graduate Scholarship program

(OGS). RFT holds a Canada Research Chair in Pharmacogenetics.

131 Significance of Chapter

In the previous chapter (Chapter 1) we characterized the functional impact of several

CYP2A6 alleles that accounted for some of the observed variability in CYP2A6 activity in vivo.

Nonetheless, even after controlling for these CYP2A6 alleles and other factors (e.g. gender and smoking) large variability in CYP2A6 activity still remained. One possible source of the remaining variation is the presence of additional genetic variants. In this chapter we identified three novel CYP2A6 alleles (CYP2A6*35, *36 and *37) one of which (CYP2A6*35) occurs at a relatively high allele frequency and is associated with lower CYP2A6 activity. The significance of this chapter is our ability to account for some of the observed variability in CYP2A6 activity by identifying and characterizing novel CYP2A6 alleles that alter CYP2A6 activity. The identification and characterization of novel CYP2A6 variants among different ethnic groups helps to refine the relationship between CYP2A6 genotype and nicotine metabolism phenotype, thus increasing the utility of genotyping in epidemiological and clinical treatment studies.

132 Chapter 3: Hepatic CYP2A6 levels and nicotine metabolism: impact of genetic, physiological, environmental, and epigenetic factors

Nael Al Koudsi, Ewa B. Hoffmann, Abbas Assadzadeh, and Rachel F. Tyndale

Reprinted with the permission from Springer. This chapter appears as published in The

European Journal of Clinical Pharmacology, 66(3): 239-251, 2010 with modifications to the figure and table numbers.

Dr. Rachel F. Tyndale and Nael Al Koudsi designed the experimental study. Nael Al Koudsi (1) isolated DNA from the K- and M-livers, (2) genotyped the majority (~75%) of the samples for

CYP2A6 alleles, (3) determined CYP2A6 protein levels for the K- and M-livers, (4) assisted with DNA sodium bisulfite treatment and subsequent PCR reactions, (5) assisted in determining the mRNA levels of CYP2A6, and (6) performed data analyses and wrote the manuscript. Ewa

B. Hoffmann (1) isolated RNA from the K- and M-livers, (2) developed the assay to measure

CYP2A6 mRNA levels, and (3) maintained HepG2 cells and two lots of human cryopreserved hepatocytes in culture. Abbas Assadzadeh (1) assisted in CYP2A6 genotyping, (2) assisted in

DNA sodium bisulfite treatment and subsequent PCR reactions, and (3) cloned all PCR products analyzed for DNA methylation. Bin Zhao performed the nicotine C-oxidation kinetic assay for all the K- and M- livers. Qian Zhou assisted in (1) DNA isolation from livers and (2) CYP2A6 genotyping.

133 Abstract

Purpose: We investigated the role of genetic, physiological, environmental, and epigenetic factors in regulating CYP2A6 expression and nicotine metabolism.

Methods: Human livers (n=67) were genotyped for CYP2A6 alleles and assessed for nicotine metabolism and CYP2A6 expression (mRNA and protein). In addition, a subset of livers (n=18), human cryopreserved hepatocytes (n=2), and HepG2 cells were used for DNA methylation analyses.

Results: Liver samples with variant CYP2A6 alleles had significantly lower CYP2A6 protein expression, nicotine C-oxidation activity, and affinity for nicotine. Female livers had significantly higher CYP2A6 protein and mRNA expression compared to male livers. Livers exposed to dexamethasone and phenobarbital had higher CYP2A6 expression and activity, however the difference was not statistically significant. Age and DNA methylation status of the

CpG island and a regulatory site were not associated with altered CYP2A6 expression and activity.

Conclusions: We identified genotype, gender, and exposure to inducers as sources of variation in CYP2A6 expression and activity, but much variation remains to be accounted for.

134 Introduction

Cytochrome P450 2A6 (CYP2A6) is primarily expressed in the liver and is involved in the metabolism of coumarin and a number of pharmaceuticals including halothane, valproic acid, tegafur, and SM-12502 (Pelkonen et al. 2000). In addition, this enzyme can activate procarcinogens such as aflatoxin B1 and tobacco-specific nitrosamines [e.g., 4- methylnitrosoamino-1-(3-pyridyl)-1-butanone and N-nitrosodiethylamine] (Pelkonen et al.

2000). Clinically, CYP2A6 is of significance due to its major role in the metabolism of nicotine, the main addictive compound in tobacco (Benowitz 2009b). Indeed, multiple genotypic and phenotypic studies have associated variable CYP2A6 with an altered risk for tobacco-related cancers and multiple smoking behaviors such as the risk of being a smoker, the number of cigarettes smoked, and the likelihood of smoking cessation (Ho and Tyndale 2007).

In vivo, the majority (70–80%) of nicotine is metabolized to cotinine (Benowitz and

Jacob 1994), and CYP2A6 mediates approximately 90% of this reaction (Nakajima et al. 1996a;

Messina et al. 1997). A common feature among in vivo studies assessing nicotine metabolism and CYP2A6 activity is the large interethnic and interindividual variability observed (Nakajima et al. 2006). Similarly, in human liver microsomes CYP2A6 mRNA, protein, and activity levels have been shown to vary >50-fold (Shimada et al. 1994; Rodriguez-Antona et al. 2001).

To date, 37 variant alleles of the CYP2A6 gene have been reported

(http://www.cypalleles.ki.se/cyp2a6.htm), many of which result in altered activity accounting for some of the observed variability in CYP2A6-mediated nicotine metabolism (Mwenifumbo et al. 2008). Although several of these alleles have been studied extensively in vivo and in in vitro cDNA expression systems, their effect on CYP2A6 expression in human livers has been assessed by few studies (Kiyotani et al. 2003; Haberl et al. 2005). To our knowledge the effect of CYP2A6 variant alleles on nicotine C-oxidation pharmacokinetic parameters (Vmax and Km)

135 in human livers has not yet been assessed, which is important as some CYP2A6 variants have substrate-specific impacts (Fukami et al. 2005). One aim of this study was to understand the mechanisms by which some of these variants are associated with altered activity in vivo. For example, do CYP2A6 variant alleles associated with lower in vivo activity, particularly for nicotine C-oxidation, encode proteins with lower protein expression, intrinsic activity, and/or apparent affinity for the substrate?

Gender and age have also been shown to influence nicotine pharmacokinetics in vivo.

Women have significantly higher rates of nicotine clearance compared to men (Benowitz et al.

2006a), while elderly (>65 years) subjects have significantly lower rates of nicotine clearance compared to younger adults (22–44 years) (Molander et al. 2001). The higher rates of nicotine clearance among females may be mediated by higher CYP2A6 activity as measured by the established in vivo phenotypic ratio of trans-3′-hydroxycotinine to cotinine (3HC/COT) (Ho et al. 2009). Another aim of this study was to investigate whether the gender- and age-related differences in nicotine metabolism were mediated by differences in metabolic factors (i.e., different CYP2A6 expression/activity). Environmental factors could also influence nicotine clearance. For example, grapefruit juice reduces nicotine metabolism by inhibiting CYP2A6

(Hukkanen et al. 2006), while drugs such as phenobarbital and dexamethasone induce CYP2A6

(Maurice et al. 1991).

While CYP2A6 genetic variants, gender, and environment explain some of the variation in CYP2A6 expression and activity, there still remains additional variation (Mwenifumbo et al.

2008). Another source of variation could be epigenetic regulation (Gomez and Ingelman-

Sundberg 2009). Epigenetic processes are heritable, or acquired, modifications of the DNA (i.e., methylation) or its associated proteins such as histones (e.g., histone acetylation) (Schumacher and Petronis 2006). DNA methylation has been shown to affect the tissue specific and general expression of several CYPs (e.g., CYP1B1, CYP1A2, CYP2E1, and CYP2W1) (Ingelman- 136 Sundberg et al. 2007). Recently, CYP2A13 was shown to be induced following the co- treatment of NCI-H441 cells with the demethylating agent 5-Aza-2′-deoxycitidine (5-AzaC) and the histone deacetylase inhibitor trichostatin A (Ling et al. 2007). Since CYP2A6 shares a 93.5% amino acid sequence identity with CYP2A13 it is possible that CYP2A6 is also epigenetically regulated. CYP2A6 has also been shown to contain a putative important CpG island, which suggests a possible role for DNA methylation in its regulation (Ingelman-Sundberg et al. 2007).

Our final aim was to investigate the potential role of DNA methylation in regulating CYP2A6 expression.

In the current study we have used a panel of human liver samples to assess the impact of

CYP2A6 genetic variants, gender, age, and exposure to inducers (phenobarbital and dexamethasone) on microsomal CYP2A6 levels and nicotine metabolic parameters (Vmax and

Km). In addition, a smaller panel of livers (n=18), human cryopreserved hepatocytes (n=2), and

HepG2 cells were utilized to investigate the potential role of DNA methylation in regulating

CYP2A6. In a linear regression model CYP2A6 genotype, gender, and exposure to inducers were associated with altered CYP2A6 expression and activity.

137 Material and Methods

Human Livers

The tissue samples studied here are a compilation from three liver banks: (1) 27 livers from the L-series (Messina et al. 1997), (2) 17 livers from the K-series (Campbell et al. 1987), and (3) 23 livers from the Biocentre in Basel, Switzerland (Meier et al. 1983). For clarity and consistency the livers from the Biocentre in Basel will be referred to as the M-series livers. The characteristics and sources of all the livers have been described previously (Meier et al. 1983;

Campbell et al. 1987; Messina et al. 1997). The ethnic origin of the liver sample donors was unknown, and the cause of death and drug use was known for some of the samples. Mean age of the organ donors was 30 years (range 2–64) and the gender distribution was as follows: 35 males, 26 females, and 6 unknown. Of note, the K- and M-livers were also assessed for

CYP2B6 and CYP2D6 protein expression (unpublished data). As a proxy measure for liver quality we did not observe any generalized reduction in the expression of all three CYPs among the different livers.

Cell culture

Human hepatocarcinoma cell line HepG2 cells were generously provided by Dr. David

Riddick (University of Toronto, ON, Canada). The HepG2 cells were maintained in a humidified incubator (37°C, 5%CO2) in MEM-α medium (Invitrogen, ON, Canada) supplemented with 10% fetal bovine serum (Invitrogen, ON, Canada). Two lots of human cryopreserved hepatocytes, lot SHM (male, 51 years old, low CYP2A6 mRNA and coumarin hydroxylase activity) and lot VEP (male, 56 years old, high CYP2A6 mRNA and coumarin hydroxylase activity), were purchased from Celsis In Vitro Technologies (MD, USA). The hepatocytes were maintained in InVitroGRO CP medium (Celsis In Vitro Technologies)

138 supplemented with Torpedo Mix (Celsis In Vitro Technologies) on collagen-coated plates (BD Biocoat plates) at 37°C and 5% CO2.

Nicotine C-oxidation kinetic assay

Liver microsomes were prepared by differential centrifugation as previously described

(Messina et al. 1997), and protein content was determined by the Bradford protein assay (Bio-

Rad Labs, ON, Canada). Nicotine C-oxidation kinetic parameters for the L-livers and two K- livers (K20 and K27) were measured previously by Messina et al. (1997). The remaining K- and

M-livers were assayed as previously described (Messina et al. 1997), with minor modifications.

A range of nicotine concentrations (50, 100, 200, and 500 µM) and 20 min incubation were used. One sample (K27) was assessed using both assays yielding similar results (Vmax = 28 vs.

23 nmol mg protein-1 h-1, Km=64 vs. 58 µM). In addition, the Vmax and Km values for K27 were similar to those reported by Messina et al. (1997) suggesting minimal effects of storage.

Other studies have demonstrated no significant decreases in CYP2A6 catalytic activity

(coumarin 7-hydroxylation) following 2 to 5 years of storage (Pearce et al. 1996; Yamazaki et al. 1997). The kinetic parameters (Vmax and Km) for all the livers (n = 67) were estimated using the Michaelis-Menten equation by the computer program GraphPad Prism (GraphPad

Software, CA, USA, version 5.0 for Windows). The mean ± SD Vmax values of the K-, L-, and

M-livers were all similar, indicative of comparable liver quality and preservation.

CYP2A6 genotype

DNA from all liver tissues (n = 67) was isolated by / (Invitrogen,

Canada) extraction and ethanol precipitation. The DNA samples were genotyped for the following CYP2A6 alleles: CYP2A6*1B, *2, *4, *5, *6, *9, *12, *14, *17, *20, *21, *23, *24,

*25, *26, *27, *28, *35, and the duplication *1×2 (Schoedel et al. 2004; Mwenifumbo et al.

2008; Ho et al. 2009; Al Koudsi et al. 2009a). The liver samples were grouped into two genotype groups: (1) wild-type livers (n = 48) and (2) variant livers (n = 17). The wild-type 139 group included samples positively genotyped for CYP2A6*1B, *14, and *21, but not any other allele. Those with CYP2A6*1B were included in the wild-type group for two main reasons.

First, CYP2A6*1B is a wild-type allele usually included in the wild-type reference group in the literature (Nakajima et al. 2006; Mwenifumbo et al. 2008), along with other variants of the

CYP2A6*1 allele (CYP2A6*1B-*1K, listed at http://www.cypalleles.ki. se/cyp2a6.htm). Second, in this study CYP2A6*1B did not influence CYP2A6 mRNA, protein, or activity levels.

CYP2A6*14 was included in the wild-type group since previous studies (Nakajima et al. 2006;

Mwenifumbo et al. 2008) and this current study did not find an effect on CYP2A6 expression or activity. To date, only two in vivo studies have assessed the impact of CYP2A6*21; one study showed no effect (Al Koudsi et al. 2006) and the other study associated it with lower activity

(Mwenifumbo et al. 2008). Due to the conflicting literature, and the fact that only one liver had a CYP2A6*21 allele, CYP2A6*21 was included in the wild-type group. Of note, the outcomes of all the analyses were not affected by either including or excluding CYP2A6*14 and CYP2A6*21 from the wild-type group.

CYP2A6 protein quantification

The L-livers and two K-livers (K20 and K27) were previously assessed for CYP2A6 protein levels (Messina et al. 1997). CYP2A6 protein levels for the remaining K- and M-livers were determined by Western blotting as described previously (Messina et al. 1997), with minor modifications. The amount of liver microsomal protein loaded was reduced to 5 µg instead of 30

µg. In addition, the primary (CYP2A6 monoclonal) and secondary (anti-mouse IgG horseradish peroxidase conjugate) antibodies were diluted 1:3,000 and 1:17,000, respectively. All other procedures were identical to those described by Messina et al. (1997). The CYP2A6 immunoreactivity levels of K20 and K27 obtained by Messina et al. (1997) were used as a normalization factor to provide a relative scale for all three liver banks. Thus, K20 and K27 were used as internal controls when quantifying CYP2A6 protein levels in the K- and M-livers. 140 Each sample was analyzed a total of 4 times, while the internal controls (K20 and K27) were analyzed 12 times. The CYP2A6 protein levels for K20, assessed by Western blotting, were similar to those reported by Messina et al. (1997), suggesting minimal effects of storage. Other studies have demonstrated no significant decreases in CYP2A6 microsomal protein levels following 2 to 5 years of storage (Pearce et al. 1996; Yamazaki et al. 1997).

RNA isolation, cDNA synthesis, and relative mRNA quantification

Tissues for only the K- and M-livers (n=40) were available for RNA analyses. Following homogenization of ∼0.1 g of liver tissue, total RNA was isolated with TRIzol reagent

(Invitrogen). RNA concentration was determined spectrophotometrically, and the integrity of the 18S and 28S ribosomal bands was confirmed by electrophoresis on a 1.2% agarose gel

(OnBio, ON, Canada) stained with ethidium bromide. cDNA from liver RNA was synthesized using 1 µg of total RNA, random hexamers (Invitrogen, RNA guard (Invitrogen), and M- MLV

Reverse Transcriptase (Invitrogen) according to the manufacturer’s protocols.

Primers for real-time PCR amplification of CYP2A6 and β-actin were as follows:

CYP2A6 forward primer (2A6ex2/3F): 5′-TCA AAG GCT ATG GCG TGG TA-3′ and reverse primer (2A6ex3/4R): 5′-CGA TAT TGG CGC CGC CA-3′; β-actin forward primer

(ACTBFex3): 5′-CAG AGC AAG AGA GGC ATC CT-3′ and reverse primer (ACTBRex4/3):

5′-GGT CTC AAA CAT GAT CTG GGT C-3′. Primer specificity was assessed using BLAST search.

Amplification and fluorescence detection were performed using the ABI 7500 Real-

Time PCR system (Applied Biosystems). Real-time PCR amplification was carried out in a mixture (25 µl) containing 1 µl of synthesized cDNA, 12.5 µl of 2× Power SYBR-Green Master

Mix (Applied Biosystems), and 0.3 µM of each primer. Cycling conditions consisted of two activation steps (50°C for 2 min, then 95°C for 10 min) followed by 40 cycles of melting (95°C for 15 s) and annealing/extension (60°C for 1 min). Relative quantifications of CYP2A6 gene 141 expression were obtained by normalizing to β-actin and using the comparative Ct method for relative quantification as described by the manufacturer (Real-Time PCR Chemistry Guide,

Applied Biosystems). All liver samples were assessed in triplicate.

CYP2A6 DNA methylation

Analysis of the CYP2A6 gene ± 10 kb revealed two CpG sites of potential interest: (1) a

CpG dinucleotide within a DR4-like element (−5476 bp) that has been shown to be involved in regulating CYP2A6 expression (Itoh et al. 2006) and (2) a CpG island in intron 2-exon 3 (1656–

1889 bp). DNA methylation status of these sites was determined in HepG2 cells, human cryopreserved hepatocytes (n=2), and a subset of liver samples (n=18). The livers selected had extreme CYP2A6 phenotypes (i.e., very high and low CYP2A6 mRNA, protein and activity levels, >50-fold variability), which was not accounted for by genotype as they had none of the

CYP2A6 variant alleles investigated.

DNA methylation status of the DR-4 site (8 CpGs) and the CpG island (25 CpGs) was evaluated using sodium bisulfite genomic sequencing. Genomic DNA was modified using the

EpiTect Bisulfite kit (Qiagen, ON, Canada) in accordance with the manufacturer’s protocol. The modified DNA was then subjected to a two-step semi-nested PCR reaction. Primers used to amplify the first PCR product of the DR-4 site and the CpG island were as follows: DR-4 site forward primer (2A6DRF2): 5′- GGT GAT ATA GTT TGG GTT TGT G-3′ and reverse primer

(2A6DRR2): 5′- AAC ATA AAA ATA CCT CAA CAT AC-3′; CpG island forward primer

(2A6CpGF4): 5′- GGG AGT TTT TTG GAG TTG T-3′ and reverse primer (2A6CpGR): 5′-

ACC TAA TCC CCA TCC C-3′. Primers used to amplify the second PCR product of the DR-4 site and the CpG island were as follows: DR-4 site forward primer 2A6DRF2 and reverse primer (2A6CpG5prR): 5′-ATC TCA ACT CAC TAC AAC CTC TA-3′; CpG-island forward primer (2A6PyroCpGF): 5′-TTT TTT AGG YGT GGT ATT TAG TAA-3′ and reverse primer

2A6CpGR. 142 Cycling conditions for the first amplification consisted of initial denaturation at 95°C for

3 min; 40 cycles of denaturation at 95°C for 30 s, annealing at 58°C for 30 s, and extension at

72°C for 30 s; followed by a final extension at 72°C for 5 min. The reaction mixture (25 µl) contained 100 ng of modified DNA, 0.2 mM of each dNTP, 0.5 mM MgCl2, 1× PCR buffer

(Sigma Aldrich, ON, Canada), 1.25 U of Jumpstart Taq DNA polymerase (Sigma Aldrich, ON,

Canada), and 0.125 µM of each primer. The cycling conditions for the second PCR reaction were identical to the first, except that the number of cycles was reduced to 25 cycles. The reaction mixture was similar to the first PCR reaction, except that the template was 0.8 µl of 10× diluted first amplification product. In addition, no MgCl2 was added and the final reaction volume was 50 µl.

PCR products were purified using PureLink PCR purification kit (Invitrogen), cloned into a pGEM-T vector system (Promega, PA, USA), and sequenced using the universal M13- forward primers. An average of eight to ten clones were sequenced per sample. Sequencing was carried out at Functional Biosciences (WI, USA). Specific amplification of CYP2A6 was confirmed by aligning the sequences of the PCR products with CYP2A6, CYP2A7, and

CYP2A13 using ClustalW2 (http://www.ebi.ac.uk/Tools/clustalw2/index.html). The BiQ

Analyzer software was used to perform quality control and derive % DNA methylation from the sequencing results (Bock et al. 2005).

Statistical analyses

Hardy-Weinberg equilibrium was tested using chi-squared test or Fisher’s exact test if five or fewer individuals were in one genotype group. The Kolmogorov-Smirnov and Shapiro-Wilk tests indicated that the phenotypic data (i.e., CYP2A6 mRNA levels, CYP2A6 protein levels,

CYP2A6 DNA methylation, Vmax, Km, and Vmax/Km) were not normally distributed.

Therefore, nonparametric tests were used in all statistical comparisons unless otherwise stated.

2 Correlation analyses were performed using the Spearman’s rank correlation test (rs =rank 143 correlation coefficient squared). Comparisons between two independent groups were carried out using the Mann-Whitney test. A linear regression model was formulated to evaluate which factors were associated with either CYP2A6 protein levels or Vmax. In this case the distribution of the dependent variables (CYP2A6 protein levels or Vmax) was log-normalized for use in the parametric regression analysis. By use of backward selection, the predictors (CYP2A6 genotype, gender, age, and exposure to inducers) were investigated for model inclusion. A predictor was considered to have a significant influence if P ≤ 0.05. The adjusted R2 value was used to assess the percentage of variation in the dependent variable that is accounted for by the overall modela.

The computed standardized Beta coefficients (Beta) for each predictor was used to measure the contribution of each predictor to the overall modela. To calculate the percentage of variation in the dependent variable that is accounted for by each predictor, the following formula was used:

(Beta value of a single predictor / sum of absolute Beta values of all predictors) x Adjusted R2 value of the total modela. The advantage of a linear regression model is that the effect of each predictor on the outcome is tested while controlling for other variables that might also be affecting the outcome. All statistical analyses were performed using SPSS (SPSS, IL, USA, version 15.0 for windows). Graphs were generated using GraphPad Prism (GraphPad Software,

CA, USA, version 5.0 for Windows).

a Sentences added following publication during thesis correction. 144 Results

Correlations among CYP2A6 mRNA, protein, and nicotine C-oxidation

Extensive interindividual variation in CYP2A6 mRNA levels, protein levels, and nicotine-to-cotinine metabolism (Vmax and Km) was observed. Relative CYP2A6 mRNA and protein levels varied more than 1,000- and 100-fold, respectively. The mean ± SD values for

Vmax, Km, and Vmax/Km were 30±26 nmol mg protein-1 h-1 (range 4.0–120), 62±28 µM

(range 16–162), and 0.6±0.5 nl mg protein-1 h-1 (range 0.04–2.6).

CYP2A6 mRNA correlated significantly with microsomal CYP2A6 protein (Figure

2 15a, rS =0.45, P<0.001). Immunoreactive CYP2A6 protein also correlated significantly with

2 nicotine C-oxidation activity (Vmax) (Figure 15b, rS =0.37, P<0.001). The correlation between

2 CYP2A6 mRNA levels and Vmax was less robust (n=40, rS =0.15, P = 0.02).

The impact of CYP2A6 genetic variation, gender, age, and exposure to inducers on

CYP2A6 expression and activity was assessed in two ways. First, consistent with previous literature publications, the effect of each factor was tested on its own. Second, we used a linear regression model in which the effect of each factor was tested while controlling for the other variables that might also be affecting the outcome. This kind of analysis was essential as many of the factors (i.e., genotype, gender, age, and exposure to inducers) were interrelated.

145

A B 10000 1000 Wildtype (n=30) Wildtype (n=48) Variant (n=8) Variant (n=17) Duplication (n=2) Duplication (n=2) s l 1000 2 r 2=0.37, p<0.001 ) e rs =0.45, p<0.001 s r v h e / l

n i A

e 100 t N o R 100 r p

m

g 6 m A / l 2 o P m Y 10 n C (

e x 10 v a i t m a l V e

R 1

0.1 1 0.01 0.1 1 10 0.001 0.01 0.1 1 10 Relative CYP2A6 Immunoreactivity Relative CYP2A6 Immunoreactivity

Figure 15 a, b Correlations among CYP2A6 protein levels, CYP2A6 mRNA levels, and nicotine C-oxidase activity. Open and closed circles denote livers in the wild-type and variant groups, respectively. Triangles denote livers positively genotyped for the duplication allele 2 (CYP2A6*1×2). A. CYP2A6 protein and mRNA levels correlated significantly (n=40, rS =0.45, P<0.001). The inset is a representative Western blot. Lane one is cDNA expressed CYP2A6 and the remaining lanes are different liver samples. B. CYP2A6 protein levels and 2 nicotine C oxidase activity correlated significantly (n=67, rS = 0.37, P < 0.001).

146 Impact of CYP2A6 genetic variants

The CYP2A6 variants investigated in this study accounted for some of the variability observed in microsomal CYP2A6 protein expression and nicotine pharmacokinetic parameters

(Vmax and Km) (Figure 16). No livers with CYP2A6*5, *6, *17, *20, *23, *24, *25, *26, *27,

*28, or *35 alleles were found. Livers with CYP2A6*2, *4, *9, *12, and *28 were identified and included in the variant group. Genotype frequencies did not deviate significantly from Hardy-

Weinberg equilibrium. Liver samples in the wild-type group (n=48) had significantly higher microsomal CYP2A6 protein expression compared to livers in the variant group (n=17) (Figure

16a; 0.90±0.90 vs. 0.40±0.40, P=0.006). However, relative CYP2A6 mRNA levels were similar between the two groups (Figure 16b; wild-type (n = 30): 291±298 vs. variant (n=8): 368±431,

P=0.8). Of note, the clustering of low CYP2A6 mRNA levels in some wild-type samples

(Figure 16b) is not likely to be related to the integrity of the RNA as these samples had relatively high mRNA levels of certain transcription factors (data not shown). The higher

CYP2A6 protein levels among the wild-type group (n=48) were associated with higher nicotine

C-oxidation activity (Vmax) compared to the variant group (n=17) (Figure 16c; 30±23 vs.

26±27 nmol mg protein-1 h-1, P=0.03). In addition, the mean apparent Km value was significantly lower among livers in the wild-type group (n=48) compared to the variant group

(n=17) (Figure 16d; 56±26 vs. 80±29, P=0.001). Finally, the higher Vmax and lower Km values among the wild-type group (n=48) resulted in higher mean Vmax/Km (catalytic efficiency) compared to the variant group (0.6±0.5 vs. 0.4±0.3 nl mg protein-1 h-1, P=0.002, data not shown).

147 A B p=0.006 10 10000

p=0.8 y t i s v l i 1000 t e c v a e l e

r A o 1 N n u R 100 m m

m 6 I

A 6 2 A P 2 Y P 10 C Y

e C

0.1 v i e t v a i l t e a l R

e 1 R

0.01 0.1 *1/*1 *1/*2 *1/*4 *1/*9 *1/*12 *1/*28 *2/*12 *9/*14 *1/*1x2 Wildtype Variant *1/*1 *1/*2 *1/*4 *1/*9 *1/*12 *9/*14 *1/*1X2 Wildtype Variant

CYP2A6 Genotype CYP2A6 Genotype C D 1000 p=0.03 400 p=0.001 ) r h / n i

e 100 t

o 100 r ) p

M g µ (

m / l m o K m n (

x 10 a m V

1 10 *1/*1 *1/*2 *1/*4 *1/*9 *1/*12 *1/*28 *2/*12 *9/*14 *1/*1x2 Wildtype Variant *1/*1 *1/*2 *1/*4 *1/*9 *1/*12 *1/*28 *2/*12 *9/*14 *1/*1x2 Wildtype Variant CYP2A6 Genotype CYP2A6 Genotype

Figure 16 A-D. Effect of CYP2A6 variants on CYP2A6 protein levels, CYP2A6 mRNA levels, nicotine C-oxidation activity, and theapparent affinity for nicotine. Open and closed circles denote livers in the wild-type and variant groups, respectively. Triangles denote livers positively genotyped for the duplication allele (CYP2A6*1×2). At least one sample with CYP2A6*1×2 also had other CYP2A6 variants, thus livers with CYP2A6*1×2 were not included in the group (wild-type vs. variant) comparisons. The horizontal line denotes the mean value in each group. Liver samples in the variant group had lower CYP2A6 protein expression (a), similar CYP2A6 mRNA levels (b), lower nicotine C-oxidation activity (Vmax) (c), and higher apparent Km (d) compared to the wild-type group.

148 Impact of gender

To minimize any potential bias due to known CYP2A6 genetic variants (Figure 16), gender analyses were restricted to livers from the wild-type group. Females (n=20) had significantly higher microsomal CYP2A6 protein levels compared to males (n=26) (Figure 17a;

1.3±1.2 vs. 0.6±0.5, P=0.03). In addition, females (n=11) had higher CYP2A6 mRNA levels compared to males (n=17) (Figure 17b; 381 ± 289 vs. 194 ± 281, P=0.03). Nicotine C-oxidation activity (Vmax) among females (n=20) was higher, although this did not reach statistical significance compared to that in males (n=26) (Figure 17c; 37±32 vs. 25±12 nmol mg protein-1 h-1, P=0.3). As expected the apparent Km was similar between females (n=20) and males (n=26)

(Figure 17d; 57±22 vs. 58±28 µM, P=0.6). Finally, Vmax/Km was not significantly higher among females (n=20) compared to males (n=26) (0.7±0.6 vs. 0.5±0.3 nl mg protein-1 h-1,

P=0.6, data not shown). When the variant livers were included in the gender analyses, the results remained the same, however the difference in CYP2A6 protein levels was no longer significant

(P=0.07).

149 A B p=0.03 10 10000 p=0.03 y t i

s 1000 v l i t e c v a e l

e r

1 A o N n u R 100 m m

6 m I

A 6 2 A P 2 Y

P 10 C

Y e C

v

0.1 i e t v a i l t e a l R

e 1 R

0.01 0.1 Females (n=20) Males (n=26) Females (n=11) Males (n=17) C D 1000 400 p=0.3 p=0.6 ) r h / n i 100 e t

o 100 r ) p

M g µ ( m

/ l m o K m n (

x

a 10 m V

1 10 Females (n=20) Males (n=26) Females (n=20) Males (n=26)

Figure 17 a-d. Impact of gender on CYP2A6 protein levels, mRNA levels, and nicotine pharmacokinetics. Females had significantly higher CYP2A6 A protein and B mRNA levels compared to males. C Females had higher but not significantly different mean Vmax value compared to males. D The apparent Km was similar between males and females. Two samples with unknown gender were not included in the analyses.

150 Impact of age

2 Among wild-type livers, age did not correlate with either CYP2A6 protein (rS =0.04,

2 -4 P=0.2, n=44) or Vmax (rS = 7.0×10 , P=0.9, n=44). A similar lack of correlation was observed when all the livers were analyzed (Figure 18).

Impact of CYP2A6 inducers

Dexamethasone and phenobarbital are known inducers of CYP2A6 (Maurice et al.

1991). Five liver samples exposed to these drugs (alone or in combination) had higher CYP2A6 protein (1.3±1.0 vs. 0.8±0.8, P=0.09), CYP2A6 mRNA (494±356 vs. 287±322, P=0.08), and

Vmax (42±28 vs. 29±26 nmol mg protein-1 h-1, P=0.06) compared to all the other livers, although this did not reach significance.

Modeling of CYP2A6 protein level and nicotine C-oxidation activity

The model developed for CYP2A6 protein levels and nicotine C-oxidation activity included the following predictors: CYP2A6 genetic variants, gender, age, and exposure to inducers. Only CYP2A6 genetic variants (P=0.006) and gender (P=0.008) were significant predictors of CYP2A6 protein levels. The final model accounted for 19% (11% CYP2A6 genotype and 8% gender) of the observed variability in CYP2A6 protein levels. The model developed for Vmax indicated that CYP2A6 genetic variants (P=0.03) and exposure to inducers

(P=0.04) were the only significant predictors. The final model accounted for 11% (6% CYP2A6 genotype and 5% inducer exposure) of the observed variability in Vmax.

151 A 10 r 2=0.04, p=0.1 s

y t i

v i t

c 1 a

e r o

n u m m

I 0.1 6 A 2

P Y C

e v i

t 0.01

a l

e R

0.001 0 10 20 30 40 50 60 70 Age

B 1000 ) 2 r rs =0.01, p=0.4 h / n i e t

o r p

g

m / l

o 100 m

n (

x a

m V

10

1 0 10 20 30 40 50 60 70 Age

Figure 18a, b. Impact of age on CYP2A6 protein levels and nicotine C-oxidation activity 2 (Vmax). Age did not significantly correlate with either A CYP2A6 protein levels (rS =0.04, 2 P=0.1, n=61) or B nicotine C-oxidation activity (Vmax) (rS =0.01, P=0.4, n=61). Six samples with unknown age were not included in the analyses. 152 DNA methylation status of human livers, human cryopreserved hepatocytes, and HepG2 cells

The DR-4 site studied herein contains a total of eight CpG dinucleotides, one of which is located where the nuclear receptor pregnane X receptor (PXR) and the peroxisome proliferator- activated receptor-γ coactivator (PGC-1α) bind and modulate CYP2A6 expression (Itoh et al.

2006). This specific CpG dinucleotide within the DR-4 site will be referred to as the transcription factor site (TF site). Among human liver samples, the average DNA methylation levels of the total DR-4 site and the specific TF site were 85% (range 65– 97%) and 84% (range

66–100%), respectively. Liver samples with extreme CYP2A6 phenotype (i.e., very high vs. low

CYP2A6 levels/activity) had similar levels of methylation at both the DR-4 site and the TF site

(Figure 19). Average DNA methylation level (mean ± 95% CI) of the DR-4 site (85±1.9%) was much higher compared to the methylation at the CpG island (5.0±1.1%) of the same liver samples (n=3). DNA methylation levels (mean ± 95% CI) of the CpG island were similar between liver samples with high (n=1) and low (n=2) CYP2A6 levels/activity (5.3± 2.2 vs.

4.9±1.3%, P=0.9).

153 Eur J Clin Pharmacol

Fig. 5 a, b DNA methylation status of the DR-4 and TF sites among liver samples with either very high or very low CYP2A6 levels/activity. Each repre- sents a single liver sample. The error bars represent the 95% CI in methylation deduced from an average of 8–10 clones per sample. a Male liver samples with very high (n=4) and low (n=6) CYP2A6 levels/activity had similar levels of DNA methylation (mean ± 95% CI) at the DR site (86±4 vs. 88±4%, P=0.2) and the TF site (81±10 vs. 88±7%, P=0.2). b Female liver samples with very high (n=4) and low (n=4) CYP2A6 levels/activity had similar levels of DNA methylation (mean ± 95% CI) at the DR site (83±4 vs. 83±4%, P=0.6) and the TF site (86±8 vs. 80±9%, P=0.3)

Figure 19 a, b. DNA methylation status of the DR-4 and TF sites among liver samples with either very high or very low CYP2A6 levels/activity. Each bar represents a single liver sample. The error bars represent the 95% CI in methylation deduced from an average of 8–10 clones per sample. a Male liver samples with very high (n=4) and low (n=6) CYP2A6 levels/activity had similar levels of DNA methylation (mean ± 95% CI) at the DR-4 site (86±4 vs. 88±4%, P=0.2) and the TF site (81±10 vs. 88±7%, P=0.2). b Female liver samples with very high (n=4) and low (n=4) CYP2A6 levels/activity had similar levels of DNA methylation (mean ± 95% CI) at the DR-4 site (83±4 vs. 83±4%, P=0.6) and the TF site (86±8 vs. 80±9%, P=0.3). 154

Fig. 6 Comparison of DNA methylation levels at the CpG island, CpG island DNA methylation (mean ± 95% CI) compared to human DR-4 site, and TF site between human liver samples and HepG2 cells. livers (n=3) (65±7.4 vs. 5.0±1.1%, P<0.001). In addition, HepG2 Each bar represents the average methylation level, while the error cells had significantly lower levels of DNA methylation compared to bars represent the 95% CI in methylation deduced from an average of human livers (n=18) at both the DR-4 site (14±12 vs. 85±1.9%, P< 8–10 clones per sample. HepG2 cells had significantly higher levels of 0.001) and the TF site (14±35 vs. 84±4.2%, P<0.001) Eur J Clin Pharmacol

Fig. 5 a, b DNA methylation status of the DR-4 and TF sites among liver samples with either very high or very low CYP2A6 levels/activity. Each bar repre- sents a single liver sample. The error bars represent the 95% CI in methylation deduced from an average of 8–10 clones per sample. a Male liver samples with very high (n=4) and low (n=6) CYP2A6 levels/activity had similar levels of DNA methylation (mean ± 95% CI) at the DR site (86±4 vs. 88±4%, P=0.2) and the TF site (81±10 vs. 88±7%, P=0.2). b Female liver samples with very high (n=4) and low (n=4) CYP2A6 levels/activity had similar levels of DNA methylation (mean ± 95% CI) at the DR site (83±4 vs. 83±4%, P=0.6) and the TF site (86±8 vs. 80±9%, P=0.3) HepG2 cells, which express very low levels of CYP2A6, had an opposite pattern of

methylation compared to human livers (i.e., high methylation in the CpG island and low

methylation in the DR-4 and TF sites, Figure 20). The cryopreserved hepatocyte lot with high

CYP2A6 mRNA and activity had higher levels of methylation (mean ± 95% CI) at the CpG

island (53±5.0%) compared to the sample with low CYP2A6 mRNA activity (5.5 ± 2.5%). The

overall methylation (mean ± 95% CI) of the DR-4 site was similar between the two

cryopreserved hepatocytes (86 ± 2.8 vs. 89 ± 3.7%). However, the specific CpG dinucleotide

within the PXR/ PGC-1α binding site (i.e., TF site) was 0% methylated in the hepatocyte lot

with high CYP2A6 mRNA and activity compared to 100% in the lot with low CYP2A6 mRNA

and activity.

Figure 20. Comparison of DNA methylation levels at the CpG island, DR-4 site, and TF site Fig. 6 Comparison of DNAbetween methylation human liver levels samples at the and CpG HepG2 island, cells. EachCpG bar island represents DNA methylation the average (mean methylation ± 95% CI) compared to human DR-4 site, and TF site betweenlevel, while human the liver error samples bars represent and HepG2 the 95% cells. CI in methylationlivers (n=3) deduced (65±7.4 from vs. an 5.0±1.1%, average of 8P–<0.001).10 In addition, HepG2 Each bar represents theclones average per methylation sample. HepG2 level, cells while had thesignificantlyerror highercells levels had significantly of CpG island lower DNA levels methylation of DNA methylation compared to bars represent the 95% CI(mean in methylation ± 95% CI) deducedcompared from to human an average livers of(n = 3)human (65 ± livers 7.4 vs. (n 5.0=18) ± at 1.1%, both P the < 0.001).DR-4 site In (14±12 vs. 85±1.9%, P< 8–10 clones per sample. HepG2addition, cells HepG2 had significantlycells had significantly higher levels lower of levels0.001) of DNA and methylation the TF site (14±35compared vs. to 84±4.2%, human P<0.001) livers (n=18) at both the DR-4 site (14±12 vs. 85±1.9%, P< 0.001) and the TF site (14 ± 35 vs. 84 ± 4.2%, P < 0.001).

155 Discussion

Consistent with previous studies (reviewed in (Pelkonen et al. 2000)), we observed substantial interindividual variation in CYP2A6 expression and activity. The significant correlation between CYP2A6 mRNA levels and CYP2A6 protein levels suggests an important role for pre-translational regulation in mediating CYP2A6 protein expression.

One focus of this study was to test the impact of CYP2A6 genetic variants on CYP2A6 expression and nicotine C-oxidation pharmacokinetics in liver tissue. The low gene deletion

(CYP2A6*4) allele frequency and absence of alleles predominantly present among individuals of African origin (e.g., CYP2A6*17) (Nakajima et al. 2006) suggested that our population is likely of Caucasian origin. Livers in the variant group (CYP2A6*2, *4, *9, *12, and *28) had significantly lower CYP2A6 protein expression compared to the wild-type group. This suggests lower CYP2A6 protein expression as one likely mechanism by which these alleles are associated with lower in vivo CYP2A6 activity. CYP2A6*2 encodes an enzyme that is unable to incorporate heme, making it unstable and more likely to be degraded (Yamano et al. 1990).

CYP2A6*4 is a gene deletion that results in lower transcript levels and consequently lower protein expression. CYP2A6*9 contains a SNP (−48T>G) in the TATA-box that has been associated with lower CYP2A6 mRNA and activity in human livers (Kiyotani et al. 2003).

CYP2A6*12 is a hybrid allele of CYP2A6 and CYP2A7 in which the 5′-region and exons 1 and 2 are of CYP2A7 origin while exons 3 onwards are of CYP2A6 origin. In vitro, COS-1 cells expressing CYP2A6.12 had approximately 50% lower immunodetectable CYP2A6, suggesting that it is an unstable enzyme that is more rapidly degraded (Oscarson et al. 2002). CYP2A6*28 is a haplotype (N418D and E419D) that has been associated with lower CYP2A6 activity in vivo

(Mwenifumbo et al. 2008). It is possible that the two non-synonymous SNPs in CYP2A6*28 might result in an unstable enzyme that is more readily degraded, however this remains to be

156 tested. Alternatively, the coding SNPs in CYP2A6*28 might be in linkage disequilibrium with polymorphisms in the gene promoter or within introns that affect transcription and mRNA splicing, and as a result might influence protein expression (Mwenifumbo et al. 2008).

Although the variant group had lower CYP2A6 protein levels compared to the wild-type group, CYP2A6 mRNA levels were similar. It is possible that the variants might be affecting

CYP2A6 protein levels at the translational or post-translational stage (e.g., enzyme degradation and/or stabilization) as opposed to affecting transcription or mRNA stability. This is thought to occur with CYP2A6*2. On the other hand a correlation between the mRNA and protein levels is expected for CYP2A6*4 and CYP2A6*9. CYP2A6*4 (gene deletion) mRNA should not be amplified in our real-time PCR reaction, while CYP2A6*9 has been associated with reduced transcriptional efficiency (Pitarque et al. 2001). We observed a trend for lower mRNA transcript levels among livers with CYP2A6*12. Of note, our CYP2A6 primers used in the real-time PCR reaction are able to amplify the CYP2A6*12 mRNA. An association of CYP2A6*12 with lower mRNA levels has been previously observed and suggests that this allele is regulated at both the transcriptional and translational level (Haberl et al. 2005).

Consistent with the correlation between CYP2A6 protein levels and Vmax, the variant group had a lower mean Vmax compared to the wild-type group. In terms of the effect of genotype on the apparent Km (i.e., affinity) for nicotine, the variant group had a significantly higher mean Km value (i.e., lower affinity) compared to the wild-type group. More specifically, samples with CYP2A6*2 or CYP2A6*12 had relatively higher mean Km values compared to the wild-type, while samples with CYP2A6*4 or CYP2A6*9 had roughly a similar mean Km value compared to the wild-type. This was anticipated since Km is an innate characteristic of an enzyme that is altered by structural changes (e.g., CYP2A6*2 and *12) but not by changes in expression levels (e.g., CYP2A6*4 and *9).

157 Among wild-type livers, female livers had approximately 2.0-, 1.9-, and 1.5-fold higher levels of CYP2A6 protein, CYP2A6 mRNA, and nicotine C-oxidation activity, respectively, compared to males. Although CYP2A6 expression (mRNA and protein) was significantly higher among female livers, the differences in nicotine C-oxidation activity did not reach statistical significance. This was likely due to the high variability in nicotine C-oxidation activity and the lack of absolute correlation between CYP2A6 levels and nicotine C-oxidation activity. Our findings suggest that the higher in vivo nicotine clearance rates (Benowitz et al. 2006a) and

CYP2A6 activity (Ho et al. 2009) observed among females are likely due to higher hepatic

CYP2A6 protein expression. Pregnancy and oral contraceptive use have been shown to further increase the rates of nicotine clearance suggesting a hormonal effect on CYP2A6 (Hukkanen et al. 2005). Indeed, CYP2A6 is induced by estrogen in an estrogen receptor-alpha (ER-α)- dependent manner (Higashi et al. 2007). It is important to note that the concentration of estrogen

(10 nM) used to induce CYP2A6 in vitro by Higashi et al. (2007) is at least 10-fold greater than the circulating plasma estrogen levels in premenopausal women (0.1-1.0 nM) and 5-fold lower than the peak estrogen levels (58 nM) in pregnant women (Ganong 2001)a. Whether estrogen is capable of inducing CYP2A6 at premenopausal physiological levels (0.1-1.0 nM) has not been tested yeta. However, unpublished studies in Dr. Matthew’s laboratory (University of Toronto) suggest that maximal recruitment of ER-α to its DNA response elements in human breast carcinoma cells (T-47D) occurs following treatment with a physiologically relevant (1.0 nM) estrogen concentration (Raymond Lo, personal communication)a. If this also occurs in human hepatocytes it would suggest that the higher hepatic CYP2A6 expression in women might, at least in part, be mediated by estrogena. The lack of an association between gender and CYP2A6 protein expression in previous studies (Shimada et al. 1994; Messina et al. 1997; Parkinson et al.

2004) is likely due to the fact that these studies did not genotype for CYP2A6 variants. As we a Sentences added following publication during thesis correction. 158 have demonstrated, CYP2A6 genetic variants have a significant effect on CYP2A6 expression, which might mask gender differences.

There was no significant correlation between age and CYP2A6 expression/activity. This is consistent with previous human liver studies (Shimada et al. 1994; Parkinson et al. 2004). In contrast, an in vivo nicotine pharmacokinetic study found that elderly (>65 years) subjects had significantly lower rates of nicotine clearance compared to younger adults (22–44 years)

(Molander et al. 2001). It is possible that we did not find an association between age and

CYP2A6 levels as our age range was below that of 65 years; our oldest liver was from a 64- year-old organ donor. Alternatively, the lower in vivo nicotine clearance rates among elderly individuals could be due to changes in other factors associated with age such as decrease in liver mass and blood flow (Kinirons and O'Mahony 2004).

When all the factors (CYP2A6 genetic variants, gender, age, and exposure to inducers) were included in a model to predict variability in CYP2A6 protein levels, only CYP2A6 variants and gender were significant predictors. The final model accounted for only 19% of the observed variability in CYP2A6 levels, suggesting there still remains a large amount of variability that is unaccounted for. We cannot rule out the contribution of other unidentified genetic variants. In addition, variability in the expression of nuclear transcription factors that regulate cytochrome

P450 (CYP) expression could also affect CYP2A6 levels. For example, CYP2A6 mRNA levels have been shown to significantly correlate with mRNA levels of the nuclear receptors CAR and

HNF4-α (Wortham et al. 2007). We were able to capture a smaller amount of variation (11%) in nicotine C-oxidation activity (Vmax). This is likely due to the fact that other enzymes

(CYP2B6) might also play a small role in this metabolic pathway (Yamazaki et al. 1999). In addition, variation in CYP activity could also be mediated by variation in the activity of CYP cofactors (e.g., NADPH-cytochrome P450 oxidoreductase) (Hart et al. 2008a). The NADPH- cytochrome P450 oxidoreductase (POR) transfers electrons that are essential for CYP activity 159 and therefore might be a rate-limiting factor for CYP mediated reactionsa. Liver-specific POR knockout mice have a profound reduction in their hepatic drug metabolism capacity (Gu et al.

2003; Henderson et al. 2003)a. In humans, many coding and noncoding SNPs have been identified in the POR gene (POR) (http://www.cypalleles.ki.se/por.htm, accessed July 14,

2010)a. Some of these variants have been associated with altered hepatic enzymatic activities of major CYPs including CYP2A6, CYP2D6, CYP1A2, and CYP3A4 (Agrawal et al. 2008; Hart et al. 2008a; Gomes et al. 2009a)a. For example, the allele POR*28 was recently shown to be associated with higher clearance in vivo and it better predicted the variability in midazolam clearance compared to genetic variants in CYP3A4, CYP3A5, and CYP3A7 (Oneda et al. 2009)a. Seventeen samples from our liver bank (K-series livers) have previously been assessed for POR protein and activity (cytochrome c-reduction) levels (Ramji 1998; Ramji et al.

2003)a. Among the 17 K-series livers we did not observe a significant correlation between rates

2 of cotinine formation and POR activity (n=17, rS =0.02, P = 0.6), suggesting that POR does not contribute to the variability in cotinine formation at least within the K-series liversa.

In our pilot epigenetic investigation we did not observe a prominent role for DNA methylation in regulating CYP2A6 expression in human livers. Nonetheless, two interesting but preliminary results were observed. First, HepG2 cells, which express low levels of CYP2A6, had an opposite pattern of DNA methylation compared to liver cells, suggesting a possible role for DNA methylation in reducing CYP2A6 expression within this cell line. We treated the

HepG2 cells with the demethylating agent 5-AzaC using a previously published paradigm

(Dannenberg and Edenberg 2006) but did not observe any changes in its DNA methylation status. The second interesting observation was that the cryopreserved hepatocyte lot with high

CYP2A6 activity was fully demethylated at the TF site while the low activity lot was fully methylated. The TF site is a CpG dinucleotide within a DR-4-like element to which PXR/PGC- a Sentences added following publication during thesis correction. 160 1α is thought to bind and modulate CYP2A6 expression (Itoh et al. 2006). Thus, the lack of methylation might make the site more accessible to the transcription factors resulting in higher expression/activity. It is important to note that the lack of an association between CYP2A6 expression and DNA methylation among the human livers is not evidence that CYP2A6 is not epigenetically regulated. Our investigation was limited to a small number of livers and only two

DNA regions; there are many CpG-dense sites within the CYP2A6 gene that might be involved in epigenetic regulation. Recently, the allele specific expression of CYP2A7 was reported to be dependent on allele haplotype sequence and its methylation status, suggesting an interesting combinatorial effect between sequence variation and methylation status (Kerkel et al. 2008). In addition, CYP2A13 expression has been shown to be induced by epigenetic modulators (Ling et al. 2007). Given that CYP2A6 is highly homologous to CYP2A7 (96.5% identity) and CYP2A13

(93.5% identity), it is likely that CYP2A6 might be epigenetically regulated, perhaps at other loci.

In conclusion we have shown that CYP2A6 genetic variants markedly influence the expression of CYP2A6 and nicotine pharmacokinetics in human liver microsomes. Moreover, our results indicate that the higher in vivo nicotine clearance rates and CYP2A6 activity observed among females are likely due to higher hepatic CYP2A6 protein expression. Finally, our study suggests that even with our current state of knowledge, unexplained variation in

CYP2A6 expression and activity still remains. Identifying sources of interindividual variation will help refine the relationship between CYP2A6 and phenotypic measures and improve our understanding of the influence of CYP2A6 on clinical outcomes.

161 Acknowledgments

We thank Bin Zhao, Qian Zhou, Fariba Baghai Wadji, Sharon Miksys, Linda Liu, and

Amandeep Mann for their technical assistance and Dr. Arturas Petronis for his valuable input in planning and analyzing the epigenetic experiments. This work was supported by the Centre for

Addiction and Mental Health and Canadian Institutes for Health Research (CIHR) MOP86471.

N.K. receives funding from CIHR-Strategic Training Program in Tobacco Use in Special

Populations (TUSP) and Ontario Graduate Scholarship program (OGS). R.F.T. holds a Canada

Research Chair in Pharmacogenetics.

162 Significance of Chapter

This was the first study to utilize a liver bank in order to elucidate the possible mechanisms by which 1) genetic variations in CYP2A6, 2) gender, and 3) age are associated with altered nicotine metabolism in vivo. In addition, this was the first pilot study to investigate the potential role of epigenetics (DNA methylation) in regulating CYP2A6 expression and activity. The advantage of using a liver bank is that the enzyme(s) are in their innate environment in contrast to heterologous expression systems (e.g. bacteria, yeast, or mammalian cell lines).

The significance of this chapter to the thesis is as follows. We identified a potential important role for transcriptional mechanisms in regulating CYP2A6. This suggests that variability in regulatory factors such as transcription factors (e.g. CAR and HNF-4α) should be considered when studying variability in CYP2A6. With respect to genetics, we determined that the lower rates of nicotine metabolism observed among individuals with CYP2A6 genetic variants are likely due to a combination of lower CYP2A6 expression (mRNA or protein) and an altered structural conformation of the enzyme (i.e. higher Km), depending on the specific type of variant. Moreover, the higher rate of nicotine clearance observed in females compared to males is likely mediated by a higher expression of CYP2A6 protein via a transcriptional mechanism. This suggests that females are likely to have higher rates of metabolism of other

CYP2A6 substrates (e.g. coumarin). Age did not seem to influence CYP2A6 expression suggesting that the lower rates of nicotine metabolism observed among elderly individuals (>65 years) is likely mediated by factors associated with aging (e.g. reduced liver mass and blood flow). Finally, in our pilot study we did not find an association between DNA methylation and regulation of CYP2A6 expression (mRNA and protein) in human livers. However, our analysis was limited to a small number of livers and a few CpG sites.

163 Our study suggests that even with our current state of knowledge, unexplained variation in CYP2A6 expression and nicotine C-oxidation still remains. Continuing to identify sources of interindividual variation will help refine the relationship between CYP2A6 and phenotypic measures and improve our understanding of the influence of CYP2A6 on clinical outcomes.

164 Chapter 4: Hepatic CYP2B6 is altered by genetic, physiologic, and environmental factors but plays little role in nicotine metabolism

Nael Al Koudsi and Rachel F. Tyndale

Reprinted with the permission from Taylor & Francis. This chapter appears as published in

Xenobiotica, 40(6): 381-92, 2010 with modifications to the figure and table numbers.

Dr. Rachel F. Tyndale and Nael Al Koudsi designed the experimental study. Liver samples and data from a few experiments in Chapter 3 are used in this chapter. The contribution of Nael Al

Koudsi and others to this study is as follows. Nael Al Koudsi (1) genotyped all the samples for the different CYP2B6 alleles, (2) determined CYP2B6 protein levels in the livers, (3) performed data analyses and (4) wrote the manuscript.

165 Abstract

1. Human cytochrome P450 2B6 (CYP2B6) is predominantly expressed in the liver and it plays a major role in the metabolism of several therapeutically important drugs and environmental toxicants.

2. The objective was twofold: (1) to determine the role of genetic, physiological, and environmental factors in predicting hepatic CYP2B6 protein expression; and (2) to investigate the role of CYP2B6 in nicotine C-oxidation.

3. Human livers (n = 40) were assessed for CYP2B6 protein and genotype.

4. Linear regression analyses indicated that CYP2B6 genotype (10%), gender (14%), and exposure to inducers (21%), but not age, were predictors of CYP2B6 protein amounts. Livers with at least one CYP2B6*5 or *6 allele were associated with lower CYP2B6. Female livers and livers exposed to inducers (phenobarbital and/or dexamethasone) were associated with higher

CYP2B6.

5. A weak correlation between CYP2B6 and nicotine C-oxidation activity was observed, which was abrogated when controlling for CYP2A6 protein levels. CYP2B6*6 was not associated with different nicotine kinetics.

6. In summary, CYP2B6 protein expression was associated with genotype, gender, and exposure to inducers, but not with nicotine C-oxidation activity.

166 Introduction

Cytochromes P450 (CYPs) are a family of heme-containing enzymes that mediate the metabolism of a wide variety of exogenous and endogenous substrates (Nelson et al. 1996). In humans, the CYP2B subfamily consists of only one functional protein CYP2B6 (Hoffman et al.

2001). CYP2B6 is thought to play a role in the complete or partial metabolism of numerous

(greater than 50) substrates (Turpeinen et al. 2006) that include the procarcinogen aflatoxin B1

(Aoyama et al. 1990), the antineoplastic agent cyclophosphamide (Roy et al. 1999), the drug of abuse methylenedioxymethamphetamine (MDMA, ‘ecstasy’) (Kreth et al. 2000), the narcotics ketamine (Yanagihara et al. 2001) and propofol (Court et al. 2001), the antiretroviral efavirenz

(Ward et al. 2003), and the anti-smoking cessation drug bupropion (Faucette et al. 2000).

Initially hepatic CYP2B6 was thought to be expressed in low levels and present in only a fraction of the total population (Mimura et al. 1993; Shimada et al. 1994). However, recent studies utilizing more specific and selective antibodies for CYP2B6 have indicated a greater incidence and quantity of CYP2B6 protein in the liver (Stresser and Kupfer 1999; Hesse et al.

2004). Substantial interindividual and interethnic variation has been observed in the expression and activity of CYP2B6 in human livers (Lamba et al. 2003; Parkinson et al. 2004). Because variation in CYP2B6 expression/activity can result in altered therapeutic or toxic responses to its substrates, it is important to identify the sources of the variable CYP2B6 expression/activity.

CYP2B6 is highly polymorphic (see (Zanger et al. 2007) for a comprehensive review).

To date, there are 29 numbered CYP2B6 alleles, some of which result in either no change, increased, decreased, or loss of activity. Many of the alleles are rare but some are more common. For example, CYP2B6*6 (Q172H and K262R) occurs at an allelic frequency ranging from 14% to 62% among different world populations (Zanger et al. 2007). CYP2B6*6 has been associated with lower CYP2B6 protein expression, accounting for some of the observed

167 interindividual variability in hepatic CYP2B6 protein expression (Desta et al. 2007). In addition to genetic factors, non-genetic factors influence CYP2B6 expression. These include environmental influences, with the classical example being the induction of CYP2B6 by inducers such as phenobarbital (Sueyoshi et al. 1999) and dexamethasone (Strom et al. 1996).

The influence of gender and age remains controversial. While some studies have found no influence of gender on CYP2B6 expression (Shimada et al. 1994; Stresser and Kupfer 1999;

Lang et al. 2001; Parkinson et al. 2004; Desta et al. 2007), Lamba et al. (2003) have reported higher levels of CYP2B6 mRNA (3.9-fold), protein (1.7-fold), and activity (1.6-fold) among female livers. In terms of age, two studies reported higher amounts of CYP2B6 protein expression during post-infancy compared with infancy (Tateishi et al. 1997; Croom et al. 2009), while other studies have found no influence of age (range = 2–75 years) on hepatic CYP2B6 protein expression (Shimada et al. 1994; Stresser and Kupfer 1999; Lang et al. 2001; Parkinson et al. 2004; Desta et al. 2007). A limitation of many previous studies assessing variability in

CYP2B6 expression includes the lack of appropriate methods to control for variables that could be interrelated (for example, genotype, age, gender, and inducer exposure). Our first aim was to use a linear regression model to investigate the influence of genetic, physiological, and environmental factors on hepatic CYP2B6 protein expression. The advantage of this approach is that the effect of each factor on the outcome (that is, CYP2B6 expression) is tested while statistically controlling for the other variables (for example, genotype and/ or induction).

The second aim of this paper was to investigate the potential role of CYP2B6 in nicotine metabolism; more specifically, whether CYP2B6*6 is associated with different nicotine kinetics in a human liver bank. Nicotine is the primary addictive substance in tobacco smoke (Benowitz

2009b) and some studies have reported an association between the rate of nicotine metabolic inactivation to cotinine and smoking behaviours (for example, risk of being a smoker, number of cigarettes smoked, and ability to quit) (Ho and Tyndale 2007). The majority (70–80%) of 168 nicotine absorbed in vivo is metabolized to cotinine in a reaction that is mostly (approximately

90%) mediated by CYP2A6 (Benowitz and Jacob 1994; Nakajima et al. 1996a; Messina et al.

1997). Compared with CYP2A6, CYP2B6 metabolizes nicotine with a lower affinity and may play a greater role among individuals who have genetically slow/absent CYP2A6 (Yamazaki et al. 1999). For example, Ring et al. (2007) reported an association between CYP2B6*6 and faster rates of nicotine metabolism, especially among individuals with genetically determined reduced

CYP2A6 activity. However, a more recent study did not find any differences in nicotine plasma levels obtained from nicotine patch between individuals with, or without, CYP2B6*6 even when the population was stratified by CYP2A6 genotype (Lee et al. 2007). One limitation of such in vivo association studies is that the contribution of CYP2A6 to nicotine metabolism in the different CYP2B6 genotype groups is unaccounted for. We are able to mitigate this impact by controlling for CYP2A6 protein in this in vitro study.

169 Materials and Methods

Human livers

The liver tissue samples (n=40) studied here were recently assessed for CYP2A6 protein expression and nicotine C-oxidation activity (Al Koudsi et al. 2009). The livers have also been assessed for CYP2D6 protein expression (unpublished data). As a proxy measure for liver quality we did not observe any generalized reduction in the expression of all three CYPs

(CYP2A6, CYP2D6, and CYP2B6) among the different livers. The characteristics and sources of the livers have been previously described (Meier et al. 1983; Campbell et al. 1987; Messina et al. 1997). The ethnic origins of the liver samples were unknown; however, an extensive analysis of CYP2A6 genotype indicates a predominant Caucasian population (Al Koudsi et al. 2009). The cause of death was known for some of the samples. The drug history for each liver sample has been previously described (Meier et al. 1983; Campbell et al. 1987; Messina et al. 1997).

Briefly, 27 of the 40 liver samples were exposed to drugs including two known CYP2B6 inducers, phenobarbital and dexamethasone (n=5/27), as well cimetidine, amitriptyline, , , clonidine, dopamine, ranitidine, labetalol, furosemide, , lidocaine, mannitol, vasopressin, insulin, propranolol and thiopental. The impact of most of these drugs on CYP2B6 protein expression was not possible to assess as only one or two livers were exposed to each drug. However the five livers exposed to known CYP2B6 inducers phenobarbital and dexamethasone had significantly higher CYP2B6 expression. Mean age of the organ donors was 35 years (range=13–64) and the gender distribution was as follows:

23 males, 13 females, and four unknown.

CYP2B6 genotyping

DNA from liver tissues was isolated by phenol/chloroform (Invitrogen, Canada) extraction and ethanol precipitation. DNA samples from all the livers were genotyped for

170 CYP2B6*4 (K262R), *9 (Q172H), and *6 (Q172H and K262R) using a previously described haplotyping assay (Lee et al. 2007a). In addition, the samples were genotyped for the 1459C>T

SNP (R487C) in CYP2B6*5 (when present alone) or in CYP2B6*7 (when present in haplotype with Q172H and K262R). A novel three-step assay was developed to detect the SNP 1459C>T

(R487C). As an additional precaution to avoid the amplification of CYP2B pseudogenes the first two steps of the assay included gene-specific amplifications. In the first step, CYP2B6 (1063 bp) was specifically amplified from intron 8 to exon 9 utilizing the forward and reverse primers

2B6*51F1: 5′-GGCTAGCCTGGCCAATATGATGAT-3′ and 2B6*51R: 5′-

ATTAGCCAAGCGTGGTAGTGCATG-3′, respectively. The 25 µl reaction mixture contained:

1× PCR buffer (10 mM Tris pH 8.8, 50 mM KCl), 0.2 mM of each dNTP, 1.1 mM MgCl2, 0.25

µM of each primer, 1.25 U of Taq Polymerase (Fermentas, Life Sciences, Burlington, ON,

Canada), and 50 ng of genomic DNA. The cycling conditions were as follows: initial denaturation at 94°C for 1 min followed by 30 cycles each consisting of denaturation at 94°C for 20 s, annealing at 61°C for 40 s, and extension at 72°C for 2 min followed by a final extension at 72°C for 7 min.

The second step was also a CYP2B6 gene specific PCR, in which a fragment of exon 9 in

CYP2B6 (206 bp) was amplified. Two artificial restriction sites BglII and AlwI were introduced by the forward and reverse primers 2B6*52F-BGL 5’-

AGAACTTCTCCATGGCCA*GATCTGTGGCCCCA-3’ and 2B6*52R-ALW 5’-

AGGCAGGAAGTTGCGGGGGATCAGAGC*CATTG-3’, respectively. The underlined nucleotides have been introduced to create the restriction sites and the asterisk (*) indicates the exact cutting location. Introduction of the artificial restriction sites by the forward and reverse primers ensured that incomplete digestion by either enzyme is not a limiting factor in determining the genotype. The 25 µl reaction mixture contained: 1× PCR buffer (10 mM Tris pH 8.8, 50 mM KCl), 0.1 mM of each dNTP, 1.2 mM MgCl2, 0.25 µM of each primer, 0.3 U of 171 Taq Polymerase (Fermentas, Life Sciences), and 0.8 µl of undiluted PCR product from the first reaction. The cycling conditions were as follows: initial denaturation at 94°C for 1 min followed by 30 cycles each consisting of denaturation at 94°C for 30 s, annealing at 68°C for 45 s, and extension at 72°C for 2 min followed by a final extension at 72°C for 7 min.

In the final third step, restriction fragment length polymorphism (RFLP) was used to detect the 1459C>T SNP. A map of the restriction sites and the predicted fragment sizes resulting from the digestion of the second step PCR product is shown in Figure 21(A). A total of three reactions were performed for each sample: BglII, AlwI, and uncut. Each BglII digestion reaction contained 3 µl of the second step PCR product, 2 µl of the 10× NE buffer 3 and 20 U of

BglII in a total volume of 20 µl. Each AlwI digestion reaction contained 3 µl of the second step

PCR product, 2 µl of the 10× NE buffer 4 and 5 U of BglII in a total volume of 20 µl. The uncut reaction consisted of 3 µl of the second step PCR product and 2 µl of either buffer (NE buffer 4 or 3) in a total volume of 20 µl. All reactions (BglII, AlwI, and uncut) were left to digest overnight (16 h) at 37°C followed by an enzyme inactivation step at 70°C for 15 min. The digested products (20 µl) were analysed by electrophoresis using a 3% gel (2% NuSieve and 1% agarose) stained with ethidium bromide. The expected RFLP patterns for the three possible genotypes are shown in Figure 21(B). The assay was verified by sequencing (ACGT

Corporation, Toronto, ON, Canada) the second PCR product using the primers 2B6*52F-BGL and 2B6*52R-ALW.

172 CYP2B6 variability and nicotine metabolism 3

drugs on CYP2B6 protein expression was not possible "e second step was also a CYP2B6 gene speci!c to assess as only one or two livers were exposed to each PCR, in which a fragment of exon 9 in CYP2B6 (206 bp) drug. However the !ve livers exposed to known CYP2B6 was ampli!ed. Two arti!cial restriction sites BglII and inducers phenobarbital and dexamethasone had signi!- AlwI were introduced by the forward and reverse prim- cantly higher CYP2B6 expression. Mean age of the organ ers 2B6*52F-BGL 5a-AGAACTTCTCCATGGCCA*GATC donors was 35 years (range = 13–64) and the gender dis- TGTGGCCCCA-3a and 2B6*52R-ALW 5a-AGGCAGGA tribution was as follows: 23 males, 13 females, and four AGTTGCGGGGGATCAGAGC*CATTG-3a, respectively. unknown. "e underlined nucleotides have been introduced to create the restriction sites and the asterisk (*) indicates the exact cutting location. Introduction of the arti!cial CYP2B6 genotyping restriction sites by the forward and reverse primers DNA from liver tissues was isolated by phenol/chlo- ensured that incomplete digestion by either enzyme is roform (Invitrogen, Canada) extraction and ethanol not a limiting factor in determining the genotype. "e precipitation. DNA samples from all the livers were geno- 25 µl reaction mixture contained: 1× PCR bu$er (10 mM typed for CYP2B6*4 (K262R), *9 (Q172H), and *6 (Q172H Tris pH 8.8, 50 mM KCl), 0.1 mM of each dNTP, 1.2 mM

and K262R) using a previously described haplotyping MgCl2, 0.25 µM of each primer, 0.3 U of Taq Polymerase assay (Lee et al. 2007a). In addition, the samples were (Fermentas, Life Sciences), and 0.8 µl of undiluted PCR genotyped for the 1459C>T SNP (R487C) in CYP2B6*5 product from the !rst reaction. "e cycling conditions (when present alone) or in CYP2B6*7 (when present in were as follows: initial denaturation at 94°C for 1 min haplotype with Q172H and K262R). A novel three-step followed by 30 cycles each consisting of denaturation at assay was developed to detect the SNP 1459C>T (R487C). 94°C for 30 s, annealing at 68°C for 45 s, and extension As an additional precaution to avoid the ampli!cation at 72°C for 2 min followed by a !nal extension at 72°C of CYP2B pseudogenes the !rst two steps of the assay for 7 min. included gene-speci!c ampli!cations. In the !rst step, In the !nal third step, restriction fragment length CYP2B6 (1063 bp) was speci!cally ampli!ed from intron polymorphism (RFLP) was used to detect the 1459C>T 8 to exon 9 utilizing the forward and reverse primers SNP. A map of the restriction sites and the predicted 2B6*51F1: 5a-GGCTAGCCTGGCCAATATGATGAT-3a fragment sizes resulting from the digestion of the second and 2B6*51R: 5a-ATTAGCCAAGCGTGGTAGTGCATG-3a, step PCR product is shown in Figure 1(A). A total of three respectively. "e 25 µl reaction mixture contained: reactions were performed for each sample: BglII, AlwI,

For personal use only. 1× PCR bu$er (10 mM Tris pH 8.8, 50 mM KCl), 0.2 mM of and uncut. Each BglII digestion reaction contained 3 µl

each dNTP, 1.1 mM MgCl2, 0.25 µM of each primer, 1.25 U of the second step PCR product, 2 µl of the 10× NE bu$er of Taq Polymerase (Fermentas, Life Sciences, Burlington, 3 and 20 U of BglII in a total volume of 20 µl. Each AlwI ON, Canada), and 50 ng of genomic DNA. "e cycling digestion reaction contained 3µl of the second step PCR conditions were as follows: initial denaturation at 94°C product, 2 µl of the 10× NE bu$er 4 and 5 U of BglII in a for 1 min followed by 30 cycles each consisting of dena- total volume of 20 µl. "e uncut reaction consisted of 3µl turation at 94°C for 20 s, annealing at 61°C for 40 s, and of the second step PCR product and 2 µl of either bu$er extension at 72°C for 2 min followed by a !nal extension (NE bu$er 4 or 3) in a total volume of 20 µl. All reactions Xenobiotica Downloaded from informahealthcare.com by University of Toronto at 72°C for 7 min. (BglII, AlwI, and uncut) were left to digest overnight

(A) (B) 188 bp 100 bp C/C C/T T/T 73 bp 115 bp ladder 18 bp UBA UBA UBA 206 206 206 BglII BglII 200 188 188 179 179 115 115 AlwI AlwI 100 93 93 86 bp 93 bp 27 bp 86 86 73 73 179 bp

Figure 1. 2 1RFLP RFLP assay assay to detect to detect the 1459C>T the 1459C>T SNP in SNP CYP2B6 in CYP2B6. (A) Map .of (A) restriction Map of sites restriction and predicted sites and fragment predicted sizes fragment following sizesdigestion following with either digestion withBglII oreither Alw I.Bgl BglIIII orwill Alw cutI. the Bgl PCRII willproduct cut atthe the PCR arti !productcial restriction at the artificialsite introduced restriction by the site forward introduced primer by and the when forward a T is primerpresent andat position when a T is 1459. AlwI will cut the PCR product at the arti!cial restriction site introduced by the reverse primer and when a C is present at position 1459. (B) present at position 1459. AlwI will cut the PCR product at the artificial restriction site introduced by the reverse primer and when a C is RFLP band patterns and sizes for the three possible genotypes (C/C, C/T, and T/T) at position 1459. U = uncut, B = BglII, and A = AlwI. present at position 1459. (B) RFLP band patterns and sizes for the three possible genotypes (C/C, C/T, and T/T) at position 1459. U= uncut, B= BglII, and A= AlwI.

173 CYP2B6 protein quantification

Liver microsomes were prepared by differential centrifugation as previously described

(Messina et al. 1997), and protein content was determined by the Bradford protein assay (Bio-

Rad Labs, ON, Canada). For Western blotting liver microsomal proteins (0.3–10 µg) were resolved on a 10% separating and 4.5% stacking sodium dodecylsulphate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were transferred overnight onto nitrocellulose membranes by wet electroblotting. Membranes were then incubated for 1h in a blocking solution of 3% w/v skim milk powder, 0.5% BSA, and 0.1% Triton X-100 in Tris buffer (TBS-

T, 50 mM Tris, 150 mM NaCl, pH 7.4). Following blocking, membranes were incubated for 2 h at room temperature with a rabbit anti-human CYP2B6 polyclonal antibody (Chemicon

International, CA, USA) diluted 1:1000 in 0.1% BSA in TBS-T. Membranes were then washed three times (5 min each) with TBS-T and reblocked with the same initial blocking solution for 1 h. The horseradish peroxidase-conjugated anti-rabbit secondary antibody (Thermo Scientific,

Rockford, IL, USA) diluted 1:3000 in 0.1% BSA in TBS-T was then added for 1 h at room temperature, followed by three washes with TBS-T (5min each). CYP2B6 was then detected using chemiluminescence (Pierce Chemical Company, Rockford, IL, USA) and exposure to auto-radiographic film (Ultident Scientific, PQ, Canada) for 30 s to 2 min. MCID Elite imaging software (Interfocus Imaging Ltd, Linton, UK) was used to quantify the densities of the visualized bands. Figure 22 is a representative western blot of all the liver samples loaded at

2µg microsomal protein. Samples with either low or high detection intensity (labelled as ‘*’ in

Figure 22) were rerun at five higher or lower microsomal concentrations, respectively, to ensure quantification in the linear detection range. CYP2B6 immunoreactivity of one liver sample

(K20) was used as a normalization factor to provide a relative scale for all samples. Each sample was analysed at least four times.

174 # * * * * * * * *

cDNA expressed Human liver microsomes CYP2B6 * # * *

Figure 22. Representative immunoblot of CYP2B6. The first sample in each row is human cDNA expressed CYP2B6 (indicated by arrows) while the remaining lanes represent different liver samples. A hash (#) represents the sample used as an internal control. An asterisk * represents samples that were subsequently loaded with different amounts of microsomal protein to insure linearity of detection as described in the materials and methods section.

Nicotine to cotinine kinetics, CYP2A6 protein quantification, and CYP2A6 genotyping

The amount of CYP2A6 protein and the metabolism of nicotine to cotinine by the human

liver microsomes has been previously assessed (Al Koudsi et al. 2009). In addition, the liver

samples were genotyped for multiple decrease or loss of activity CYP2A6 variant alleles

(CYP2A6*2, *4, *5, *6, *9, *12, *17, *20, *23, *24, *25, *26, *27, *28, and *35) (Al Koudsi et

al. 2009). Livers with one or more copy of these alleles composed the CYP2A6 variant group

(n=8), while the remaining livers formed the CYP2A6 wild-type group (n = 27) (Al Koudsi et al.

2009). Livers (n=2) genotyped as having the duplication allele (CYP2A6*1x2) were excluded in

analyses which included CYP2A6 genotype (Al Koudsi et al. 2009).

Statistical analyses

The Hardy–Weinberg equilibrium was tested using Chi-square, or Fisher’s exact test if

five or fewer individuals were in one genotype group. The Kolmogorov–Smirnov and Shapiro–

Wilk tests indicated that the phenotypic data (that is, CYP2B6 protein levels, CYP2A6 protein

levels, and Vmax) were not normally distributed. Therefore, non-parametric tests were used in

all statistical comparisons unless otherwise stated. Comparisons between two independent

groups were carried out using the Mann–Whitney U-test. A linear regression model was 175 formulated to evaluate which factors were associated with CYP2B6 protein levels. In this case the dependent variable (CYP2B6 protein level) was logged in order to normalize its distribution for this parametric test. By use of backward selection, the predictors (CYP2B6 genotype, gender, age, and exposure to inducers) were investigated for model inclusion. The reference group for the CYP2B6 genotype was the wild-type genotype group (CYP2B6*1/*1). A predictor was considered to have a significant influence if p≤0.05. The adjusted R2 value was used to assess the percentage of variation in the dependent variable that is accounted for by the overall modela.

The computed standardized Beta coefficients (Beta) for each predictor was used to measure the contribution of each predictor to the overall modela. To calculate the percentage of variation in the dependent variable that is accounted for by each predictor, the following formula was used:

(Beta value of a single predictor / sum of absolute Beta values of all predictors) x Adjusted R2 value of the total modela. The advantage of a linear regression model is that the effect of each predictor on the outcome is tested while controlling for other variables that might also be affecting the outcome.

Correlation analyses were performed using the Spearman’s rank correlation test (rS, rank correlation coefficient). The correlation between CYP2B6 and Vmax was further tested using partial regression analysis to determine the role of collinearity of CYP2A6. This allows us to test the correlation between CYP2B6 and Vmax while statistically controlling for variability in

CYP2A6 protein. The variables (CYP2B6 protein levels, CYP2A6 protein levels, and Vmax) were log normalized for the partial regression analyses for this parametric test. All statistical analyses were performed using SPSS (SPSS, Inc., Chicago, IL, USA; version 15.0 for windows). Graphs were generated using GraphPad Prism (GraphPad Software, Inc., La Jolla,

CA, USA, version 5.0 for Windows).

a Sentences added following publication during thesis correction. 176 Results

CYP2B6 protein expression

Extensive interindividual variation in CYP2B6 protein expression was observed (Figure

22). The amount of CYP2B6 protein varied greater than 200-fold (range = 0.07–20.8 relative units).

The impact of CYP2B6 genetic variation, gender, age, and exposure to inducers on

CYP2B6 protein expression was assessed in two ways. First, in order to compare to data in previous publications, the effect of each factor was tested alone. Second, we used a linear regression model in which the effect of each factor was tested while controlling for the other variables that might also be affecting the outcome. This analysis was essential as many of the factors (that is, genotype, gender, age, and exposure to inducers) were interrelated.

Impact of CYP2B6 genotype

The allele frequencies of CYP2B6 *4, *5, and *6 were 3.8% (n=80 alleles), 6.8% (n=74 alleles), and 28% (n=80 alleles), respectively. No CYP2B6*9 alleles were detected. Livers with none of the tested CYP2B6 variants (that is, CYP2B6*4, *5, and *6) were referred to as wild- type CYP2B6 livers. Genotype frequencies did not deviate from Hardy–Weinberg equilibrium

(p>0.05). Due to the conflicting results from in vitro expression studies regarding the effects of

CYP2B6*4 on protein expression (Jinno et al. 2003; Wang et al. 2006a), and the low number of samples with CYP2B6*4 (n = 3), we did not include them in the subsequent analyses. The effect of CYP2B6*5 and *6 on CYP2B6 protein expression is shown in Figure 23. Relative CYP2B6 protein expression (mean ± standard deviation (SD)) among livers with CYP2B6*1/*5 (n=5,

3.1±5.3), CYP2B6*1/*6 (n=18, 1.4±3.3), or CYP2B6*6/*6 (n=1, 0.3) was lower than livers in the wild-type CYP2B6 group (n=13, 4.0±6.5), although this did not reach statistical significance due to the large variation within each group. When combined as a variant group, livers with

177 CYP2B6*5 and *6 (n=24) had lower levels of CYP2B6 compared with the wild- type group

(n=13) (1.7±3.7 versus 4.0±6.5, p=0.28), however this did not reach statistical significance.

Livers (n=2) with the highest CYP2B6 protein expression in the variant group were both exposed to inducers (phenobarbital and/or dexamethasone). Eliminating all induced livers (n=5) from the analyses lowered the mean CYP2B6 expression in the CYP2B6 variant group; however, the difference in CYP2B6 expression between the variant (n=20) and wild-type (n=12) groups remained non-significant (0.6±0.4 versus 4.0±6.7, p=0.18).

100 p=0.28 y t i v i t

c 10 a e r o n u m m I 1 6 B 2 P Y C

e v i t 0.1 a l e R

0.01 Wildtype *1/*5 *1/*6 *6/*6 Wildtype Variant CYP2B6 Genotype

Figure 23. Impact of genetic variation in CYP2B6 on CYP2B6 protein expression. Each symbol represents a liver sample. Symbols (n=5) surrounded by a rectangle (□) are livers from people known to be exposed to established CYP2B6 inducers. One sample in the CYP2B6*1/*5 group (labelled ▼) was genotyped as heterozygous for CYP2B6*5 and CYP2B6*6 thus this liver could either be a heterozygous CYP2B6*1/*7 (that is, all SNPs on the same allele) or a compound heterozygous CYP2B6*5/*6. The horizontal solid and dashed lines denote the mean value in each group including or excluding the induced livers, respectively. The wild-type group (n = 13) represents liver samples without variant alleles. Livers in the variant group (n = 24) are a combination of livers with CYP2B6*1/*5 (n = 5), *1/*6 (n = 18), and *6/*6 (n = 1). Direct comparisons (Mann–Whitney U-test) between the different individual and combined genotype groups compared with the wild-type group revealed no significant effect of genotype on CYP2B6 expression, even when the five induced livers were excluded from the analyses. 178 Impact of gender, inducers, and age

Females (n=11) had significantly higher microsomal CYP2B6 protein levels compared with males (n=23) (Figure 24; 5.8±7.5 versus 0.8±0.9, p = 0.03). This gender effect was not significant in the absence of induced livers (females (n=8): 4.5±7.6 versus males (n=21):

0.6±0.3, p=0.1). Five liver samples exposed to dexamethasone and phenobarbital (alone or in combination) had significantly higher CYP2B6 protein expression compared with the remaining livers (n=32) (Figure 24; 6.8±6.4 versus 1.9±4.3, p=0.003). Age was not correlated with

2 CYP2B6 protein expression (Figure 25; rS = 0.01, p=0.54, n=32). Eliminating induced livers

2 did not affect the relationship between age and CYP2B6 expression (rS = 0.004, p=0.77, n=27).

100 p=0.03 p=0.003 y t i v i t

c 10 a e r o n u m m I 1 6 B 2 P Y C

e v i t 0.1 a l e R

0.01 All Male Female No inducer Inducer

Figure 24. Impact of gender and inducers on CYP2B6 protein expression. Each symbol represents a liver sample. Symbols (n=5) surrounded by a rectangle (□) are livers from people exposed to known CYP2B6 inducers. The horizontal solid and dashed lines denote the mean value in each group including or excluding the induced livers, respectively. Three samples with unknown gender were not included in the gender analysis.

179 100 2 rs =0.01, p=0.54 y t i v i t

c 10 a e r o n u m m I 1 6 B 2 P Y C

e v i

t 0.1 a l e R

0.01 0 10 20 30 40 50 60 70 Age (years)

Figure 25. Impact of age on CYP2B6 protein expression. Each symbol represents a liver sample. Symbols (n = 5) surrounded by a rectangle (□) are livers from people exposed to known CYP2B6 inducers. Five samples with unknown age were not included in the analyses.

Modeling of CYP2B6 protein level

To test the impact of a single predictor on CYP2B6 protein expression while controlling for the effects of the other predictors we developed a linear regression model which included

CYP2B6 genotype, gender, age, and exposure to inducers. All the predictors except for age

(p=0.5) had a significant effect on CYP2B6 protein expression. Livers with at least one CYP2B6 variant allele (that is, CYP2B6*5 or *6) were associated with significantly lower (p=0.04)

CYP2B6 expression, while liver samples of female origin (p=0.006) or those exposed to inducers (p=0.003) were associated with higher CYP2B6 expression. The final model accounted for 45% of the total observed variability in CYP2B6 protein expression (10% CYP2B6 genotype, 14% gender, and 21% inducer exposure).

180 Correlation of CYP2B6 with nicotine C-oxidation activity and CYP2A6

We previously reported a significant correlation between CYP2A6 protein and nicotine

2 C-oxidation activity (Vmax) (rS = 0.37, p<0.001) (Al Koudsi et al. 2009). Here, we observed a

2 significant albeit less robust correlation between CYP2B6 protein and Vmax (Figure 26(A); rS

= 0.15, p=0.02, n=37). A significant correlation between CYP2B6 and CYP2A6 protein

2 expression (Figure 26(B); rS = 0.21, p=0.004, n=37) was also observed. Since CYP2A6 correlated with both Vmax and CYP2B6 we tested whether the observed correlation between

CYP2B6 and Vmax was influenced by CYP2A6. Using partial regression analyses, in which the variability in CYP2A6 protein was statistically controlled for, the correlation between CYP2B6

2 and Vmax was abrogated (rS = 0.09, p=0.06, n=37), suggesting that much of the relationship between CYP2B6 and Vmax was due to the correlation between CYP2A6 and CYP2B6.

181 A B 100 2 100 2 rs =0.15, p=0.02 rs =0.21, p=0.004 y y t t i i v v i i t t c c 10 10 a a e e r r o o n n u u m m m m I I

1 1 6 6 B B 2 2 P P Y Y C C

e e v v i i t 0.1 t 0.1 a a l l e e R R

0.01 0.01 10 100 200 0.01 0.1 1 10 Vmax (nmol/mg of protein/hr) Relative CYP2A6 Immunoreactivity

Figure 26. Correlations between (A) CYP2B6 and Vmax and (B) CYP2B6 and CYP2A6. Each symbol represents a liver sample. Symbols (n=5) surrounded by a rectangle (□) are livers from people exposed to known CYP2B6 inducers. (A) A weak but significant correlation 2 between CYP2B6 protein expression and Vmax for nicotine C-oxidation was observed (rS = 0.15, p=0.02, n=37). Controlling for variability 2 in CYP2A6 abrogated the correlation between CYP2B6 and Vmax (rS = 0.09, p=0.06, n=37). Eliminating induced livers did not 2 significantly influence the relationship between CYP2B6 and Vmax (rS = 0.13, p=0.03, n=28). (B) CYP2B6 and CYP2A6 protein levels 2 were significantly correlated (rS = 0.21, p=0.004, n=37). Eliminating induced livers weakened the relationship between CYP2B6 and 2 CYP2A6 (rS =0.10, p=0.08, n=23).

182 Since CYP2B6 is thought to play a greater role in nicotine metabolism among those with compromised CYP2A6 activity we tested the correlation between CYP2B6 and Vmax when the livers were separated by CYP2A6 genotype. Among genetically wild-type CYP2A6 livers

2 CYP2B6 and Vmax were not significantly correlated (rS = 0.08, p=0.16, n=27). Among livers with decreased activity CYP2A6 genotypes there was a correlation between CYP2B6 and Vmax

2 (rS = 0.70, p=0.004, n=8), although the correlation was principally due to two samples with relatively high CYP2B6 expression and Vmax values. Among these livers with variant CYP2A6

2 genotypes, the correlation between CYP2A6 and Vmax was also significant (rS = 0.78, p=0.004, n=8) suggesting an important role for CYP2A6 and perhaps a derivative role for

2 CYP2B6 (as seen among all livers) due to its correlation with CYP2A6 (rS = 0.39, p=0.05, n=8), although this could not be assessed reliably due to the small numbers (n=8). Collectively, our data suggest a minor role for CYP2B6 in nicotine C-oxidation activity in most livers while a greater role might be apparent among livers with compromised CYP2A6 activity; though the latter observation should be interpreted with caution due to the small number of livers (n=8).

Impact of CYP2B6*6 on nicotine C-oxidation activity

There was no statistically significant association between CYP2B6*6 and nicotine C- oxidation activity (Vmax, p=0.9) even when the livers were separated by CYP2A6 genotype groups (Table 17). Among CYP2A6 wild-type livers, the CYP2B6*6 group trended towards having higher Vmax (p=0.09) and CYP2A6 protein levels (p=0.05) compared with the wild- type CYP2B6 group, while the number of livers with CYP2A6 variant alleles was too small for

CYP2B6 genotype comparisons. When the variability in CYP2A6 protein was controlled for, using a linear regression model, this trend of a higher Vmax in the CYP2B6*6 group among

CYP2A6 wild-type livers was abrogated (p-value changed from 0.09 to 0.5). Thus, our data suggest that once one controls for CYP2A6, there is little to no role of CYP2B6 genotype on

Vmax. 183 Table 17. Association of CYP2B6*6 with nicotine C-oxidation activity and CYP2A6.

Total livers Livers with Livers with wildtype CYP2A6 variant CYP2A6 Vmax CYP2A6 Vmax CYP2A6 Vmax CYP2A6 protein protein protein CYP2B6 38±32 0.58±0.64 26±13 0.56±0.68 65±39 0.76±0.45 Wildtype (n=17) (n=17) (n=13) (n=13) (n=3) (n=3)

CYP2B6*6 31±18 0.83±0.61 34±20 0.94±0.60 20±9.5 0.37±0.44 (n=20) (n=20) (n=14) (n=14) (n=5) (n=5)

P-value 0.90 0.10 0.09 0.05 0.30 0.30

Note: The CYP2B6*6 group includes all livers with at least one CYP2B6*6 allele. The CYP2B6 wild-type group included all livers without a CYP2B6*6 allele. Vmax units are nmol mg-1 of protein h-1. CYP2A6 represents the mean relative CYP2A6 protein levels present in each group. P-values are from direct comparisons (Mann–Whitney U-test) between CYP2B6 wild-type and CYP2B6*6 groups.

184 Discussion

When the effects of CYP2B6 genotype, gender, age, and exposure to inducers were analysed separately, only gender and exposure to inducers were significantly associated with higher CYP2B6 expression. It is important to note that in some cases direct comparisons between groups might lead to erroneous conclusions since the outcome (in this case CYP2B6 expression) is influenced by factors that are related. For example, using a direct comparison, we did not observe an effect of CYP2B6 genotype on CYP2B6 expression; however, distribution of genders and the presence of more livers in the variant group than wild-type group having been exposed to inducers contributes to this lack of genotype effect. We therefore assessed the effect of each variable using a linear regression model. One advantage of this is that it permits the identification and quantification of the relative importance of each variable while statistically controlling for the remaining variables. In addition, it provides us with the per cent variability in the dependent variable that could be accounted for by the predictors tested.

The final model accounted for approximately half (45%) of the observed variability in

CYP2B6 protein expression, with CYP2B6 genotype contributing 10%. Livers with at least one

CYP2B6 variant allele (that is, CYP2B6*5 or *6) had lower CYP2B6 protein expression, consistent with previous human liver studies assessing the effect of CYP2B6 genotype on

CYP2B6 protein expression and activity (Lang et al. 2001; Desta et al. 2007). The molecular mechanism by which CYP2B6*5 results in lower CYP2B6 protein expression is currently unknown. It is interesting to note that although CYP2B6*5 is associated with lower protein expression (Lang et al. 2001; Desta et al. 2007), bupropion hydroxylation (Desta et al. 2007), and S-mephenytoin N-demethylation (Lang et al. 2001), its efavirenz hydroxylation is similar to that of the wild-type genotype in human livers (Desta et al. 2007). In addition, this allele

(CYP2B6*5) does not affect in vivo efavirenz metabolism or clearance (Burger et al. 2006;

185 Rotger et al. 2007). The apparent discordance between CYP2B6 protein expression and efavirenz hydroxylation for CYP2B6*5 could be due to an increased specific activity towards efavirenz which would mask the effect of lower protein expression, however this remains to be tested.

The lower CYP2B6 protein expression associated with CYP2B6*6 involves a post- transcriptional mechanism whereby the variant 516G>T (Q172H) is thought to cause aberrant splicing resulting in mRNA transcripts missing exons 4 to 6 (Hofmann et al. 2008). In addition to lower protein expression, CYP2B6*6 has been associated with lower bupropion and efavirenz hydroxylation (Desta et al. 2007). Efavirenz is a commonly used antiretroviral that is prescribed as part of an initial therapy for HIV-infection (Staszewski et al. 1999). At least half of the patients that receive efavirenz experience central nervous system (CNS) side-effects, thought to reflect higher efavirenz plasma concentrations (Marzolini et al. 2001; Csajka et al. 2003). There is a strong association between high efavirenz plasma levels and CYP2B6*6 genotype in a number of studies (reviewed by (Telenti and Zanger 2008)). Indeed, Gatanaga et al. (2007) were able to successfully employ CYP2B6 genotyping to reduce the therapeutic dose of efavirenz and improve the CNS-related side-effects. CYP2B6*6 could therefore be important for future pharmacogenetic studies, especially since it occurs at relatively high (14–62%) allelic frequencies among different ethnic populations resulting in approximately 26–85% of individuals having at least one CYP2B6*6 allele (Zanger et al. 2007).

In the model, gender accounted for approximately 14% of the variability observed in

CYP2B6 protein levels. Livers from female donors had significantly higher CYP2B6 expression compared with male donors. This is thought to be mediated by a transcriptional mechanism since the major CYP2B6 regulator CAR, and CYP2B6 mRNA levels, were both found to be higher in female livers (Lamba et al. 2003). In rodents, the expression of hepatic CYP2B is sexually dimorphic and is regulated by the secretion pattern of growth hormone (GH) 186 (Murayama et al. 1991). Human GH secretion pattern is also sexually dimorphic and is a regulator of CYP (CYP1A2 and CYP3A4) expression (Liddle et al. 1998; Jaffe et al. 2002; Dhir et al. 2006; Waxman and O'Connor 2006), raising the possibility that CYP2B6 might also be regulated in a sex-dependent manner by GH (Waxman and O'Connor 2006).

Consistent with previous studies we did not observe a significant effect of age (13–64 years) on CYP2B6 expression (Shimada et al. 1994; Stresser and Kupfer 1999; Lang et al. 2001;

Parkinson et al. 2004; Desta et al. 2007). Some reports suggest that the protein expression of

CYP2B6 is lower during infancy compared with post-infancy (Tateishi et al. 1997; Croom et al.

2009), this could not be tested in our liver bank. The underlying mechanism or the biological basis for higher CYP2B6 expression following infancy is currently unknown. Finally, livers from individuals exposed to the prototypical CYP2B6 inducers (dexamethasone and/or phenobarbital) were associated with significantly higher CYP2B6 expression. In our model, exposure to inducers accounted for the largest (21%) variability in CYP2B6 protein expression, highlighting inducibility as a major contributor to the levels of CYP2B6 expression (Mo et al.

2009). Consistent with previous reports (Maurice et al. 1991; Donato et al. 2000; Madan et al.

2003), the fold induction observed here for CYP2B6 (approximately four-fold) was greater than that previously observed for CYP2A6 (approximately two-fold) among the same liver samples

(Al Koudsi et al. 2009).

CYP2B6 genotype, gender, age, and exposure to inducers accounted for approximately half (45%) of the observed variability in CYP2B6 protein expression. Sources of the remaining unaccounted variation may include other CYP2B6 variants that were not investigated in this study (for example, -82T>C). In addition variability (genetic and environmental) in the expression of nuclear transcription factors which regulate CYP expression could also affect

CYP2B6 levels. For example, CYP2B6 mRNA levels in human livers are correlated with mRNA levels of the nuclear receptors CAR (Chang et al. 2003; Lamba et al. 2003; Wortham et 187 al. 2007), HNF4-α (Wortham et al. 2007) and PXR (Chang et al. 2003). Finally, CYP2B6 could also be induced by other therapeutic agents, including ritonavir, rifampicin, and cyclophosphamide, which may have not been reported in our liver bank (Gervot et al. 1999;

Faucette et al. 2004).

When the variability in CYP2A6 protein expression was controlled for, CYP2B6

2 accounted for only 9% of the variability in Vmax (rS = 0.09, p = 0.06), suggesting only a minor role for CYP2B6 in nicotine C-oxidation. This is consistent with a number of other lines of evidence:

• Antibody and chemical inhibition studies suggest that CYP2B6 is responsible for

approximately 10–20% of nicotine’s metabolism to cotinine, while the remaining 80–

90% of the reaction is mediated by CYP2A6 (Nakajima et al. 1996a; Messina et al.

1997).

• Expressed CYP2B6 has approximately a ten-fold lower affinity for nicotine compared

with CYP2A6 (Km = 105 versus 11 µM, respectively) (Yamazaki et al. 1999).

• In human livers, CYP2B6 is thought to be present at lower levels (range=1–4% of total

CYP) compared with CYP2A6 (range=1–10% of total CYP) (Hanna et al. 2000;

Pelkonen et al. 2000).

Nonetheless, it has been suggested that CYP2B6 may play a greater role in nicotine metabolism among individuals that have reduced CYP2A6 activity. We observed a stronger correlation between CYP2B6 protein levels and nicotine C-oxidation activity among livers with

CYP2A6 variant alleles compared with those with wild-type CYP2A6; however, the former correlation must be interpreted with caution due to the small number of CYP2A6 variant livers

(n = 8).

Originally, Ring et al. (2007) reported that individuals with CYP2B6*6 had a faster rate of nicotine and cotinine clearance in the total population and among individuals genotyped as 188 reduced activity CYP2A6 metabolizers (Ring et al. 2007). Since the majority of cotinine is metabolized to trans-3′-hydroxycotinine (Hukkanen et al. 2005) in a reaction that is primarily

(approximately 100%) metabolized by CYP2A6 (Nakajima et al. 1996), this raised the question of whether CYP2B6*6 was in genetic linkage disequilibrium (LD) with CYP2A6 or associated with greater CYP2A6 activity. In a subsequent study, Lee et al. (2007) reported no evidence for genetic LD between CYP2A6 and CYP2B6 and also demonstrated no effect of CYP2B6*6 on nicotine plasma levels obtained from nicotine patch, even when the population was stratified by

CYP2A6 genotype. Consistent with this we did not observe any genetic LD between CYP2A6 and CYP2B6 and did not observe an effect of CYP2B6*6 on nicotine C-oxidation activity

(Vmax). When the data were separated by CYP2A6 genotype it was interesting to observe trends of opposing associations between CYP2B6*6, Vmax, and CYP2A6 protein levels in the different CYP2A6 genotype groups. When the variability in CYP2A6 protein was controlled for, the observed trend of higher Vmax associated with CYP2B6*6 among the CYP2A6 wild-type livers was abrogated. This indicates that variation in levels of CYP2A6 protein is the main source of variation in nicotine C-oxidation activity and that this can confound the interpretation of CYP2B6 genotype association studies with nicotine metabolism, even when the population is stratified by CYP2A6 genotype.

A correlation between CYP2B6 and CYP2A6 protein expression was observed. Of note, this correlation was not due to a generalized CYP characteristic of the liver bank as the expression of another CYP (CYP2D6) did not correlate with either CYP2B6 or CYP2A6 (data not shown). A correlation between CYP2A6 and CYP2B6 protein levels has been previously reported (Forrester et al. 1992) and is thought to be transcriptionally mediated since their mRNA levels were also correlated (Miles et al. 1989). This suggests the presence of several factors that might influence CYP2B6 and CYP2A6 expression similarly. For example, both dexamethasone and phenobarbital can induce the expression of CYP2B6 and CYP2A6 (Maurice et al. 1991; 189 Strom et al. 1996; Sueyoshi et al. 1999). In addition, similar nuclear factors (for example, CAR and HNF4-α) are involved in the transcriptional regulation of both CYP2B6 and CYP2A6

(Wortham et al. 2007). Finally, CYP2B6 and CYP2A6 are located within a 350 kb CYP2 gene cluster on chromosome 19 (Hoffman et al. 2001). The opposing transcriptional start sites and the close proximity (approximately 141kb apart) of these two genes indicates a possible shared 5′ regulatory region. While we did not observe any genetic LD between the CYP2B6 and CYP2A6

SNPs investigated herein, a large sequencing study by Haberl et al. (2005) reported genetic LD between several SNPs within CYP2B6 (-1848C>A, -801G>T, and -82T>C) and the CYP2A6*12 allele (Haberl et al. 2005). It is interesting to note that while -82T>C in CYP2B6 is associated with increased transcription in a reporter gene assay (Zukunft et al. 2005), CYP2A6*12 is associated with lower CYP2A6 mRNA and protein expression (Haberl et al. 2005; Al Koudsi et al. 2009). Thus, this is an example in which this haplotype might result in higher and lower

CYP2B6 and CYP2A6 expression, respectively. Other examples of genetic LD between

CYP2B6 and CYP2A6 which result in lower or higher expression for both CYPs have yet to be reported.

190 Conclusions

In conclusion, we observed a correlation between CYP2B6 and CYP2A6 protein expression, and once the variability in CYP2A6 protein was controlled for only a very minor role for CYP2B6 in nicotine to cotinine metabolism was observed. In addition, CYP2B6*6 did not alter cotinine formation even when the livers were stratified by CYP2A6 genotype. CYP2B6 variant alleles (*5 and *6) were associated with lower CYP2B6 expression, while being female and exposed to inducers (phenobarbital and/or dexamethasone) was associated with higher

CYP2B6 expression. We could account for approximately 45% of the variation in CYP2B6 protein expression, indicative of additional sources of variation. Identifying sources of interindividual variation will help refine the relationship between CYP2B6 and phenotypic measures and improve our understanding of the influence of CYP2B6 on clinical outcomes.

191 Acknowledgments

The authors thank Ewa B. Hoffmann and William Kim for the original development of the CYP2B6*5 assay. The authors also thank Fariba Baghai Wadji, Sharon Miksys, Andy Z. X.

Zhu, and Amandeep Mann for their technical assistance.

192 Significance of Chapter

This chapter contributes to the literature through its identification of genetic, physiologic and environmental factors that alter hepatic CYP2B6 protein expression. CYP2B6 plays a major role in the metabolism of important therapeutics such the anti-retroviral drug efavirenz and the smoking cessation drug bupropion. Because variation in CYP2B6 expression/activity can result in altered therapeutic or toxic responses to its substrates, it is important to identify the sources of variable CYP2B6 expression/activity.

With respect to the thesis, the significance of this chapter is as follows. We observed a correlation between hepatic CYP2B6 and CYP2A6 protein levels, which in part mediated the correlation between CYP2B6 and nicotine’s C-oxidation to cotinine. Controlling for variability in CYP2A6 abrogated the correlation between CYP2B6 and cotinine formation, suggesting a minimal role for CYP2B6 in nicotine C-oxidation. Consistent with this we observed no influence of the genetic variant CYP2B6*6 on the rate of cotinine formation. Previous studies assessing the influence of genetic variants in CYP2B6 on nicotine metabolism in vivo might have been confounded by the variability in the main nicotine-metabolizing enzyme CYP2A6.

193 Section 3 General Discussion

3.1 CYP2A6

3.1.1 Missing genetic variability in CYP2A6

Twin studies provide important information for understanding the role of genetic and environmental factors in determining phenotype (Rahmioglu and Ahmadi 2010). Using this approach, multiple nicotine kinetic parameters including, total clearance of nicotine, clearance of nicotine via cotinine, and the metabolite ratio of 3HC/COT have all been demonstrated to exhibit an important (50-68%) genetic component (Swan et al. 2005; Swan et al. 2009). These pathways are largely mediated by CYP2A6 (Benowitz and Jacob 1994; Dempsey et al. 2004), however adjusting for known CYP2A6 genetic variants only modestly (~10%) accounted for the genetic heritability in nicotine metabolism. This suggested that there still remained uncharacterized and unidentified genetic variability in CYP2A6. One of the main objectives of this study was to characterize known, and identify new, genetic variability in CYP2A6 in order to improve future genotype-phenotype correlations.

3.1.2 CYP2A6 sequencing and identification of novel genetic variants

In 2003-5 several sequencing studies identified a number of SNPs that were detected primarily among individuals of black African descent (Saito et al. 2003; Solus et al. 2004;

Haberl et al. 2005). However, their haplotype structure and effect on CYP2A6 activity was yet to be determined. In chapter 1 we characterized some of these SNPs resulting in the identification of several novel alleles (CYP2A6*24A&B, *25, *26, *27, and *28A&B). Although these and several other CYP2A6 alleles accounted for a considerable portion of the observed variability in CYP2A6 activity, there clearly remained much variability that was unaccounted

194 for. This was evident by the large (~30-fold) range of 3HC/COT values observed among individuals with undetected CYP2A6 variants (i.e. wild-type, CYP2A6*1/*1) (Mwenifumbo et al. 2008). This suggested that there are likely more CYP2A6 variants that were either undiscovered or perhaps undetected by the sequencing studies. Indeed, our sequencing of

CYP2A6 from individuals with the 594G>C SNP led to the identification of another nonsynonymous SNP 6458A>T (N438Y). The 6458A>T SNP occurred in haplotype with

594G>C to form the allele CYP2A6*24. At the time, 6458A>T (N438Y) was not described in the literature raising the question as to why the original sequencing paper was able to detect

594G>C (exon 2) but not 6458A>T (exon 9) (Haberl et al. 2005)? Two plausible explanations are likely. First, it is possible that the 6458A>T SNP is in haplotype with other SNPs that could have prevented the binding of the primers used to amplify and sequence that specific CYP2A6 region (exon 9) by Haberl et al. (2005). We found that 6458A>T is in haplotype with a SNP

(6782C>G) located at the 3’-end of the reverse primer (5’-AGGACGGGGGTCAGAATCGAC-

3’) used by Haberl et al. (2005) to amplify and sequence exon 9. Therefore, the sequencing study by Haberl et al. (2005) is unlikely to have detected any SNPs that are in haplotype with

6782C>G. This provides an illustration of one of the challenges encountered by genotyping and sequencing studies and suggests that there are likely still more interesting SNPs in CYP2A6 to be discovered. The second possibility is that the 594G>C detected by Haberl et al. (2005) represents a unique allele in which 594G>C is not in haplotype with 6458A>T. Although plausible, this is unlikely as we have genotyped two large populations of black African descent

(n>800 total individuals) and found that 594G>C most likely occurs in obligate haplotype with

6458A>T forming CYP2A6*24 (Al Koudsi et al. 2009a). On the other hand, 6458A>T was detected by itself (i.e. not in haplotype with 594G>C) forming the novel allele CYP2A6*35. The

6458A>T SNP was also detected among Taiwanese individuals forming the rare (0.3%) alleles

CYP2A6*36 and *37. The higher (2.9 vs. 1.3%) allele frequency of CYP2A6*35 (i.e. 6458A>T) 195 compared to CYP2A6*24 (i.e. 594G>C+6458A>T) and the presence of 6458A>T in both Asian and African populations suggests that CYP2A6*24 possibly originated from CYP2A6*35. It is likely that following the introduction of the 594G>C SNP into CYP2A6*35, possibly by DNA mutation, the two SNPs (6458A>T+594G>C; CYP2A6*24) are inherited together in haplotype due to their close proximity to each other. In general, the likelihood of separating two or more

SNPs by homologous recombination decreases with smaller distances between them (Reich et al. 2001).

Another limitation of previous sequencing studies is their emphasis on exonic regions. It is likely that important functional noncoding SNPs remain to be discovered and characterized.

In fact, CYP2A6*1B and CYP2A6*9, two of the more common CYP2A6 alleles contain noncoding variations that alter enzymatic activity. CYP2A6*1B occurs at an allelic frequency ranging from 13 to 57% in different world populations and has been associated with higher in vivo rates of nicotine clearance among Caucasians (Mwenifumbo et al. 2008a). CYP2A6*9 occurs at an allelic frequency ranging from 5 to 20% in different world populations and has been associated with lower in vivo nicotine C-oxidation activity and cigarette consumption (Yoshida et al. 2003; Minematsu et al. 2006). Our walk-on strategy of sequencing the coding and noncoding regions of CYP2A6 enabled us to identify a number of interesting coding SNPs, noncoding SNPs, and nonSNP variations (e.g. indels). For example, an insertion of a short (316 bp) interspersed nuclear element (AluYa5) in the 5’-flanking (-1199/-1198) region of CYP2A6

(CYP2A6*1K) was identified in an individual that had higher than the average predicted

CYP2A6 activity. Alu elements are one of the most abundant families of repetitive elements representing approximately 10% of the human genome (Rowold and Herrera 2000; Hasler and

Strub 2006). Originally, Alu and other repetitive elements were considered to have no major function and were referred to as “junk” DNA (Hasler et al. 2007). Nowadays, several lines of evidence suggest an important role for Alu elements in regulating gene expression at both the 196 transcriptional and post-transcriptional (e.g. altering mRNA splicing and translational efficiency) levels (Batzer and Deininger 2002; Hasler and Strub 2006). The insertion of the Alu element in the 5’-flanking region of CYP2A6 (e.g. CYP2A6*1K) could influence gene transcription in several ways. It could: 1) alter the methylation status of the promoter, and/or 2) introduce additional regulatory sequences such as binding sites of transcription factors (e.g. estrogen receptor response elements) (Norris et al. 1995; Batzer and Deininger 2002). Future experiments could identify the possible transcription factor binding sites within this Alu element, establish a genotyping assay to determine the allele frequency of CYP2A6*1K in different world populations, and investigate its effect on CYP2A6 expression and activity.

3.1.3 CYP2A6 haplotype structures

To date, sequencing studies have mostly sequenced small fragments (~1 kb) of the

CYP2A6 gene and utilized statistical approaches to assign haplotypes (Haberl et al. 2005).

Although this approach could be useful in determining haplotypes within genes, its main limitation is the reliance on population data and statistical probabilities (McDonald and Evans

2005). Our sequencing approach is the first to amplify the entire CYP2A6 gene (~9.2 kb) and clone the PCR product in order to separate the parental alleles. This allows for the direct and precise determination of haplotypes without the reliance on statistical probability. In theory, the determination of haplotypes should provide a greater predictive power of the phenotype in comparison to the identification of a single genotype. This is expected because genetic variants

(e.g. SNPs) on the same allele (i.e. haplotype) can interact together to influence the overall phenotype by potentially altering multiple processes including transcription, mRNA processing/stability, translation, and protein stability/function (Figure 27). Therefore, the overall phenotype associated with a single variant (e.g. SNP1, Figure 27) will not only be influenced by the absence or presence of that specific variant, but it will also be dependent on

197 the haplotype structure of the allele (i.e. absence or presence of SNP2, 3, 4, or 5, Figure 27). It is therefore not surprising to observe variability in CYP2A6 phenotype among individuals within a specific genotype group, especially if studies do not discern haplotypes. For example, the SNP 6600G>T (R485L) in CYP2A6*8 does not seem to influence CYP2A6 activity

(Yoshida et al. 2002), however when it is in haplotype with 6558T>C (I471T) forming

CYP2A6*10, this allele is inactive towards nicotine and coumarin (Yoshida et al. 2002;

Peamkrasatam et al. 2006). Thus, studies genotyping for only 6600G>T (R485L) will likely observe variability in the phenotype associated with this genotype since individuals could either have the active CYP2A6*8 or the inactive CYP2A6*10.

Figure 27. An illustration of a haplotype structure. A number of SNPs can have different independent effects on gene transcription, mRNA stability/processing, translation, or protein stability/function resulting in a cumulative overall phenotype. Therefore, determining haplotypes could predict the overall phenotype with a greater accuracy than just analyzing a single SNP. Figure adapted and modified from McDonald and Evans (2005).

198 In addition to gaining considerable power to predicting phenotype, determination of haplotypes is of practical value and constitutes one of the major aims of the International

HapMap Project (HapMap 2003). Technically, this could allow for genotyping few carefully chosen SNPs (‘tag’ SNPs) that are able to predict or identify the common haplotypes present in a given region. This would reduce the amount of genotyping required with little loss of information. A possible example from our study includes SNPs in CYP2A6*24 and

CYP2A6*35. Since both CYP2A6*24 (i.e. 594G>C+6458A>T) and CYP2A6*35 (i.e. 6458A>T) are associated with lower CYP2A6 activity in vivo, it is possible to use 6458A>T as a ‘tag’ SNP in the future and genotype the population for only 6458A>T as it would capture both

CYP2A6*24 and *35. Recently a large (n=31,266 individuals) genome wide association meta- analyses suggested the possible utility of -2294C>T as a ‘tag’ SNP in studies investigating the association between CYP2A6 genotype and cigarette consumption (Thorgeirsson et al. 2010).

The functional consequence of -2294C>T on CYP2A6 enzymatic expression/activity is currently not known, however it is found in linkage disequilibrium with 1799T>A (L160H) present in the loss-of-function allele CYP2A6*2 (Thorgeirsson et al. 2010). Interestingly, the association between the SNPs and reduced cigarette consumption was more significant for

-2294C>T compared to 1799T>A, suggesting that -2294C>T might be altering CYP2A6 activity independently or that it might be tagging other reduced- or loss-of-function CYP2A6 variants.

3.1.4 Unique CYP2A6 haplotypes, CYP2A6*1B and CYP2A6*4 as examples

There are several examples in which the effect of certain CYP2A6 genotypes on

CYP2A6 activity has been inconsistent among different world populations. These observations are important and highlight the need to exercise caution when generalizing the functional consequence of certain CYP2A6 alleles across different world populations. One possible

199 mechanism mediating differences in predicted allelic effect on CYP2A6 activity among different world populations is the presence of diverse, yet unique, haplotypes that occur at varying frequencies across different populations. This has been observed for several CYPs including CYP1A1, CYP1A2, CYP2C18, and CYP2C19 (Speed et al. 2009; Jorge-Nebert et al.

2010). CYP2A6*1B and CYP2A6*4 will be discussed as examples since they did not exhibit the expected phenotype in this study (Chapter 1).

CYP2A6*1B contains a 58 bp CYP2A7 gene conversion in the 3’-UTR of CYP2A6 and occurs at a relatively high allelic frequency (13-57%) in different world populations (Yamano et al. 1990; Mwenifumbo and Tyndale 2007). Some studies have found an association between

CYP2A6*1B and 1) greater mRNA stability in vitro (Wang et al. 2006), 2) higher CYP2A6 mRNA, protein, and activity levels in human livers (Wang et al. 2006), and 3) higher rates of nicotine clearance in vivo (Mwenifumbo et al. 2008a). However, this has not been widely replicated (Yoshida et al. 2003; Nakajima et al. 2006; Mwenifumbo et al. 2008; Al Koudsi et al.

2009). One possibility for the lack of replication in these studies is the ethnic makeup of the participants. Some world populations might have higher or lower frequencies of functional coding and noncoding SNPs in haplotype with the 58 bp gene conversion, thereby modifying the association between CYP2A6*1B and CYP2A6 activity. For example, in Chapters 1 and 2 we characterized several decrease- and loss-of-function CYP2A6 alleles that contain the 58 bp gene conversion (e.g. CYP2A6*24, *28, and *35). If these alleles were previously considered as

CYP2A6*1B then the association between CYP2A6*1B and greater CYP2A6 activity would likely have been diminished, especially among individuals of black African descent. In a recent study, after eliminating all currently described loss-of-function alleles that contain the 58 bp gene conversion, CYP2A6*1B was associated with higher CYP2A6 activity among individuals of black African descent (Ho et al. 2009). This is consistent with results from Caucasian populations (Mwenifumbo et al. 2008a), which have a very low (<1%) frequency of loss-of- 200 function alleles (e.g. CYP2A6*5) that contain the 58 bp gene conversion (Oscarson et al. 1999).

Due to the high frequency of CYP2A6*1B, identifying and characterizing potential loss-of- function alleles that contain the 58 bp gene conversion is critical in refining the correlation between CYP2A6 genotype and phenotype.

CYP2A6*4 is a gene deletion allele which should result in a lack of enzyme function

(Kitagawa et al. 1999). Nonetheless, some CYP2A6*4 heterozygous individuals had unexpectedly similar CYP2A6 activity levels compared to the wildtype group (i.e.

CYP2A6*1/*1). This could either be due to: 1) a genotyping error (i.e. detection of false positives), 2) the presence of SNPs that could transform the normally inactive CYP2A7 protein to an active form, or 3) the presence of an active CYP2A6 allele in haplotype with the deletion allele. In total four deletion alleles (i.e. CYP2A6*4) were sequenced. These alleles are characterized by slightly different crossover junctions resulting in unique CYP2A6/CYP2A7 hybrid alleles. This suggests that the CYP2A6*4 deletion alleles detected by our current genotyping methods are not a homogenous group, but instead consist of a heterogeneous class of deletion alleles. This could introduce variability in CYP2A6 genotype-phenotype correlation studies especially if these unique deletion alleles do not share a similar effect on CYP2A6 activity.

CYP2A6 and CYP2A7 are approximately 96% identical at the amino acid level, differing in only 21 amino acids. CYP2A7 is highly polymorphic and it is possible that genetic variation in CYP2A7 could result in a CYP2A7 allele that encodes a protein that more closely resembles the active CYP2A6. For example, eight of the variant amino acid substitutions detected in the CYP2A7 portion of CYP2A6*4E&F were the wildtype amino acids found in

CYP2A6. Whether this is sufficient to result in the expression of an active protein from the hybrid CYP2A6/CYP2A7 allele is currently not known, but might explain some of the unpredicted genotype-phenotype correlations observed in the CYP2A6 gene deletion group (i.e. 201 CYP2A6*4). In addition to determining the presence of novel CYP2A6*4 alleles we identified an individual in the CYP2A6*4 group that had at least three CYP2A6 alleles. The third CYP2A6 allele was analogous to CYP2A6*12, consisting of a hybrid in which the 5’-region until intron 2 was of CYP2A6 sequence while the remaining sequence was of CYP2A7 sequence. If this new hybrid allele is functional this could be an additional possible reason for the unexpected higher

CYP2A6 activity in this individual.

Sequencing the CYP2A6*4 alleles suggests the presence of a heterogeneous group of deletion alleles that contain different crossover junctions and CYP2A6/CYP2A7 hybrid alleles.

The multiple distinct crossover junctions reported thus far (CYP2A6*4A, *B, *D, *E, *F, *G,

*H) suggest the presence of a recombination hotspot within this area. The multiple SNPs in

CYP2A7 that are within the deletion allele might encode an active protein and contribute to the variability in CYP2A6 activity, however this remains to be tested.

3.1.5 CYP2A6 genetic variation major conclusions

In Chapters 1 and 2 we identified multiple genetic variations in CYP2A6 that included

SNPs (noncoding and coding), indels, gene conversions, gene deletions, and gene duplications.

The identification of these variants contributes to the CYP2A6 literature in several important ways. A number of the characterized CYP2A6 alleles, predominantly found in populations of black African descent, were associated with a lower CYP2A6 activity in vivo. The combined decrease- and loss-of-function alleles were present in at least 10% of the study population, thereby accounting for some of the observed variability in CYP2A6 activity and nicotine metabolism. This in turn will improve CYP2A6 genotype-phenotype correlations in future studies, especially among individuals of black African descent. In addition, these and other alleles (e.g. CYP2A6*17 and *20) (Fukami et al. 2004; Fukami et al. 2005a) that have been found predominantly in populations of black African descent may, at least in part, form the

202 biological basis for this group’s lower rates of nicotine metabolism and cigarette consumption compared to Caucasians (Benowitz et al. 1999; Office of Applied Studies 2006). Our studies also provide some insight into the possible reasons for the association of a few CYP2A6 alleles with unpredicted CYP2A6 phenotypes in different populations. For example, in Chapter 1,

CYP2A6*4 was not associated with the expected lower CYP2A6 activity. Sequencing individuals with CYP2A6*4 demonstrated the presence of a heterogeneous group of alleles that might not necessarily be predicted to associate with lower CYP2A6 activity. Although significant advances in our understanding of CYP2A6 genetic variability were made, much work still remains to be done. For example, the functional consequence of many of the noncoding

SNPs identified in the 5’-flanking region, introns, and 3’-UTR is currently unknown. These can potentially contribute to variability in CYP2A6 activity by altering transcription, mRNA splicing, mRNA stability, and translational efficiency. In addition, the frequency and functional consequence of the indels discovered is currently unknown. Finally, copy number variants have recently received considerable attention as an important class of variation in human genomes

(Iafrate et al. 2004; Sebat et al. 2004; Wain et al. 2009). With our limited sequencing we identified one individual with at least three CYP2A6 alleles. This suggests that novel CYP2A6 duplication and/or deletion alleles are present that warrant the development of a systematic approach to detect copy number variants in CYP2A6. Extending our knowledge about CYP2A6 genetic variability and its effect on CYP2A6 activity will increase the utility of genotyping in epidemiological and clinical treatment studies (Fujieda et al. 2004; Malaiyandi et al. 2006).

3.1.6 CYP2A6 genetic variation and smoking behavior

Nicotine is the main psychoactive compound in tobacco and dependent smokers modulate their cigarette consumption to maintain desired nicotine plasma levels (Benowitz

2008). Due to nicotine’s extensive metabolism, enzymes involved in the metabolism and

203 elimination of nicotine are logical candidates to alter smoking behavior. Our group and others have reported an association between genetically determined slow or absent CYP2A6 activity and an altered risk of becoming nicotine dependent (O'Loughlin et al. 2004; Audrain-McGovern et al. 2007), reduced risk of being a smoker (Iwahashi et al. 2004; Schoedel et al. 2004), greater likelihood of cessation (Gu et al. 2000), lower cigarette consumption (Rao et al. 2000; Fujieda et al. 2004; Schoedel et al. 2004; Malaiyandi et al. 2006; Minematsu et al. 2006; Mwenifumbo et al. 2007a), and reduced inhalation (Malaiyandi et al. 2006a; Strasser et al. 2007). However other, mostly older, studies did not find the same associations (London et al. 1999; Loriot et al.

2001; Carter et al. 2004). Differences in methods such as the number of CYP2A6 alleles investigated, genotype-grouping strategies, and the definition of smoking phenotypes/groupings could have contributed to the discordant findings. For example, the earlier studies only investigated a few non-prevalent CYP2A6 alleles and therefore their genotype-phenotype correlation is likely to have been obscured by the presence of uncharacterized and unidentified

CYP2A6 alleles (Tiihonen et al. 2000). In addition, the low frequency of the CYP2A6 variant alleles investigated produced small genotype groups resulting in the studies being underpowered

(Tiihonen et al. 2000). On the other hand, a study conducted in an Asian population that had a higher frequency of CYP2A6 variant alleles also reported no association between CYP2A6 and smoking status. In this study though, approximately 50% of the population were older women

(40-69 years) and 90% of the women were nonsmokers (Ando et al. 2003). Thus in this case, it is possible that the association between CYP2A6 genetic variation and smoking status would have been blunted due to environmental influences such as the social pressure in Japanese culture for women not to smoke (Hammond 2009). The definition of smokers (dependent versus nondependent) is also an important component that could limit replication. For example, we have found that CYP2A6 genetic variants are associated with lower cigarette consumption only among dependent smokers (Schoedel et al. 2004). Thus, the effect of CYP2A6 genetic variation 204 on amount smoked could be masked in studies with a greater proportion of nondependent smokers.

Considering many of the novel CYP2A6 variant alleles characterized and identified in this study occurred predominantly among individuals of black African descent, smoking behaviors and its relation to CYP2A6 are briefly discussed with specific attention to this population. Given that the proportion of CYP2A6 genetic variants among individuals of African descent is greater than Caucasians, one would expect African Americans to have a lower prevalence of smoking, smoke fewer cigarettes per day, and have higher quit rates compared to

Caucasians. Indeed, African Americans have consistently been reported to smoke fewer cigarettes per day compared to Caucasians and this difference has not been attributable to social factors (e.g. employment status, , or income) (Novotny et al. 1988; Office of Applied

Studies 2006). Therefore, the higher proportion of CYP2A6 variant alleles among African

Americans compared to Caucasians might, at least in part, be responsible for their lower cigarette consumption. On the other hand, the prevalence of smoking among African Americans

(23%) is similar to that reported among Caucasians (22%) in the United States (Centres for

Disease Control and Prevention 2010). Moreover, African Americans have lower abstinence rates compared to Caucasians (Gilpin and Pierce 2002). Thus, difference in CYP2A6 genotype is not a plausible explanation for the similar smoking prevalence rates and lower quit rates among

African Americans compared to Caucasians. Alternatively, it has been suggested that differences in social factors such as education and socioeconomic status are important determinants for the observed differences in quit rates between African Americans and

Caucasians (Kiefe et al. 2001; King et al. 2004). It is important to note that when comparing the effect of certain genetic variants within or between ethnic populations, differences in genetic, environmental, and social factors could all interact to influence the behavior.

205 3.1.7 Nicotine metabolism: in vivo-in vitro correlations

The metabolism of nicotine has been thoroughly studied in humans. In vivo studies have determined that nicotine is extensively metabolized with the major pathway being its inactivation to cotinine via CYP2A6 (Benowitz and Jacob 1994; Messina et al. 1997). In our liver bank we have measured cotinine formation using various nicotine concentrations in order to compute the kinetic parameters Vmax and Km. Vmax is the maximum rate of cotinine formation while Km is a measure of the affinity of the enzyme for the substrate and is equal to the substrate concentration at which half Vmax is attained. If the enzymatic reactions are in accordance with the Michaelis-Menten assumptions (as in our study) and the substrate concentration at the enzyme site in vivo is <10% of Km (true for nicotine) then the ratio of

Vmax/Km represents the in vitro hepatic intrinsic clearance (CLH,int) ( 1994). The intrinsic clearance represents the maximum efficiency of a metabolic process and has often been used to extrapolate in vitro data to in vivo situations (Houston 1994; Iwatsubo et al. 1997).

However, factors that might alter in vivo clearance such as blood flow and plasma protein binding are not considered in this term (CLH,int) (Venkatakrishnan et al. 2001). Hepatic blood flow is likely to significantly influence nicotine’s clearance since it is a high extraction drug

(Benowitz et al. 1982). On the other hand nicotine is estimated to be less than 5% bound to plasma proteins, therefore its distribution and clearance are unlikely to be significantly altered by different plasma protein concentrations (Benowitz et al. 1982). To convert the units of CLH,int from ml/min/mg of microsomal protein to ml/min/kg several scaling factors must be used

(Obach et al. 1997; Carlile et al. 1999; Venkatakrishnan et al. 2001). These include, 50 mg microsomal protein per gram of liver and 20 g of liver per kg of body weight in humans. The following equation represents these conversion factors using the average CLH,int (i.e. Vmax/Km) value observed in our liver bank (0.01 ml/min/mg microsomal protein).

206 In vitro hepatic intrinsic clearance (CLH,int) = Vmax/Km (0.01 ml/min/mg microsomal protein) x

52.5 (mg microsomal protein/g human liver) x 21.4 (g human liver/kg body weight)

= 11.2 ml/min/kg body weight

To estimate hepatic clearance (CLH) several mathematical models that incorporate blood flow and protein binding have been described (Gad 2008). One of the simplest and most commonly used approaches that provide robust predictive capacity is the venous equilibrium model, also known as the well-stirred model (Wilkinson and Shand 1975; Houston 1994). This model assumes the liver to be a single, homogeneous, well-stirred compartment such that the drug is rapidly distributed resulting in a uniform concentration in the whole liver (Gad 2008).

The hepatic clearance CLH deduced by the venous equilibrium model is expressed by the equation below in which QH is hepatic blood flow (1500 ml/min), fub is the fraction of drug unbound in blood (0.95 for nicotine) and CLH,int is the in vitro hepatic intrinsic clearance

(calculated above as 11.2 ml/min/kg body weight or 784 ml/min assuming a body weight of 70 kg).

CLH = (QH*fub* CLH,int) / (QH + (fub* CLH,int)) = 497.7 ml/min = 7.1 ml/min/kg

The estimated hepatic intrinsic clearance value obtained from our in vitro analysis (7.1 ml/min/kg body weight) is approximately 46% of the average in vivo metabolic clearance of nicotine reported in the literature (15.4 ml/min/kg) (Hukkanen et al. 2005). Although our estimations under-predicted the observed in vivo hepatic clearance it is within the acceptable range observed in such in vitro-in vivo extrapolations (Dr. Kenneth Thummel, personal communication). There are several factors that could potentially contribute to discrepancies observed during in vitro to in vivo extrapolations (Iwatsubo et al. 1997), these include: 1) extrahepatic metabolism, however this is considered to play a small role for nicotine (Turner et al. 1975; Kyerematen and Vesell 1991; Hukkanen et al. 2005); 2) incorrect assumption of quick

207 equilibrium of drug between blood and hepatocytes; 3) presence of active transport; and 4) nonspecific binding of drugs to membrane proteins in vitro.

In vivo, substantial interindividual variability is observed in the metabolic clearance rates of nicotine (Benowitz and Jacob 1994). Similarly, we observed a large variability in the intrinsic clearance of nicotine to cotinine (i.e. Vmax/Km). Much of this variability was due to Vmax since its values varied to a greater extent compared to Km values. This would be anticipated for the majority of CYPs since Vmax is largely dependent on enzyme levels which could be altered by multiple factors including genetic, physiological (e.g. gender), and environmental (e.g. inducers); while Km is dependent on the structure of the enzyme which is less likely to be altered by environmental or physiological factors.

A significant portion of the variability observed in Vmax was accounted for by CYP2A6

2 protein levels (rS =0.37, P<0.001). Although the correlation was slightly lower than predicted it

2 is within the range (rS =0.30-0.90) previously reported in the literature (Berkman et al. 1995;

Nakajima et al. 1996a; Messina et al. 1997; Yamazaki et al. 1999). The wide range of correlation coefficients reported between CYP2A6 and the rate of cotinine formation is likely due to the different assay conditions used in the independent studies. Important factors to consider when determining CYP include 1) compliance with Michaelis-Menten assumptions, 2) use of appropriate buffer and cofactors, and 3) substrate concentration range used (Venkatakrishnan et al. 2001). For example, at high nicotine concentrations another enzyme (CYP2B6) is thought to play a role in metabolizing nicotine to cotinine and therefore the contribution of CYP2A6 will be slightly underestimated. In accordance, we observed a better correlation between CYP2A6 and the rate of cotinine formation at lower nicotine concentrations. In our study we used a range of low (50 µM) to high (500 µM) nicotine concentrations, as one of our secondary aims was to investigate the role of CYP2B6 in nicotine metabolism (Chapter 4). 208 3.1.8 CYP2A6 genetic variation: in vivo-in vitro correlations

To date, the majority of CYP2A6 variant alleles characterized in vivo have been associated with lower nicotine C-oxidation activity (i.e. cotinine formation) (Benowitz et al.

2006; Nakajima et al. 2006), reduced clearance rates of nicotine, and increased systemic exposure to nicotine (i.e. prolonged half-life and increased AUC) (Benowitz et al. 2006).

Although many of the CYP2A6 variant alleles have been studied extensively utilizing in vitro cDNA expression systems, their effect on CYP2A6 protein expression in human livers have been assessed by only a few studies (Kiyotani et al. 2003; Haberl et al. 2005), while their effect on nicotine pharmacokinetic parameters (Vmax and Km) have not been investigated. There are several advantages of utilizing human livers compared to in vitro expression systems. Firstly, human livers represent an innate environment for CYP activity in which coenzymes such as

NADPH-cytochrome P450 oxidoreductase and cytochrome b5 are present in appropriate ratios.

Secondly, in expression systems the effect of one or a few variants can be tested at a time, while in human livers the phenotype associated with a CYP2A6 allele (i.e. multiple variants) is represented. Finally, frequently used expression systems (e.g. E. coli) lack suitable posttranslational modifications (e.g. phosphorylation) that might play a role in CYP regulation

(Gillam 1998).

Our data suggest that the lower in vivo nicotine metabolism associated with the CYP2A6 variant alleles investigated is likely due to either a reduction in CYP2A6 protein expression, an alteration in CYP2A6’s structural property, or a combination of both. Our estimated in vitro-in vivo estimations suggest a ~33% percent reduction in metabolic clearance among livers with variant CYP2A6 alleles. This is similar to the ~38% reduction in metabolic clearance observed among individuals genotyped as slow metabolizers (Benowitz et al. 2006). Understanding the effect of CYP2A6 genetic variation on the expression and function of CYP2A6 in its innate

209 environment (i.e. human livers) is of clinical use as it provides an insight to whether the effects of these genetic variants are generalizable or substrate specific. For example, if a genetic variant is associated with lower CYP2A6 protein expression then it is expected to alter the metabolism of several of its substrates similarly. Although this is generally true, there have been some examples in which a genetic variant (e.g. CYP2B6*5) is associated with lower (CYP2B6) protein expression but higher enzymatic activity towards a specific substrate (e.g. efavirenz)

(Burger et al. 2006; Rotger et al. 2007). This is likely due to an increased activity towards the substrate that masks the effect of lower protein expression. We did not observe such an effect in our liver bank since the Vmax per CYP2A6 expression level values (i.e. Vmax/CYP2A6 expression) were similar between the wildtype and variant CYP2A6 groups. This also suggests that the reduction in cotinine formation in the CYP2A6 variant livers was largely driven by the lower CYP2A6 protein expression observed in this group. The underlying mechanisms for the lower CYP2A6 protein expression associated with some of the genetic variants (e.g. CYP2A6*2,

*12, and *28) is currently not well understood, but is likely to involve posttranslational mechanisms since they had similar CYP2A6 mRNA levels compared to wildtype livers. Our data in Chapter 3 suggest the possible role for the amino acid substitutions V110L, V365M, and N438Y in increasing CYP2A6’s susceptibility to degradation, however this remains to be fully characterized. Future studies could employ the elegant approaches used by Dr.

Weinshilboum’s group with thiopurine methyltransferase to characterize the role of nonsynonymous SNPs in CYP2A6 protein degradation (Salavaggione et al. 2005; Wang et al.

2005; Feng et al. 2010).

One possibility that might contribute to the lower CYP2A6 protein expression observed in livers with variant CYP2A6 alleles is the inability of the antibody used in the western blotting assessment to effectively recognize its epitope. For example if the CYP2A6 alleles investigated contain polymorphisms within, or linked to, the epitope then the antibody might recognize the 210 epitope with a different sensitivity resulting in altered associations between protein expression levels and these genetic variations. Because the epitope of the monoclonal antibody used in our study is not provided by the manufacturer, we cannot definitively rule out this possibility.

However, this is unlikely since the livers with variant CYP2A6 alleles were also associated with lower nicotine C-oxidation activity (Vmax). Of note, the difference in the rate of cotinine formation between wildtype and variant livers (1.2-fold) was less pronounced compared to the differences in the CYP2A6 protein expression (2.3-fold). There are several possible explanations for this observation. First, quantifying CYP2A6 protein using western blotting does not distinguish between functional holoenzyme and non-functional apoenzyme. This could effect the correlation between CYP2A6 protein expression and enzyme activity if there was a disproportional presence of functional versus nonfunctional enzymes in the liver samples.

Secondly, and perhaps more importantly, CYP2B6 plays a minor role in the production of cotinine from nicotine (Yamazaki et al. 1999; Al Koudsi and Tyndale 2010). Thus, for samples with low CYP2A6 the contribution of CYP2B6 would be expected to be larger which, at least in part, might explain why the difference in the enzymatic activity between CYP2A6 wildtype and variant livers was smaller than the difference in CYP2A6 protein expression. Finally, although the production of cotinine and the enzymatic function of CYP2A6 are largely dependent on

CYP2A6 protein levels, several coenzymes are also involved. These could include aldehyde oxidase, cytochrome b5, and NADPH-cytochrome P450 oxidoreductase. In our study excess aldehyde oxidase was added to avoid limiting the reaction, however variability in the expression and/or activity of cytochrome b5 and NADPH-cytochrome P450 oxidoreductase might influence the correlation between CYP2A6 protein expression and cotinine formation. For example, if a liver sample had high CYP2A6 protein expression but low cytochrome b5 and/or NADPH- cytochrome P450 oxidoreductase activity then this sample might have a lower than predicted rate of cotinine formation. 211 3.1.9 Physiological factors: in vivo-in vitro correlations

The rates of nicotine clearance (Benowitz et al. 2006a) and CYP2A6 activity (Iscan et al.

1994; Johnstone et al. 2006; Ho et al. 2009) have consistently been reported to be higher among women compared to men. Higher rates of nicotine clearance have also been observed among pregnant women (Dempsey et al. 2002) and women taking oral contraceptives that contain estrogen (Benowitz et al. 2006a), suggesting a possible hormonal regulation of CYP2A6. In vitro, CYP2A6 is induced by estrogen in an estrogen receptor-alpha (ER-α)-dependent manner

(Higashi et al. 2007). It is interesting to note that the induction of CYP2A6 by estrogen is likely dependent on the concentration and duration of exposure since a within-subject design study failed to observe any differences in nicotine or cotinine metabolism during different phases

(midfollicular vs. midluteal) of the menstrual cycle (Hukkanen et al. 2005a).

After eliminating liver samples with CYP2A6 variant alleles, we observed approximately

2.0-, 1.9-, 1.5-, and 1.4-fold higher levels of CYP2A6 protein, CYP2A6 mRNA, cotinine formation (Vmax), and intrinsic clearance (Vmax/Km) among livers from female donors compared to male donors. According to our in vivo-in vitro estimations women should have an expected 28% higher rate of nicotine metabolic clearance compared to men. Our estimation

(28%) is greater than the difference reported in vivo among women not taking oral contraceptives (14%) but is similar to that among women taking oral contraceptives (29%).

Unfortunately, in our liver bank the use of oral contraceptives was not recorded. Our observation of higher hepatic CYP2A6 protein expression among women as a likely mechanism mediating the higher rates of nicotine clearance rates observed in vivo suggests that women would metabolize other CYP2A6 substrates at a higher rate. Indeed, a number of studies have detected higher CYP2A6 activity in women compared to men using the phenotypic reactions of coumarin 7-hydroxylation or conversion of the caffeine metabolite 1,7-dimethylxanthine to 1,7-

212 dimethyluric acid (Ujjin et al. 2002; Satarug et al. 2004; Sinues et al. 2008). Thus, studies assessing the relationship between CYP2A6 genotype and phenotype should include gender as a covariate.

A question that is commonly asked is “if women have higher rates of nicotine metabolism, are they more likely to be smokers and if so do they smoke more than men?” The answer is actually quite the opposite. In Canada, the prevalence of smoking among Canadian adults aged 15 and over is lower among women compared to men (19.6% vs. 23.9%) and significantly fewer women are heavy smokers (>26 cigarettes consumed per day) compared to men (3.8% vs. 7.4%) (Gilmore 2002). There are likely several possible reasons for the discordant association between the rate of nicotine clearance and smoking behavior between genders. It is plausible that the higher rates of nicotine clearance observed in women (Benowitz et al. 2006a) are not large enough to alter smoking behavior. Differences in smoking status and consumption is often observed among genetically predicted slow metabolizers (Schoedel et al.

2004; Malaiyandi et al. 2006) who have approximately a 37% lower rate of total nicotine clearance compared to normal metabolizers (Benowitz et al. 2006). On the other hand women only have an 18% higher rate of total nicotine clearance compared to men, a difference that is much smaller than 37%, which might not be sufficient to alter smoking behavior. Differences in pharmacodynamic rather than pharmacokinetic factors might also be playing a role. A number of studies from Dr. Perkin’s laboratory suggest that the reinforcing and rewarding effects of nicotine are less robust in women compared to men (Perkins et al. 1999; Perkins 2009). For example in a choice procedure paradigm women self-administer less nicotine compared to men

(Perkins et al. 1997; Perkins et al. 2001). Moreover, women may be more responsive than men to non-nicotine stimuli (i.e. cues) associated with cigarettes such as the scent and touch of a cigarette (Perkins et al. 2001a; Perkins 2009). Together these studies suggest a relatively less important role for nicotine in cigarette smoking among women, which might be consistent with 213 the lower therapeutic benefit obtained from nicotine replacement therapy among women compared to men (Perkins and Scott 2008). Finally, differences in social context (e.g. perceived health hazards) and individual factors (e.g. concern about body weight gain) can also influence gender differences in smoking behaviors (Perkins 2001b).

Age (2-64 years) did not correlate with either CYP2A6 protein levels or rates of cotinine formation (Vmax). This is consistent with an in vivo study which observed similar steady-state nicotine plasma levels attained from nicotine patch among individuals aged 18-39, 40-59, and

60-69 years (Gourlay and Benowitz 1996). In contrast, an in vivo nicotine pharmacokinetic study found that elderly (>65 years) subjects had significantly lower rates of nicotine clearance compared to younger adults (22-44 years). Thus, studies assessing the relationship between

CYP2A6 genotype and nicotine phenotype do not need to include age as a covariate unless participants are 65 years of age and older.

Adolescence is often described as a unique period of biological vulnerability during which nicotine exposure might lead to a greater susceptibility to continuing smoking and greater dependence (Jamner et al. 2003). Our data suggest that the possible differences between adolescent and adult individuals in the response toward nicotine are unlikely due to differences in nicotine metabolism. Other factors that might be playing a role during adolescence include increased levels of novelty seeking and the lack of fully developed self-regulation mechanisms

(Steinberg 2004). Moreover, social influences such as the drive for independence and peer acceptance might also facilitate the adoption and continuation of cigarette use among adolescents (Jamner et al. 2003). It is important to note that much of the evidence for a higher vulnerability toward nicotine sensitivity/dependence among adolescents compared to adults stems from animal studies (Abreu-Villaca et al. 2003; O'Dell et al. 2004; Levin et al. 2007).

Significant caution must be exercised when generalizing from animal models to human

214 adolescents especially given the differences in species, methods of administration, doses used, and social context of self-administration (e.g. peer influences).

3.1.10 Remaining variability in hepatic CYP2A6 expression and nicotine metabolism

The genetic, environmental, and physiological factors we have investigated accounted for some but not all of the variability observed in hepatic CYP2A6 expression and nicotine to cotinine metabolism. The correlation between CYP2A6 mRNA and protein levels seems to suggest that much (~45%) of the variability in CYP2A6 protein expression is related to transcriptional mechanisms. Thus, future studies could test whether variability (genetic or environmental) in transcription factors implicated in regulating CYP2A6 (e.g. CAR and HNF-

4α) are associated with variable CYP2A6 expression and nicotine metabolism. In addition, we cannot rule out the contribution of other unidentified genetic variants in CYP2A6. Future studies could adopt the sequencing strategy used in Chapters 1 and 2 to sequence liver samples with either very high or low CYP2A6 expression in an effort to identify new genetic variants of potential clinical importance. We predict as more CYP2A6 genetic variants are identified we will be able to account for a greater percentage of the observed variability in CYP2A6, thereby enhancing genotype-phenotype correlations.

In general the activity of a CYP enzyme is dependent on the transfer of electrons from its coenzymes (e.g. NADPH-cytochrome P450 oxidoreductase). Many coding and noncoding SNPs have been identified in the NADPH-cytochrome P450 oxidoreductase gene (POR), some of which have been associated with altered enzymatic activities of major CYPs including

CYP2A6, CYP2D6, CYP1A2, and CYP3A4 (Agrawal et al. 2008; Hart et al. 2008a; Gomes et al. 2009a). An interesting observation from these studies is that the effect of a specific SNP in

POR cannot be generalized to similarly alter the catalytic activity of all CYPs (Agrawal et al.

215 2008; Gomes et al. 2009a). Future studies could investigate the potential association between genetic variation in POR and CYP2A6 activity; this might contribute to some of the variability observed in nicotine metabolism in vivo.

3.1.11 Utility of understanding variability in nicotine metabolism

Smoking is influenced by environmental and biological/genetic factors. In the United

States the prevalence of cigarette smoking has declined considerably from its peak at 42% in

1965 to 21% in 2008 (Dube et al. 2009). Much of this decline has been attributed to increased public awareness of the dangers of smoking in addition to intensive public health control policies that included changing social norms about cigarettes and increased governmental actions to regulate the use, sale, price, and advertising of cigarettes. In addition, the development of pharmaceuticals to aid smoking cessation such as nicotine replacement therapies (NRTs), bupropion, and the recently developed varenicline has also contributed to the decline in the prevalence of smoking. However, despite these advances, in the past six years the prevalence of smoking has plateaued at 21% (Dube et al. 2009), testifying to the fact that the development of novel interventions or the refinement of already effective methods is necessary to further reduce the prevalence of smoking. One possible approach could be through personalizing treatment.

The premise of NRTs is to substitute the nicotine delivered by cigarettes in order to reduce nicotine craving and withdrawal symptoms during abstinence (Schneider et al. 1977).

Thus, variability in CYP2A6 activity/nicotine metabolism can potentially alter nicotine levels obtained from NRT products thereby influencing efficacy. Using the 3HC/COT ratio as a measure of CYP2A6 activity/nicotine metabolism, the abstinence rates from nicotine patch were reported to be significantly higher among individuals with low CYP2A6 activity compared to individuals with high CYP2A6 activity at the end of an 8-week treatment (46 vs. 27%) and at 6-

216 month follow-up (30 vs. 11%) (Lerman et al. 2006). Although the plasma nicotine levels obtained from the patch were higher among individuals with low CYP2A6 activity, the treatment nicotine levels in that study accounted for a very small proportion of the variance in abstinence (Lerman et al. 2006). This suggested that CYP2A6 activity might be influencing abstinence independent of treatment. Indeed, in a more recent placebo-controlled bupropion study, individuals with low CYP2A6 activity had significantly higher abstinence rates on placebo (i.e. no pharmacological treatment only counseling) compared to individuals with high

CYP2A6 activity (32 vs. 10%) (Patterson et al. 2008). Interestingly, bupropion treatment significantly improved abstinence rates among individuals with high CYP2A6 activity (34% on bupropion vs. 10% on placebo) while individuals with low CYP2A6 activity had similar abstinence rates (32%) on bupropion and placebo (Patterson et al. 2008). Finally, a more recent study also suggests that extending the duration of treatment with nicotine patch from 8-weeks to

6-months can improve the abstinence rates only among individuals genotyped as slow CYP2A6 metabolizers (Lerman et al. 2010). Together, these studies provide a very good example of how understanding/determining variability in CYP2A6 could potentially be used to individualize therapy and improve cessation outcomes. Specifically, these studies suggest that individuals with slow CYP2A6 activity may benefit the most from extended nicotine patch therapy or counseling alone, while individuals with high CYP2A6 activity may benefit the most from treatment with bupropion.

217 3.2 CYP2B6

3.2.1 Genetic and nongenetic factors underlying variability in hepatic CYP2B6 protein expression

In our study, approximately 45% of the observed variability in hepatic CYP2B6 protein expression could be accounted for by genetic, physiological and environmental factors. Livers from female organ donors were associated with greater CYP2B6 expression compared to males.

This is consistent with some (Lamba et al. 2003) but not all studies (Stresser and Kupfer 1999;

Lang et al. 2001; Parkinson et al. 2004; Desta et al. 2007). Differences in sample sizes, statistical power, and investigation of confounders (e.g. genotype and/or exposure to inducers) are all possible reasons for the conflicting results. For example, in our study, the effect of genotype on CYP2B6 expression was masked by the presence of a greater portion of induced livers in the variant group. Thus, the effect of genotype was only significant after controlling for the effects of induction. It is important for future studies to consider the effects of possible confounding factors and utilize appropriate statistical methods (e.g. linear regression) to assess the effect of age, gender, and genotype on general CYP expression.

The higher CYP2B6 protein expression observed among female livers is thought to be mediated by a transcriptional mechanism since CYP2B6 mRNA levels, were also found to be higher in female livers (Lamba et al. 2003). Higher CYP2B6 expression has also been observed among ER-α positive breast tumors compared to ER-α negative breast tumors, suggesting a possible role for estrogen in regulating CYP2B6 (Bieche et al. 2004; Tozlu et al. 2006). Very recently, Dr. Matthew’s laboratory identified an estrogen response element in the upstream regulatory region of CYP2B6 (-1669/-1657) to which ER-α is able to bind and activate CYP2B6 transcription in specific human breast carcinoma cells (T-47D) (Lo et al. 2010). Limitations of the study by Lo et al. (2010) include the use of estrogen concentrations (10 nM) much higher

218 than the circulating plasma estrogen levels in premenopausal women (0.1-1.0 nM) and the absence of experiments using relevant hepatic cells such as primary human hepatocytes.

Nonetheless, if CYP2B6 is regulated by estrogen in human livers, coregulation by estrogen could potentially be one mechanism underlying the correlation we observed between hepatic

CYP2A6 and CYP2B6 protein levels.

In vivo, the metabolism of CYP2B6 substrates (e.g. bupropion) is not consistently higher among women compared to men (Hsyu et al. 1997; Stewart et al. 2001). There are likely several possible reasons for the apparent discordant findings. First, differences in CYP2B6 activity between genders might not be dramatic and could easily be overlooked by the contribution of other factors such as induction or genetic variation; many studies do not control for these factors. Secondly and perhaps more importantly, even though estrogen might have an inducing effect on CYP2B6 protein expression it has also been demonstrated to inhibit CYP2B6 activity in vitro and in vivo (Palovaara et al. 2003; Walsky et al. 2006). However, the in vivo study by

Palovaara et al. (2003) has some limitations that merit mentioning. First, estrogen was provided in a formula of either hormone replacement therapy or oral contraceptive that included estrogen and . Thus, the study is unable to discern whether the observed effects are due to the estrogen or progesterone component. Secondly, the authors failed to note the induction of a parallel non-CYP2B6 pathway, which resulted in an increase in the total clearance of bupropion and a lower bupropion AUC. It is possible that the CYP2B6 activity in women is likely to be dependent on the balance between estrogen’s inducing and inhibitory effects, which could be influenced by external factors such as oral contraceptive use (Palovaara et al. 2003).

Interestingly the study (Stewart et al. 2001) that observed gender differences in bupropion metabolism in vivo was among adolescents (13-19 years), a population whose use of oral contraceptives is generally lower than the adult population (Wilkins et al. 2000). In adults no gender differences were observed in bupropion metabolism, however the use of oral 219 contraceptives was not controlled for. It will be interesting to determine gender differences in bupropion metabolism among adults while controlling for oral contraceptive use.

3.2.2 Association between CYP2B6 genetic variation, nicotine metabolism and smoking behavior

Interest in investigating the association between CYP2B6 genetic variation and nicotine metabolism stems largely from the study by Ring et al. (2007). In that study, CYP2B6*6 was associated with faster rates of nicotine and cotinine clearance in the total population and among individuals genotyped as predicted reduced activity CYP2A6 metabolizers (Ring et al. 2007).

The observation of higher rates of cotinine clearance associated with CYP2B6*6 was a little surprising given that the majority of cotinine is metabolized to trans-3’-hydroxycotinine

(Hukkanen et al. 2005) in a reaction that is primarily (approximately 100%) catalyzed by

CYP2A6 (Nakajima et al. 1996). This suggests that CYP2B6*6 in that study was likely associated with greater CYP2A6 activity. Interestingly, we observed an association between

CYP2B6*6 and higher CYP2A6 protein expression and rate of cotinine formation (Vmax) among CYP2A6 wildtype livers. However when the higher levels of CYP2A6 associated with

CYP2B6*6 were statistically controlled for, CYP2B6*6 was no longer associated with higher rates of cotinine formation suggesting that CYP2B6*6 is unlikely to affect nicotine metabolism in vivo. Consistent with this, Lee et al. (2007) demonstrated no effect of CYP2B6*6 on nicotine plasma levels obtained from nicotine patch, even when the population was stratified by CYP2A6 genotype (Lee et al. 2007). One possible limitation of in vivo studies investigating the association between CYP2B6*6 and nicotine metabolism is that the contribution of the main nicotine metabolizing enzyme (CYP2A6) is unaccounted for. This is likely to be of concern given the significant correlation we observed between CYP2A6 and CYP2B6 protein levels. As a result, it is possible for false positive associations between CYP2B6 and nicotine metabolism

220 to be reported due to the correlation between the two CYPs. For example we reported a significant correlation between CYP2B6 and the rate of cotinine formation, however when

CYP2A6’s contribution to cotinine formation was controlled for the correlation was abrogated

2 and in fact CYP2B6 only accounted for nine percent of the variability in cotinine formation (rS

= 0.09, p = 0.06), suggesting a minor role for CYP2B6 in the metabolism of nicotine. This is supported by the fact that cDNA expressed CYP2B6 has approximately 8% of the catalytic efficiency of CYP2A6 in metabolizing nicotine to cotinine (Yamazaki et al. 1999) and in general the expression of CYP2B6 in human livers is thought to be lower (range=1–4% of total

CYP) compared to CYP2A6 (range=1–10% of total CYP) (Hanna et al. 2000; Pelkonen et al.

2000).

The possible mechanism(s) mediating the association between CYP2B6*6 and higher

CYP2A6 protein expression among the CYP2A6 wildtype livers is currently not known. One possibility is that CYP2B6*6 might be in linkage disequilibrium (LD) with genetic variants in

CYP2A6 that are associated with higher CYP2A6 expression. CYP2B6 and CYP2A6 are located within a 350 kb CYP2 gene cluster on chromosome 19 and lie approximately 141 kb apart from each other (Hoffman et al. 2001). Although regions of functional LD have been mostly reported to be within 3-20 kb pairs (Salisbury et al. 2003), large (>100 kb pair) haplotype blocks have also been described (Ardlie et al. 2002). We did not observe any genetic LD between the

CYP2B6 and CYP2A6 SNPs investigated herein, however a large sequencing study provides some evidence for genetic LD between several SNPs within CYP2B6 (-1848C>A, -801G>T, and -82T>C) and the CYP2A6*12 allele (Haberl et al. 2005). Whether CYP2B6*6 is in genetic

LD with variants in CYP2A6 that result in higher CYP2A6 expression is currently not known.

Due to the minor role of CYP2B6 in the major nicotine-metabolizing pathway, genetic variability in CYP2B6 is not predicted to significantly influence smoking behaviors via this mechanism. However, a recent large genome-wide association meta-analysis observed a 221 significant association between an intronic SNP (24428G>A) in CYP2B6 and reduced cigarette consumption (Thorgeirsson et al. 2010). The functional impact of the intronic SNP (24428G>A) on CYP2B6 activity is currently not known. Genotype data from the Hapmap project do not indicate that this SNP (24428G>A) in CYP2B6 is in LD with functional SNPs in either CYP2B6 or CYP2A6 (http://hapmap.ncbi.nlm.nih.gov/, accessed May 16, 2010). Thus, the current mechanism associating 24428G>A in CYP2B6 with cigarette consumption is currently not known and warrants further investigation. On the other hand, our group has found no association between CYP2B6*6 and a variety of smoking behaviors including risk of smoking, cigarettes per day, and duration of smoking (Lee et al. 2007a; Faheim 2009). This is in contrast to CYP2A6 in which individuals genotyped as slow CYP2A6 metabolizers are associated with a decreased risk for smoking, lower cigarette consumption per day, and decreased duration of smoking

(Schoedel et al. 2004).

Genetic variation in CYP2B6 however has the potential to alter smoking cessation rates since it is the main enzyme involved in the metabolism of bupropion to its major metabolite (Faucette et al. 2000). Bupropion was the first non-nicotine product approved to aid smoking cessation (Zyban) and is also used in the treatment of depression

(Wellbutrin) (Siu and Tyndale 2007). Bupropion’s biological is not fully understood (Siu and Tyndale 2007), but evidence suggests that in addition to its nicotinic properties (Slemmer et al. 2000) it can also inhibit the reuptake of norepinephrine and dopamine (Stahl et al. 2004). The nicotinic receptor antagonist properties of bupropion may, in part, attenuate the rewarding effects of nicotine (Slemmer et al. 2000), while the reuptake inhibition of dopamine and norepinephrine may, in part, reduce cravings and alleviate certain withdrawal symptoms experienced in abstinent smokers (Shiffman et al. 2000;

Durcan et al. 2002). The efficacy and safety of bupropion as a smoking cessation drug is related to the plasma concentrations of itself and its metabolites (Johnston et al. 2001; Johnston et al. 222 2002). For example, patients receiving 100, 150, or 300 mg of bupropion a day for seven weeks were 1.42-, 1.69-, and 2.84-times more likely to quit smoking (Johnston et al. 2001). Whereas the efficacy of bupropion (i.e. probability of quitting) was most strongly correlated with bupropion plasma levels (r=0.87, p<0.05) the side effects of insomnia and dry mouth were most correlated with bupropion metabolites erythrohydrobupropion and threohydrobupropion, respectively (Johnston et al. 2001; Johnston et al. 2002). This suggests that variables that alter bupropion plasma levels by either its intake (e.g. dosage) or perhaps by its removal (e.g. metabolism) might play an important role in altering the clinical efficacy and safety of bupropion.

Genetic variation in CYP2B6, more specifically CYP2B6*6, has been associated with reduced bupropion hydroxylation in vitro (Desta et al. 2007) and a trend for lower rates of bupropion clearance in vivo (Loboz et al. 2006). An individual homozygous for CYP2B6*6 had a bupropion clearance rate that was approximately 40% lower than the average clearance rate observed among wildtype CYP2B6 individuals (5.3 vs. 9.0 L/hr/kg of body weight) (Loboz et al.

2006). Together the in vitro and in vivo data suggest a possible role for genetic variation in

CYP2B6 in altering bupropion treatment outcomes. The possible consequence of CYP2B6*6 on bupropion smoking cessation outcomes is of interest due to the high (14-62%) allelic frequency of CYP2B6*6 among different world populations (Zanger et al. 2007). In 2007 our laboratory hypothesized that subjects with at least one copy of CYP2B6*6 will likely have higher plasma bupropion levels due to its slower metabolism, resulting in higher rates of abstinence compared to individuals that do not have CYP2B6*6 (i.e. CYP2B6*1/*1, wildtype individuals) (Lee et al.

2007a). In contrast to this hypothesis, individuals treated with bupropion that had at least one copy of CYP2B6*6 had similar abstinence rates compared to the wildtype CYP2B6 group (Lee et al. 2007a). Nonetheless, several other interesting observations were made. For example within the CYP2B6*6 group, individuals treated with bupropion had a higher abstinence rate 223 compared to individuals treated with placebo (33% vs. 15%, p=0.01) while in the CYP2B6*1

(i.e. wildtype) group the abstinence rates were similar between bupropion and placebo treated individuals (31% vs. 32%, p=0.93) (Lee et al. 2007a). In addition, individuals treated with bupropion or placebo within the CYP2B6*6 group maintained their abstinence rates at the end of treatment for a period of six months, while the abstinence rates among the bupropion- or placebo-treated CYP2B6*1 individuals declined over time (Lee et al. 2007a). These observations suggest that individuals with CYP2B6*6 should be treated with bupropion as they benefit the most from it, while treatment of the CYP2B6*1 group with bupropion confers no additional therapeutic advantage compared to placebo (Lee et al. 2007a). Before any genotype- based recommendations for treatment with bupropion can be made these observations warrant replication in different smoking (i.e. heavy and light smokers) and world populations.

The mechanism(s) underlying the low abstinence rates achieved from placebo treatment among individuals with CYP2B6*6 is currently not known. One possibility is that within this population CYP2B6*6 was associated with higher CYP2A6 activity. This has been observed in our liver bank among CYP2A6 wildtype livers as discussed earlier. Higher CYP2A6 activity measured by the 3HC/COT ratio has been associated with lower abstinence rates achieved from placebo treatment (Patterson et al. 2008). Therefore the possible association between CYP2B6*6 and higher CYP2A6 activity could be mediating the low abstinence rates observed among the

CYP2B6*6 group. Interestingly, in the study by Patterson et al. (2008) individuals with high

CYP2A6 activity benefited the most from bupropion treatment, while individuals with low

CYP2A6 activity achieved no further benefit from bupropion (Patterson et al. 2008). This is similar to the observations by Lee et al. (2007) in which the CYP2B6*6 group benefited the most from bupropion treatment compared to placebo, while the CYP2B6*1 group did not achieve any further benefit from bupropion. In the future, studies assessing the association between CYP2B6 genetic variation and smoking cessation aided with bupropion need to control 224 for variability in CYP2A6 activity, preferably by measuring the 3HC/COT ratio. The lower placebo abstinence rates among CYP2B6*6 individuals might also involve another mechanism that is unrelated to nicotine metabolism. CYP2B6 is expressed in the brain (Miksys et al. 2003), although at a lower level compared to the liver, and is involved in the minor metabolism of several endogenous substrates including serotonin (Fradette et al. 2004). The impact of

CYP2B6*6 on CYP2B6 activity is substrate specific; it has been associated with greater cyclophosphamide hydroxylation (Nakajima et al. 2007), but lower bupropion and efavirenz hydroxylation (Desta et al. 2007). The impact of CYP2B6*6 on serotonin metabolism is unknown, but if we postulate that it results in greater activity towards serotonin then individuals with CYP2B6*6 might have lower levels of serotonin resulting in lower positive mood among these individuals. Lower baseline positive mood has been identified as a risk factor for relapse during smoking cessation trials (Doran et al. 2006). Thus, individuals with CYP2B6*6 might have lower abstinence rates due to the associated higher rate of serotonin metabolism. This hypothesis merits examination in future studies. The association between genetic variation in

CYPs and personality trait has been previously reported. For example, CYP2D6 is thought to be involved in the synthesis of certain neurotransmitters (e.g. dopamine) (Hiroi et al. 1998) and genetic variation in CYP2D6 has been associated with personality traits such as harm avoidance

(Bertilsson et al. 1989; Roberts et al. 2004).

3.3 Conclusions

Our findings contribute significantly to the current knowledge of CYP2A6 genetic variation and factors influencing variability in hepatic CYP2A6 expression, CYP2B6 expression, and nicotine metabolism. Several novel decrease- and loss-of-function CYP2A6 alleles were identified, predominantly among individuals of black African descent. These alleles are likely to have an important role in refining future CYP2A6 genotype-phenotype associations

225 since approximately 10% of the black African population is predicted to have at least one copy of these alleles. In addition, the finding that these alleles are predominantly present in populations of black African descent might contribute to their slower nicotine metabolism and lower cigarette consumption compared to their Caucasian counterpart.

The mechanisms underlying the lower in vivo rates of nicotine clearance associated with multiple CYP2A6 genetic variants were not previously fully understood. Our data suggest that this association is likely due to either a reduction in CYP2A6 protein expression, an alteration in

CYP2A6’s structural properties, or a combination of both. We also identified higher CYP2A6 protein expression as one possible mechanism mediating the higher rates of nicotine metabolism observed in women compared to men. Finally we observed a correlation between CYP2A6 and

CYP2B6 and when the contribution of CYP2A6 to the metabolism of nicotine to cotinine

(Vmax) was controlled for, CYP2B6 did not significantly (~9%) account for the variability in

Vmax. This suggested that CYP2B6 plays a minor role in nicotine metabolism and suggests caution in interpreting studies assessing the potential role of CYP2B6 in smoking behaviors, as this might be confounded by CYP2B6’s correlation with CYP2A6. The identification of variables (whether genetic, physiological, or environmental) that contribute to CYP2A6 variability is critical in order to improve and refine the association between CYP2A6 genotype, nicotine metabolism, and smoking behaviors. Similarly, if nicotine is used therapeutically (e.g. nicotine replacement therapies) then a better understanding of the variability in its metabolism could provide an opportunity to personalize therapy in order to improve efficacy while reducing potential adverse effects (e.g. nausea).

226 References

Abreu-Villaca, Y., F. J. Seidler, D. Qiao, C. A. Tate, M. M. Cousins, I. Thillai, et al. (2003). Short-term adolescent nicotine exposure has immediate and persistent effects on cholinergic systems: critical periods, patterns of exposure, dose thresholds. Neuropsychopharmacology 28(11): 1935-49. Adhikiari, B., J. Kahende, A. Malarcher, T. Pechacek and V. Tong (2008). Smoking- Attributable Mortality, Years of Potential Life Lost, and Productivity Losses: United States, 2000-2004. Morbidity and Mortality Weekly Report 57(45): 1226-1228. Agrawal, V., N. Huang and W. L. Miller (2008). Pharmacogenetics of P450 oxidoreductase: effect of sequence variants on activities of CYP1A2 and CYP2C19. Pharmacogenet Genomics 18(7): 569-76. Aguiar, M., R. Masse and B. F. Gibbs (2005). Regulation of cytochrome P450 by posttranslational modification. Drug Metab Rev 37(2): 379-404. Ahluwalia, J. S., K. Okuyemi, N. Nollen, W. S. Choi, H. Kaur, K. Pulvers, et al. (2006). The effects of nicotine gum and counseling among African American light smokers: a 2 x 2 factorial design. Addiction 101(6): 883-91. Al Koudsi, N., J. S. Ahluwalia, S. K. Lin, E. M. Sellers and R. F. Tyndale (2009a). A novel CYP2A6 allele (CYP2A6*35) resulting in an amino-acid substitution (Asn438Tyr) is associated with lower CYP2A6 activity in vivo. Pharmacogenomics J 9(4): 274-82. Al Koudsi, N., E. B. Hoffmann, A. Assadzadeh and R. F. Tyndale (2009). Hepatic CYP2A6 levels and nicotine metabolism: impact of genetic, physiological, environmental, and epigenetic factors. Eur J Clin Pharmacol 66(3): 239-51. Al Koudsi, N., J. C. Mwenifumbo, E. M. Sellers, N. L. Benowitz, G. E. Swan and R. F. Tyndale (2006). Characterization of the novel CYP2A6*21 allele using in vivo nicotine kinetics. Eur J Clin Pharmacol 62(6): 481-4. Al Koudsi, N. and R. F. Tyndale (2005). Genetic influences on smoking: a brief review. Ther Drug Monit 27(6): 704-9. Al Koudsi, N. and R. F. Tyndale (2010). Hepatic CYP2B6 is altered by genetic, physiologic, and environmental factors but plays little role in nicotine metabolism. Xenobiotica 40(6): 381-92. Al-Salmy, H. S. (2001). Individual variation in hepatic aldehyde oxidase activity. IUBMB Life 51(4): 249-53. Albuquerque, E. X., E. F. Pereira, M. Alkondon and S. W. Rogers (2009). Mammalian nicotinic acetylcholine receptors: from structure to function. Physiol Rev 89(1): 73-120. Anandatheerthavarada, H. K., G. Biswas, J. Mullick, N. B. Sepuri, L. Otvos, D. Pain, et al. (1999). Dual targeting of cytochrome P4502B1 to endoplasmic reticulum and mitochondria involves a novel signal activation by cyclic AMP-dependent phosphorylation at ser128. EMBO J 18(20): 5494-504. Ando, M., N. Hamajima, N. Ariyoshi, T. Kamataki, K. Matsuo and Y. Ohno (2003). Association of CYP2A6 gene deletion with cigarette smoking status in Japanese adults. J Epidemiol 13(3): 176-81. Aoyama, T., S. Yamano, P. S. Guzelian, H. V. Gelboin and F. J. Gonzalez (1990). Five of 12 forms of vaccinia -expressed human hepatic cytochrome P450 metabolically activate aflatoxin B1. Proc Natl Acad Sci U S A 87(12): 4790-3. Arcavi, L., P. Jacob, 3rd, M. Hellerstein and N. L. Benowitz (1994). Divergent tolerance to metabolic and cardiovascular effects of nicotine in smokers with low and high levels of cigarette consumption. Clin Pharmacol Ther 56(1): 55-64.

227 Ardlie, K. G., L. Kruglyak and M. Seielstad (2002). Patterns of linkage disequilibrium in the human genome. Nat Rev Genet 3(4): 299-309. Arenaz, I., J. Vicente, A. Fanlo, P. Vasquez, J. C. Medina, B. Conde, et al. (2009). Haplotype structure and allele frequencies of CYP2B6 in Spaniards and Central Americans. Fundam Clin Pharmacol. Ariyoshi, N., Y. Sawamura and T. Kamataki (2001). A novel single nucleotide polymorphism altering stability and activity of CYP2a6. Biochem Biophys Res Commun 281(3): 810-4. Ariyoshi, N., H. Sekine, K. Nakayama, K. Saito, A. Miyamoto and T. Kamataki (2004). Identification of deletion-junction site of CYP2A6*4B allele lacking entire coding region of CYP2A6 in Japanese. Pharmacogenetics 14(10): 701-5. Ariyoshi, N., Y. Takahashi, M. Miyamoto, Y. Umetsu, S. Daigo, T. Tateishi, et al. (2000). Structural characterization of a new variant of the CYP2A6 gene (CYP2A6*1B) apparently diagnosed as heterozygotes of CYP2A6*1A and CYP2A6*4C. Pharmacogenetics 10(8): 687-93. Armitage, A., C. Dollery, T. Houseman, E. Kohner, P. J. Lewis and D. Turner (1978). Absorption of nicotine from small cigars. Clin Pharmacol Ther 23(2): 143-51. Armitage, A. K., C. T. Dollery, C. F. George, T. H. Houseman, P. J. Lewis and D. M. Turner (1975). Absorption and metabolism of nicotine from cigarettes. Br Med J 4(5992): 313- 6. Ashton, H., R. Stepney and J. W. Thompson (1979). Self-titration by cigarette smokers. Br Med J 2(6186): 357-60. Asimus, S., T. N. Hai, N. Van Huong and M. Ashton (2008). Artemisinin and CYP2A6 activity in healthy subjects. Eur J Clin Pharmacol 64(3): 283-92. Attwood, J. T., R. L. Yung and B. C. Richardson (2002). DNA methylation and the regulation of gene transcription. Cell Mol Life Sci 59(2): 241-57. Audrain-McGovern, J., N. Al Koudsi, D. Rodriguez, E. P. Wileyto, P. G. Shields and R. F. Tyndale (2007). The role of CYP2A6 in the emergence of nicotine dependence in adolescents. Pediatrics 119(1): e264-74. Ayesh, D., M. Al-Waiz, J. Crothers, S. Cholerton, S. C. Mitchell, J. R. Idle, et al. (1988). Deficient nicotine N-oxidation in two sisters with trimethylaminuria. Br J Clin Pharmacol 25: 664P-665P. Baker, J. R., S. Satarug, P. E. Reilly, R. J. Edwards, N. Ariyoshi, T. Kamataki, et al. (2001). Relationships between non-occupational cadmium exposure and expression of nine cytochrome P450 forms in human liver and kidney cortex samples. Biochem Pharmacol 62(6): 713-21. Bandiera, S., S. Weidlich, V. Harth, P. Broede, Y. Ko and T. Friedberg (2005). Proteasomal degradation of human CYP1B1: effect of the Asn453Ser polymorphism on the post- translational regulation of CYP1B1 expression. Mol Pharmacol 67(2): 435-43. Bao, Z., X. Y. He, X. Ding, S. Prabhu and J. Y. Hong (2005). Metabolism of nicotine and cotinine by human cytochrome P450 2A13. Drug Metab Dispos 33(2): 258-61. Batzer, M. A. and P. L. Deininger (2002). Alu repeats and human genomic diversity. Nat Rev Genet 3(5): 370-9. Beckett, A. H., J. W. Gorrod and P. Jenner (1972). A possible relation between pKa 1 and and the amounts excreted in urine of some tobacco alkaloids given to man. J Pharm Pharmacol 24(2): 115-20. Benowitz, N. L. (1986). Clinical pharmacology of nicotine. Annu Rev Med 37: 21-32. Benowitz, N. L. (1988). Drug therapy. Pharmacologic aspects of cigarette smoking and nicotine addition. N Engl J Med 319(20): 1318-30.

228 Benowitz, N. L. (2008). Clinical pharmacology of nicotine: implications for understanding, preventing, and treating tobacco addiction. Clin Pharmacol Ther 83(4): 531-41. Benowitz, N. L. (2009b). Pharmacology of nicotine: addiction, smoking-induced disease, and therapeutics. Annu Rev Pharmacol Toxicol 49: 57-71. Benowitz, N. L., J. T. Bernert, R. S. Caraballo, D. B. Holiday and J. Wang (2009). Optimal serum cotinine levels for distinguishing cigarette smokers and nonsmokers within different racial/ethnic groups in the United States between 1999 and 2004. Am J Epidemiol 169(2): 236-48. Benowitz, N. L., A. Hansson and P. Jacob, 3rd (2002). Cardiovascular effects of nasal and transdermal nicotine and cigarette smoking. Hypertension 39(6): 1107-12. Benowitz, N. L., B. Herrera and P. Jacob, 3rd (2004). Mentholated cigarette smoking inhibits nicotine metabolism. J Pharmacol Exp Ther 310(3): 1208-15. Benowitz, N. L. and P. Jacob, 3rd (1984). Daily intake of nicotine during cigarette smoking. Clin Pharmacol Ther 35(4): 499-504. Benowitz, N. L. and P. Jacob, 3rd (1985). Nicotine renal excretion rate influences nicotine intake during cigarette smoking. J Pharmacol Exp Ther 234(1): 153-5. Benowitz, N. L. and P. Jacob, 3rd (1993). Nicotine and cotinine elimination pharmacokinetics in smokers and nonsmokers. Clin Pharmacol Ther 53(3): 316-23. Benowitz, N. L. and P. Jacob, 3rd (1994). Metabolism of nicotine to cotinine studied by a dual stable isotope method. Clin Pharmacol Ther 56(5): 483-93. Benowitz, N. L. and P. Jacob, 3rd (2000). Effects of cigarette smoking and carbon monoxide on nicotine and cotinine metabolism. Clin Pharmacol Ther 67(6): 653-9. Benowitz, N. L. and P. Jacob, 3rd (2001). Trans-3'-hydroxycotinine: disposition kinetics, effects and plasma levels during cigarette smoking. Br J Clin Pharmacol 51(1): 53-9. Benowitz, N. L., P. Jacob, 3rd, C. Denaro and R. Jenkins (1991). Stable isotope studies of nicotine kinetics and bioavailability. Clin Pharmacol Ther 49(3): 270-7. Benowitz, N. L., P. Jacob, 3rd, I. Fong and S. Gupta (1994a). Nicotine metabolic profile in man: comparison of cigarette smoking and transdermal nicotine. J Pharmacol Exp Ther 268(1): 296-303. Benowitz, N. L., P. Jacob, 3rd, R. T. Jones and J. Rosenberg (1982). Interindividual variability in the metabolism and cardiovascular effects of nicotine in man. J Pharmacol Exp Ther 221(2): 368-72. Benowitz, N. L., F. Kuyt, P. Jacob, 3rd, R. T. Jones and A. L. Osman (1983). Cotinine disposition and effects. Clin Pharmacol Ther 34(5): 604-11. Benowitz, N. L., C. N. Lessov-Schlaggar, G. E. Swan and P. Jacob, 3rd (2006a). Female sex and oral contraceptive use accelerate nicotine metabolism. Clin Pharmacol Ther 79(5): 480- 8. Benowitz, N. L., E. J. Perez-Stable, I. Fong, G. Modin, B. Herrera and P. Jacob, 3rd (1999). Ethnic differences in N-glucuronidation of nicotine and cotinine. J Pharmacol Exp Ther 291(3): 1196-203. Benowitz, N. L., E. J. Perez-Stable, B. Herrera and P. Jacob, 3rd (2002a). Slower metabolism and reduced intake of nicotine from cigarette smoking in Chinese-Americans. J Natl Cancer Inst 94(2): 108-15. Benowitz, N. L., H. Porchet, L. Sheiner and P. Jacob, 3rd (1988a). Nicotine absorption and cardiovascular effects with use: comparison with cigarettes and nicotine gum. Clin Pharmacol Ther 44(1): 23-8. Benowitz, N. L., G. E. Swan, P. Jacob, 3rd, C. N. Lessov-Schlaggar and R. F. Tyndale (2006). CYP2A6 genotype and the metabolism and disposition kinetics of nicotine. Clin Pharmacol Ther 80(5): 457-67. 229 Benowitz, N. L., S. Zevin and P. Jacob, 3rd (1998). Suppression of nicotine intake during ad libitum cigarette smoking by high-dose transdermal nicotine. J Pharmacol Exp Ther 287(3): 958-62. Berg, J. Z., J. Mason, A. J. Boettcher, D. K. Hatsukami and S. E. Murphy (2010). Nicotine metabolism in African Americans and European Americans: variation in glucuronidation by ethnicity and UGT2B10 haplotype. J Pharmacol Exp Ther 332(1): 202-9. Berkman, C. E., S. B. Park, S. A. Wrighton and J. R. Cashman (1995). In vitro-in vivo correlations of human (S)-nicotine metabolism. Biochem Pharmacol 50(4): 565-70. Berlin, I., M. J. Gasior and E. T. Moolchan (2007). Sex-based and hormonal contraception effects on the metabolism of nicotine among adolescent tobacco-dependent smokers. Nicotine Tob Res 9(4): 493-8. Bertilsson, L., C. Alm, C. De Las Carreras, J. Widen, G. Edman and D. Schalling (1989). Debrisoquine hydroxylation polymorphism and personality. Lancet 1(8637): 555. Bieche, I., I. Girault, E. Urbain, S. Tozlu and R. Lidereau (2004). Relationship between intratumoral expression of genes coding for xenobiotic-metabolizing enzymes and benefit from adjuvant tamoxifen in estrogen receptor alpha-positive postmenopausal breast carcinoma. Res 6(3): R252-63. Bird, A. P. (1980). DNA methylation and the frequency of CpG in animal DNA. Nucleic Acids Res 8(7): 1499-504. Bock, C., S. Reither, T. Mikeska, M. Paulsen, J. Walter and T. Lengauer (2005). BiQ Analyzer: visualization and quality control for DNA methylation data from bisulfite sequencing. Bioinformatics 21(21): 4067-8. Boffetta, P., S. Hecht, N. Gray, P. Gupta and K. Straif (2008). Smokeless tobacco and cancer. Lancet Oncol 9(7): 667-75. Botto, F., E. Seree, S. el Khyari, G. de Sousa, A. Massacrier, M. Placidi, et al. (1994). Tissue- specific expression and methylation of the human CYP2E1 gene. Biochem Pharmacol 48(6): 1095-103. Brandange, S. and L. Lindblom (1979). The enzyme "aldehyde oxidase" is an iminium oxidase. Reaction with nicotine delta 1'(5') iminium ion. Biochem Biophys Res Commun 91(3): 991-6. Brunnemann, K. D. and D. Hoffmann (1974). The pH of tobacco smoke. Food Cosmet Toxicol 12(1): 115-24. Bumpus, N. N., U. M. Kent and P. F. Hollenberg (2006). Metabolism of efavirenz and 8- hydroxyefavirenz by P450 2B6 leads to inactivation by two distinct mechanisms. J Pharmacol Exp Ther 318(1): 345-51. Bumpus, N. N., C. Sridar, U. M. Kent and P. F. Hollenberg (2005). The naturally occurring cytochrome P450 (P450) 2B6 K262R mutant of P450 2B6 exhibits alterations in substrate metabolism and inactivation. Drug Metab Dispos 33(6): 795-802. Burger, D., I. van der Heiden, C. la Porte, M. van der Ende, P. Groeneveld, C. Richter, et al. (2006). Interpatient variability in the pharmacokinetics of the HIV non-nucleoside reverse transcriptase inhibitor efavirenz: the effect of gender, race, and CYP2B6 polymorphism. Br J Clin Pharmacol 61(2): 148-54. Byrd, G. D., K. M. Chang, J. M. Greene and J. D. deBethizy (1992). Evidence for urinary excretion of glucuronide conjugates of nicotine, cotinine, and trans-3'-hydroxycotinine in smokers. Drug Metab Dispos 20(2): 192-7. Campbell, M. E., D. M. Grant, T. Inaba and W. Kalow (1987). Biotransformation of caffeine, paraxanthine, theophylline, and by polycyclic aromatic hydrocarbon- inducible cytochrome(s) P-450 in human liver microsomes. Drug Metab Dispos 15(2): 237-49. 230 Carlile, D. J., N. Hakooz, M. K. Bayliss and J. B. Houston (1999). Microsomal prediction of in vivo clearance of CYP2C9 substrates in humans. Br J Clin Pharmacol 47(6): 625-35. Carter, B., T. Long and P. Cinciripini (2004). A meta-analytic review of the CYP2A6 genotype and smoking behavior. Nicotine Tob Res 6(2): 221-7. Cashman, J. R., S. B. Park, Z. C. Yang, S. A. Wrighton, P. Jacob, 3rd and N. L. Benowitz (1992). Metabolism of nicotine by human liver microsomes: stereoselective formation of trans-nicotine N'-oxide. Chem Res Toxicol 5(5): 639-46. Cashman, J. R., J. Zhang, J. Leushner and A. Braun (2001). Population distribution of human flavin-containing monooxygenase form 3: gene polymorphisms. Drug Metab Dispos 29(12): 1629-37. Cauffiez, C., J. M. Lo-Guidice, S. Quaranta, D. Allorge, D. Chevalier, S. Cenee, et al. (2004). Genetic polymorphism of the human cytochrome CYP2A13 in a French population: implication in lung cancer susceptibility. Biochem Biophys Res Commun 317(2): 662-9. Cauffiez, C., N. Pottier, G. Tournel, J. M. Lo-Guidice, D. Allorge, D. Chevalier, et al. (2005). CYP2A13 genetic polymorphism in French Caucasian, Gabonese and Tunisian populations. Xenobiotica 35(7): 661-9. CDC (1988). The Health Consequences of Smoking: Nicotine Addiction: A report of the Surgeon General, U.S. Department of Health and Human Services. Celius, T., A. Pansoy, J. Matthews, A. B. Okey, M. C. Henderson, S. K. Krueger, et al. (2010). Flavin-containing monooxygenase-3: induction by 3-methylcholanthrene and complex regulation by xenobiotic chemicals in hepatoma cells and mouse liver. Toxicol Appl Pharmacol In press. Celius, T., S. Roblin, P. A. Harper, J. Matthews, P. C. Boutros, R. Pohjanvirta, et al. (2008). Aryl hydrocarbon receptor-dependent induction of flavin-containing monooxygenase mRNAs in mouse liver. Drug Metab Dispos 36(12): 2499-505. Centres for Disease Control and Prevention. (2010). "Morbidity and Mortality Weekly Report, Vol.59, No.11." from http://www.cdc.gov/mmwr/pdf/wk/mm5911.pdf. Chang, T. K., S. M. Bandiera and J. Chen (2003). Constitutive androstane receptor and pregnane X receptor gene expression in human liver: interindividual variability and correlation with CYP2B6 mRNA levels. Drug Metab Dispos 31(1): 7-10. Chen, G., A. S. Blevins-Primeau, R. W. Dellinger, J. E. Muscat and P. Lazarus (2007). Glucuronidation of nicotine and cotinine by UGT2B10: loss of function by the UGT2B10 Codon 67 (Asp>Tyr) polymorphism. Cancer Res 67(19): 9024-9. Chen, G., R. W. Dellinger, C. J. Gallagher, D. Sun and P. Lazarus (2008). Identification of a prevalent functional missense polymorphism in the UGT2B10 gene and its association with UGT2B10 inactivation against tobacco-specific nitrosamines. Pharmacogenet Genomics 18(3): 181-91. Chen, G. F., Y. M. Tang, B. Green, D. X. Lin, F. P. Guengerich, A. K. Daly, et al. (1999). Low frequency of CYP2A6 gene polymorphism as revealed by a one-step polymerase chain reaction method. Pharmacogenetics 9(3): 327-32. Cho, J. Y., H. S. Lim, J. Y. Chung, K. S. Yu, J. R. Kim, S. G. Shin, et al. (2004). Haplotype structure and allele frequencies of CYP2B6 in a Korean population. Drug Metab Dispos 32(12): 1341-4. Christian, K., M. Lang, P. Maurel and F. Raffalli-Mathieu (2004). Interaction of heterogeneous nuclear ribonucleoprotein A1 with cytochrome P450 2A6 mRNA: implications for post- transcriptional regulation of the CYP2A6 gene. Mol Pharmacol 65(6): 1405-14. Clarke, P. B. S., M. Quik, F. Adlkofer and K. Thurau (1995). Effects of nicotine on biological systems II. Berlin, Advances in Pharmacological Sciences.

231 Corrigall, W. A., K. B. Franklin, K. M. Coen and P. B. Clarke (1992). The mesolimbic dopaminergic system is implicated in the reinforcing effects of nicotine. Psychopharmacology (Berl) 107(2-3): 285-9. Corringer, P. J., S. Bertrand, S. Bohler, S. J. Edelstein, J. P. Changeux and D. Bertrand (1998). Critical elements determining diversity in agonist binding and desensitization of neuronal nicotinic acetylcholine receptors. J Neurosci 18(2): 648-57. Cosgrove, K. P., J. Batis, F. Bois, P. K. Maciejewski, I. Esterlis, T. Kloczynski, et al. (2009). beta2-Nicotinic acetylcholine receptor availability during acute and prolonged abstinence from . Arch Gen Psychiatry 66(6): 666-76. Coulie, B., M. Camilleri, A. E. Bharucha, W. J. Sandborn and D. Burton (2001). Colonic motility in chronic ulcerative proctosigmoiditis and the effects of nicotine on colonic motility in patients and healthy subjects. Aliment Pharmacol Ther 15(5): 653-63. Court, M. H., S. X. Duan, L. M. Hesse, K. Venkatakrishnan and D. J. Greenblatt (2001). Cytochrome P-450 2B6 is responsible for interindividual variability of propofol hydroxylation by human liver microsomes. Anesthesiology 94(1): 110-9. Crawford, E. L., D. A. Weaver, J. P. DeMuth, C. M. Jackson, S. A. Khuder, M. W. Frampton, et al. (1998). Measurement of cytochrome P450 2A6 and 2E1 gene expression in primary human bronchial epithelial cells. Carcinogenesis 19(10): 1867-71. Croom, E. L., J. C. Stevens, R. N. Hines, A. D. Wallace and E. Hodgson (2009). Human hepatic CYP2B6 developmental expression: the impact of age and genotype. Biochem Pharmacol 78(2): 184-90. Csajka, C., C. Marzolini, K. Fattinger, L. A. Decosterd, J. Fellay, A. Telenti, et al. (2003). Population pharmacokinetics and effects of efavirenz in patients with human immunodeficiency virus infection. Clin Pharmacol Ther 73(1): 20-30. Daigo, S., Y. Takahashi, M. Fujieda, N. Ariyoshi, H. Yamazaki, W. Koizumi, et al. (2002). A novel mutant allele of the CYP2A6 gene (CYP2A6*11 ) found in a cancer patient who showed poor metabolic phenotype towards tegafur. Pharmacogenetics 12(4): 299-306. Dannenberg, L. O. and H. J. Edenberg (2006). Epigenetics of gene expression in human hepatoma cells: expression profiling the response to inhibition of DNA methylation and histone deacetylation. BMC Genomics 7: 181. Davis, R. A., M. F. Stiles, J. D. deBethizy and J. H. Reynolds (1991). Dietary nicotine: a source of urinary cotinine. Food Chem Toxicol 29(12): 821-7. Dawson, E. B., D. R. Evans, W. A. Harris and W. J. McGanity (1999). The effect of ascorbic acid supplementation on the nicotine metabolism of smokers. Prev Med 29(6 Pt 1): 451- 4. Degtyarenko, K. N. and A. I. Archakov (1993). Molecular evolution of P450 superfamily and P450-containing monooxygenase systems. FEBS Lett 332(1-2): 1-8. Dempsey, D., P. Jacob, 3rd and N. L. Benowitz (2002). Accelerated metabolism of nicotine and cotinine in pregnant smokers. J Pharmacol Exp Ther 301(2): 594-8. Dempsey, D., P. Tutka, P. Jacob, 3rd, F. Allen, K. Schoedel, R. F. Tyndale, et al. (2004). Nicotine metabolite ratio as an index of cytochrome P450 2A6 metabolic activity. Clin Pharmacol Ther 76(1): 64-72. Denton, T. T., X. Zhang and J. R. Cashman (2004). Nicotine-related alkaloids and metabolites as inhibitors of human cytochrome P-450 2A6. Biochem Pharmacol 67(4): 751-6. Desta, Z., T. Saussele, B. Ward, J. Blievernicht, L. Li, K. Klein, et al. (2007). Impact of CYP2B6 polymorphism on hepatic efavirenz metabolism in vitro. Pharmacogenomics 8(6): 547-58.

232 Dhir, R. N., W. Dworakowski, C. Thangavel and B. H. Shapiro (2006). Sexually dimorphic regulation of hepatic isoforms of human cytochrome p450 by growth hormone. J Pharmacol Exp Ther 316(1): 87-94. Di Chiara, G. and A. Imperato (1988). Drugs abused by humans preferentially increase synaptic dopamine concentrations in the mesolimbic system of freely moving rats. Proc Natl Acad Sci U S A 85(14): 5274-8. di Iulio, J., A. Fayet, M. Arab-Alameddine, M. Rotger, R. Lubomirov, M. Cavassini, et al. (2009). In vivo analysis of efavirenz metabolism in individuals with impaired CYP2A6 function. Pharmacogenet Genomics 19(4): 300-9. Di, Y. M., V. D. Chow, L. P. Yang and S. F. Zhou (2009). Structure, Function, Regulation and Polymorphism of Human Cytochrome P450 2A6. Curr Drug Metab. Dicke, K. E., S. M. Skrlin and S. E. Murphy (2005). Nicotine and 4-(methylnitrosamino)-1-(3- pyridyl)-butanone metabolism by cytochrome P450 2B6. Drug Metab Dispos 33(12): 1760-4. Ding, S., B. G. Lake, T. Friedberg and C. R. Wolf (1995). Expression and alternative splicing of the cytochrome P-450 CYP2A7. Biochem J 306 ( Pt 1): 161-6. Ding, X. and L. S. Kaminsky (2003). Human extrahepatic cytochromes P450: function in xenobiotic metabolism and tissue-selective chemical toxicity in the respiratory and gastrointestinal tracts. Annu Rev Pharmacol Toxicol 43: 149-73. Domino, E. F., E. Hornbach and T. Demana (1993). The nicotine content of common vegetables. N Engl J Med 329(6): 437. Donato, M. T., P. Viitala, C. Rodriguez-Antona, A. Lindfors, J. V. Castell, H. Raunio, et al. (2000). CYP2A5/CYP2A6 expression in mouse and human hepatocytes treated with various in vivo inducers. Drug Metab Dispos 28(11): 1321-6. Doran, N., B. Spring, B. Borrelli, D. McChargue, B. Hitsman, R. Niaura, et al. (2006). Elevated positive mood: a mixed blessing for abstinence. Psychol Addict Behav 20(1): 36-43. Draper, A. J., A. Madan and A. Parkinson (1997). Inhibition of coumarin 7-hydroxylase activity in human liver microsomes. Arch Biochem Biophys 341(1): 47-61. Drocourt, L., J. C. Ourlin, J. M. Pascussi, P. Maurel and M. J. Vilarem (2002). Expression of CYP3A4, CYP2B6, and CYP2C9 is regulated by the vitamin D receptor pathway in primary human hepatocytes. J Biol Chem 277(28): 25125-32. Dube, S. R., K. Asman, A. Malarcher and R. Carabollo (2009). Cigarette Smoking Among Adults and Trends in Smoking Cessation: United States, 2008. Morbidity and Mortality Weekly Report 58(44): 1227-1232. Durcan, M. J., G. Deener, J. White, J. A. Johnston, D. Gonzales, R. Niaura, et al. (2002). The effect of bupropion sustained-release on cigarette craving after smoking cessation. Clin Ther 24(4): 540-51. Edwards, R. J., R. J. Price, P. S. Watts, A. B. Renwick, J. M. Tredger, A. R. Boobis, et al. (2003). Induction of cytochrome P450 enzymes in cultured precision-cut human liver slices. Drug Metab Dispos 31(3): 282-8. Engeland, C., C. Mahoney, E. Mohr, V. Ilivitsky and V. J. Knott (2002). Acute nicotine effects on auditory sensory memory in tacrine-treated and nontreated patients with Alzheimer's disease: an event-related potential study. Pharmacol Biochem Behav 72(1-2): 457-64. Epping-Jordan, M. P., S. S. Watkins, G. F. Koob and A. Markou (1998). Dramatic decreases in brain reward function during nicotine withdrawal. Nature 393(6680): 76-9. Evans, W. E. and H. L. McLeod (2003). Pharmacogenomics--drug disposition, drug targets, and side effects. N Engl J Med 348(6): 538-49. Ezzati, M. and A. D. Lopez (2003). Estimates of global mortality attributable to smoking in 2000. Lancet 362(9387): 847-52. 233 Faheim, S. (2009). Impact of genetic polymorphisms within the CYP2B6 gene on smoking behaiours. Department of Pharmacology and Toxicology. University of Toronto. Master of Science. Faucette, S. R., R. L. Hawke, E. L. Lecluyse, S. S. Shord, B. Yan, R. M. Laethem, et al. (2000). Validation of bupropion hydroxylation as a selective marker of human cytochrome P450 2B6 catalytic activity. Drug Metab Dispos 28(10): 1222-30. Faucette, S. R., H. Wang, G. A. Hamilton, S. L. Jolley, D. Gilbert, C. Lindley, et al. (2004). Regulation of CYP2B6 in primary human hepatocytes by prototypical inducers. Drug Metab Dispos 32(3): 348-58. Feng, Q., S. Vannaprasaht, Y. Peng, S. Angsuthum, Y. Avihingsanon, V. C. Yee, et al. (2010). Thiopurine S-methyltransferase pharmacogenetics: functional characterization of a novel rapidly degraded variant allozyme. Biochem Pharmacol 79(7): 1053-61. Fernandez-Salguero, P., S. M. Hoffman, S. Cholerton, H. Mohrenweiser, H. Raunio, A. Rautio, et al. (1995). A genetic polymorphism in coumarin 7-hydroxylation: sequence of the human CYP2A genes and identification of variant CYP2A6 alleles. Am J Hum Genet 57(3): 651-60. Feyerabend, C., R. M. Ings and M. A. Russel (1985). Nicotine pharmacokinetics and its application to intake from smoking. Br J Clin Pharmacol 19(2): 239-47. Filozof, C., M. C. Fernandez Pinilla and A. Fernandez-Cruz (2004). Smoking cessation and weight gain. Obes Rev 5(2): 95-103. Forrester, L. M., C. J. Henderson, M. J. Glancey, D. J. Back, B. K. Park, S. E. Ball, et al. (1992). Relative expression of cytochrome P450 isoenzymes in human liver and association with the metabolism of drugs and xenobiotics. Biochem J 281 ( Pt 2): 359-68. Fradette, C., N. Yamaguchi and P. Du Souich (2004). 5-Hydroxytryptamine is biotransformed by CYP2C9, 2C19 and 2B6 to hydroxylamine, which is converted into . Br J Pharmacol 141(3): 407-14. Fujieda, M., H. Yamazaki, T. Saito, K. Kiyotani, M. A. Gyamfi, M. Sakurai, et al. (2004). Evaluation of CYP2A6 genetic polymorphisms as determinants of smoking behavior and tobacco-related lung cancer risk in male Japanese smokers. Carcinogenesis 25(12): 2451-8. Fukami, T., M. Nakajima, E. Higashi, H. Yamanaka, H. L. McLeod and T. Yokoi (2005a). A novel CYP2A6*20 allele found in African-American population produces a truncated protein lacking enzymatic activity. Biochem Pharmacol 70(5): 801-8. Fukami, T., M. Nakajima, E. Higashi, H. Yamanaka, H. Sakai, H. L. McLeod, et al. (2005). Characterization of novel CYP2A6 polymorphic alleles (CYP2A6*18 and CYP2A6*19) that affect enzymatic activity. Drug Metab Dispos 33(8): 1202-10. Fukami, T., M. Nakajima, H. Sakai, H. L. McLeod and T. Yokoi (2006). CYP2A7 polymorphic alleles confound the genotyping of CYP2A6*4A allele. Pharmacogenomics J 6(6): 401- 12. Fukami, T., M. Nakajima, H. Yamanaka, Y. Fukushima, H. L. McLeod and T. Yokoi (2007). A novel duplication type of CYP2A6 gene in African-American population. Drug Metab Dispos 35(4): 515-20. Fukami, T., M. Nakajima, R. Yoshida, Y. Tsuchiya, Y. Fujiki, M. Katoh, et al. (2004). A novel polymorphism of human CYP2A6 gene CYP2A6*17 has an amino acid substitution (V365M) that decreases enzymatic activity in vitro and in vivo. Clin Pharmacol Ther 76(6): 519-27. Gad, S. C. (2008). Preclinical development handbook, ADME and biopharmaceutical properties, John Wiley & Sons.

234 Gaedigk, A., S. D. Simon, R. E. Pearce, L. D. Bradford, M. J. Kennedy and J. S. Leeder (2008). The CYP2D6 activity score: translating genotype information into a qualitative measure of phenotype. Clin Pharmacol Ther 83(2): 234-42. Gambier, N., A. M. Batt, B. Marie, M. Pfister, G. Siest and S. Visvikis-Siest (2005). Association of CYP2A6*1B genetic variant with the amount of smoking in French adults from the Stanislas cohort. Pharmacogenomics J 5(4): 271-5. Ganong, W. F. (2001). Review of Medical Physiology, Twentieth Edition, McGraw-Hill Companies. Garattini, E., M. Fratelli and M. Terao (2008). Mammalian aldehyde oxidases: genetics, evolution and biochemistry. Cell Mol Life Sci 65(7-8): 1019-48. Gardiner-Garden, M. and M. Frommer (1987). CpG islands in vertebrate genomes. J Mol Biol 196(2): 261-82. Garfinkel, D. (1958). Studies on pig liver microsomes. I. Enzymic and pigment composition of different microsomal fractions. Arch Biochem Biophys 77(2): 493-509. Gatanaga, H., T. Hayashida, K. Tsuchiya, M. Yoshino, T. Kuwahara, H. Tsukada, et al. (2007). Successful efavirenz dose reduction in HIV type 1-infected individuals with cytochrome P450 2B6 *6 and *26. Clin Infect Dis 45(9): 1230-7. Gervot, L., B. Rochat, J. C. Gautier, F. Bohnenstengel, H. Kroemer, V. de Berardinis, et al. (1999). Human CYP2B6: expression, inducibility and catalytic activities. Pharmacogenetics 9(3): 295-306. Ghosheh, O. and E. M. Hawes (2002). Microsomal N-glucuronidation of nicotine and cotinine: human hepatic interindividual, human intertissue, and interspecies hepatic variation. Drug Metab Dispos 30(12): 1478-83. Ghosheh, O. and E. M. Hawes (2002a). N-glucuronidation of nicotine and cotinine in human: formation of cotinine glucuronide in liver microsomes and lack of catalysis by 10 examined UDP-glucuronosyltransferases. Drug Metab Dispos 30(9): 991-6. Ghotbi, R., A. Gomez, L. Milani, G. Tybring, A. C. Syvanen, L. Bertilsson, et al. (2009). Allele- specific expression and gene methylation in the control of CYP1A2 mRNA level in human livers. Pharmacogenomics J 9(3): 208-17. Gillam, E. M. (1998). Human cytochrome P450 enzymes expressed in bacteria: reagents to probe molecular interactions in toxicology. Clin Exp Pharmacol Physiol 25(11): 877-86. Gillam, E. M., A. M. Aguinaldo, L. M. Notley, D. Kim, R. G. Mundkowski, A. A. Volkov, et al. (1999). Formation of indigo by recombinant mammalian cytochrome P450. Biochem Biophys Res Commun 265(2): 469-72. Gilmore, J. (2002). Report on smoking in Canda 1985 to 2001. H. S. D. Statistics Canada. Gilpin, E. A. and J. P. Pierce (2002). Demographic differences in patterns in the incidence of smoking cessation: United States 1950-1990. Ann Epidemiol 12(3): 141-50. Giovino, G. A., S. Sidney, J. C. Gfroerer, P. M. O'Malley, J. A. Allen, P. A. Richter, et al. (2004). Epidemiology of menthol cigarette use. Nicotine Tob Res 6 Suppl 1: S67-81. Gomes, A. M., S. Winter, K. Klein, M. Turpeinen, E. Schaeffeler, M. Schwab, et al. (2009a). Pharmacogenomics of human liver cytochrome P450 oxidoreductase: multifactorial analysis and impact on microsomal drug oxidation. Pharmacogenomics 10(4): 579-99. Gomez, A. and M. Ingelman-Sundberg (2009). Pharmacoepigenetics: its role in interindividual differences in drug response. Clin Pharmacol Ther 85(4): 426-30. Gomez, A., M. Karlgren, D. Edler, M. L. Bernal, S. Mkrtchian and M. Ingelman-Sundberg (2007). Expression of CYP2W1 in colon tumors: regulation by gene methylation. Pharmacogenomics 8(10): 1315-25.

235 Goodwin, B., L. B. Moore, C. M. Stoltz, D. D. McKee and S. A. Kliewer (2001). Regulation of the human CYP2B6 gene by the nuclear pregnane X receptor. Mol Pharmacol 60(3): 427-31. Goodz, S. D. and R. F. Tyndale (2002). Genotyping human CYP2A6 variants. Methods Enzymol 357: 59-69. Gori, G. B., N. L. Benowitz and C. J. Lynch (1986). Mouth versus deep airways absorption of nicotine in cigarette smokers. Pharmacol Biochem Behav 25(6): 1181-4. Gorrod, J. W. and A. R. Hibberd (1982). The metabolism of nicotine-delta 1'(5')-iminium ion, in vivo and in vitro. Eur J Drug Metab Pharmacokinet 7(4): 293-8. Gorrod, J. W. and P. Jacob III (1999). Analytical determination of nicotine and related compounds and their metabolites. Amesterdam; , Elsevier. Gorrod, J. W. and J. Wahren (1993). Nicotine and related alkaloids: absorption, distribution, metabolism, excretion. London, Chapman and Hall. Gourlay, S. G. and N. L. Benowitz (1996). The benefits of stopping smoking and the role of nicotine replacement therapy in older patients. Drugs Aging 9(1): 8-23. Gourlay, S. G. and N. L. Benowitz (1997). Arteriovenous differences in plasma concentration of nicotine and catecholamines and related cardiovascular effects after smoking, nicotine nasal spray, and intravenous nicotine. Clin Pharmacol Ther 62(4): 453-63. Gries, J. M., N. Benowitz and D. Verotta (1996). Chronopharmacokinetics of nicotine. Clin Pharmacol Ther 60(4): 385-95. Grunberg, N. E., D. J. Bowen, V. A. Maycock and S. M. Nespor (1985). The importance of sweet taste and caloric content in the effects of nicotine on specific food consumption. Psychopharmacology (Berl) 87(2): 198-203. Gu, D. F., L. J. Hinks, N. E. Morton and I. N. Day (2000). The use of long PCR to confirm three common alleles at the CYP2A6 locus and the relationship between genotype and smoking habit. Ann Hum Genet 64(Pt 5): 383-90. Gu, J., Y. Weng, Q. Y. Zhang, H. Cui, M. Behr, L. Wu, et al. (2003). Liver-specific deletion of the NADPH-cytochrome P450 reductase gene: impact on plasma homeostasis and the function and regulation of microsomal cytochrome P450 and heme . J Biol Chem 278(28): 25895-901. Guan, S., M. Huang, E. Chan, X. Chen, W. Duan and S. F. Zhou (2006a). Genetic polymorphisms of cytochrome P450 2B6 gene in Han Chinese. Eur J Pharm Sci 29(1): 14-21. Guan, S., M. Huang, X. Li, X. Chen, E. Chan and S. F. Zhou (2006). Intra- and inter-ethnic differences in the allele frequencies of cytochrome P450 2B6 gene in Chinese. Pharm Res 23(9): 1983-90. Gullsten, H., J. A. Agundez, J. Benitez, E. Laara, J. M. Ladero, M. Diaz-Rubio, et al. (1997). CYP2A6 gene polymorphism and risk of liver cancer and cirrhosis. Pharmacogenetics 7(3): 247-50. Gyamfi, M. A., M. Fujieda, K. Kiyotani, H. Yamazaki and T. Kamataki (2005). High prevalence of cytochrome P450 2A6*1A alleles in a black African population of Ghana. Eur J Clin Pharmacol 60(12): 855-7. Haberl, M., B. Anwald, K. Klein, R. Weil, C. Fuss, A. Gepdiremen, et al. (2005). Three haplotypes associated with CYP2A6 phenotypes in Caucasians. Pharmacogenet Genomics 15(9): 609-24. Hadidi, H., K. Zahlsen, J. R. Idle and S. Cholerton (1997). A single amino acid substitution (Leu160His) in cytochrome P450 CYP2A6 causes switching from 7-hydroxylation to 3- hydroxylation of coumarin. Food Chem Toxicol 35(9): 903-7.

236 Haiman, C. A., D. O. Stram, L. R. Wilkens, M. C. Pike, L. N. Kolonel, B. E. Henderson, et al. (2006). Ethnic and racial differences in the smoking-related risk of lung cancer. N Engl J Med 354(4): 333-42. Hakooz, N. and I. Hamdan (2007). Effects of dietary broccoli on human in vivo caffeine metabolism: a pilot study on a group of Jordanian volunteers. Curr Drug Metab 8(1): 9- 15. Hammond, S. K. (2009). Global patterns of nicotine and tobacco consumption. Handb Exp Pharmacol(192): 3-28. Hammons, G. J., Y. Yan-Sanders, B. Jin, E. Blann, F. F. Kadlubar and B. D. Lyn-Cook (2001). Specific site methylation in the 5'-flanking region of CYP1A2 interindividual differences in human livers. Life Sci 69(7): 839-45. Hanna, I. H., J. R. Reed, F. P. Guengerich and P. F. Hollenberg (2000). Expression of human cytochrome P450 2B6 in Escherichia coli: characterization of catalytic activity and expression levels in human liver. Arch Biochem Biophys 376(1): 206-16. HapMap (2003). The International HapMap Project. Nature 426(6968): 789-96. Hart, S. N., S. Wang, K. Nakamoto, C. Wesselman, Y. Li and X. B. Zhong (2008a). Genetic polymorphisms in cytochrome P450 oxidoreductase influence microsomal P450- catalyzed drug metabolism. Pharmacogenet Genomics 18(1): 11-24. Hart, S. N. and X. B. Zhong (2008). P450 oxidoreductase: genetic polymorphisms and implications for drug metabolism and toxicity. Expert Opin Drug Metab Toxicol 4(4): 439-52. Harvey, D. M., S. Yasar, S. J. Heishman, L. V. Panlilio, J. E. Henningfield and S. R. Goldberg (2004). Nicotine serves as an effective reinforcer of intravenous drug-taking behavior in human cigarette smokers. Psychopharmacology (Berl) 175(2): 134-42. Hasler, J., T. Samuelsson and K. Strub (2007). Useful 'junk': Alu RNAs in the human transcriptome. Cell Mol Life Sci 64(14): 1793-800. Hasler, J. and K. Strub (2006). Alu elements as regulators of gene expression. Nucleic Acids Res 34(19): 5491-7. Hecht, S. S., S. G. Carmella and S. E. Murphy (1999). Effects of watercress consumption on urinary metabolites of nicotine in smokers. Cancer Epidemiol Biomarkers Prev 8(10): 907-13. Henderson, C. J., D. M. Otto, D. Carrie, M. A. Magnuson, A. W. McLaren, I. Rosewell, et al. (2003). Inactivation of the hepatic cytochrome P450 system by conditional deletion of hepatic cytochrome P450 reductase. J Biol Chem 278(15): 13480-6. Henningfield, J. E. and S. R. Goldberg (1983). Nicotine as a reinforcer in human subjects and laboratory animals. Pharmacol Biochem Behav 19(6): 989-92. Henningfield, J. E. and R. M. Keenan (1993). Nicotine delivery kinetics and abuse liability. J Consult Clin Psychol 61(5): 743-50. Henningfield, J. E., K. Miyasato and D. R. Jasinski (1985). Abuse liability and pharmacodynamic characteristics of intravenous and inhaled nicotine. J Pharmacol Exp Ther 234(1): 1-12. Henningfield, J. E., J. M. Stapleton, N. L. Benowitz, R. F. Grayson and E. D. London (1993a). Higher levels of nicotine in arterial than in venous blood after cigarette smoking. Drug Alcohol Depend 33(1): 23-9. Herning, R. I., R. T. Jones, N. L. Benowitz and A. H. Mines (1983). How a cigarette is smoked determines blood nicotine levels. Clin Pharmacol Ther 33(1): 84-90. Hesse, L. M., P. He, S. Krishnaswamy, Q. Hao, K. Hogan, L. L. von Moltke, et al. (2004). Pharmacogenetic determinants of interindividual variability in bupropion hydroxylation by cytochrome P450 2B6 in human liver microsomes. Pharmacogenetics 14(4): 225-38. 237 Higashi, E., T. Fukami, M. Itoh, S. Kyo, M. Inoue, T. Yokoi, et al. (2007). Human CYP2A6 is induced by estrogen via estrogen receptor. Drug Metab Dispos 35(10): 1935-41. Hiratsuka, M., Y. Hinai, Y. Konno, H. Nozawa, S. Konno and M. Mizugaki (2004). Three novel single nucleotide polymorphisms (SNPs) of the CYP2B6 gene in Japanese individuals. Drug Metab Pharmacokinet 19(2): 155-8. Hiratsuka, M., Y. Takekuma, N. Endo, K. Narahara, S. I. Hamdy, Y. Kishikawa, et al. (2002). Allele and genotype frequencies of CYP2B6 and CYP3A5 in the Japanese population. Eur J Clin Pharmacol 58(6): 417-21. Hiroi, T., S. Imaoka and Y. Funae (1998). Dopamine formation from tyramine by CYP2D6. Biochem Biophys Res Commun 249(3): 838-43. Ho, M. K., J. C. Mwenifumbo, N. Al Koudsi, K. S. Okuyemi, J. S. Ahluwalia, N. L. Benowitz, et al. (2009). Association of nicotine metabolite ratio and CYP2A6 genotype with smoking cessation treatment in African-American light smokers. Clin Pharmacol Ther 85(6): 635-43. Ho, M. K., J. C. Mwenifumbo, B. Zhao, E. M. Gillam and R. F. Tyndale (2008). A novel CYP2A6 allele, CYP2A6*23, impairs enzyme function in vitro and in vivo and decreases smoking in a population of Black-African descent. Pharmacogenet Genomics 18(1): 67-75. Ho, M. K. and R. F. Tyndale (2007). Overview of the pharmacogenomics of cigarette smoking. Pharmacogenomics J 7(2): 81-98. Hodgson, E. and R. L. Rose (2007). The importance of cytochrome P450 2B6 in the human metabolism of environmental chemicals. Pharmacol Ther 113(2): 420-8. Hoffman, S. M., P. Fernandez-Salguero, F. J. Gonzalez and H. W. Mohrenweiser (1995). Organization and evolution of the cytochrome P450 CYP2A-2B-2F subfamily gene cluster on human chromosome 19. J Mol Evol 41(6): 894-900. Hoffman, S. M., D. R. Nelson and D. S. Keeney (2001). Organization, structure and evolution of the CYP2 gene cluster on human chromosome 19. Pharmacogenetics 11(8): 687-98. Hofmann, M. H., J. K. Blievernicht, K. Klein, T. Saussele, E. Schaeffeler, M. Schwab, et al. (2008). Aberrant splicing caused by single nucleotide polymorphism c.516G>T [Q172H], a marker of CYP2B6*6, is responsible for decreased expression and activity of CYP2B6 in liver. J Pharmacol Exp Ther 325(1): 284-92. Houston, J. B. (1994). Utility of in vitro drug metabolism data in predicting in vivo metabolic clearance. Biochem Pharmacol 47(9): 1469-79. Howson, A. L., S. Batth, V. Ilivitsky, A. Boisjoli, M. Jaworski, C. Mahoney, et al. (2004). Clinical and attentional effects of acute nicotine treatment in Tourette's syndrome. Eur Psychiatry 19(2): 102-12. Hsyu, P. H., A. Singh, T. D. Giargiari, J. A. Dunn, J. A. Ascher and J. A. Johnston (1997). Pharmacokinetics of bupropion and its metabolites in cigarette smokers versus nonsmokers. J Clin Pharmacol 37(8): 737-43. Huang, S., D. G. Cook, L. J. Hinks, X. H. Chen, S. Ye, J. A. Gilg, et al. (2005). CYP2A6, MAOA, DBH, DRD4, and 5HT2A genotypes, smoking behaviour and cotinine levels in 1518 UK adolescents. Pharmacogenet Genomics 15(12): 839-50. Hukkanen, J., S. G. Gourlay, S. Kenkare and N. L. Benowitz (2005a). Influence of menstrual cycle on cytochrome P450 2A6 activity and cardiovascular effects of nicotine. Clin Pharmacol Ther 77(3): 159-69. Hukkanen, J., P. Jacob Iii, M. Peng, D. Dempsey and N. L. Benowitz (2010). Effects of nicotine on cytochrome P450 2A6 and 2E1 activities. Br J Clin Pharmacol 69(2): 152-9. Hukkanen, J., P. Jacob, 3rd and N. L. Benowitz (2005). Metabolism and disposition kinetics of nicotine. Pharmacol Rev 57(1): 79-115. 238 Hukkanen, J., P. Jacob, 3rd and N. L. Benowitz (2006). Effect of grapefruit juice on cytochrome P450 2A6 and nicotine renal clearance. Clin Pharmacol Ther 80(5): 522-30. Iafrate, A. J., L. Feuk, M. N. Rivera, M. L. Listewnik, P. K. Donahoe, Y. Qi, et al. (2004). Detection of large-scale variation in the human genome. Nat Genet 36(9): 949-51. Ingelman-Sundberg, M., S. C. Sim, A. Gomez and C. Rodriguez-Antona (2007). Influence of cytochrome P450 polymorphisms on drug therapies: pharmacogenetic, pharmacoepigenetic and clinical aspects. Pharmacol Ther 116(3): 496-526. Ingram, J. R., J. Rhodes, B. K. Evans and G. A. Thomas (2005). Preliminary observations of oral nicotine therapy for inflammatory bowel disease: an open-label phase I-II study of tolerance. Inflamm Bowel Dis 11(12): 1092-6. Isaac, P. F. and M. J. Rand (1972). Cigarette smoking and plasma levels of nicotine. Nature 236(5345): 308-10. Iscan, M., H. Rostami, T. Guray, O. Pelkonen and A. Rautio (1994). Interindividual variability of coumarin 7-hydroxylation in a Turkish population. Eur J Clin Pharmacol 47(4): 315- 8. Itoh, M., M. Nakajima, E. Higashi, R. Yoshida, K. Nagata, Y. Yamazoe, et al. (2006). Induction of human CYP2A6 is mediated by the pregnane X receptor with peroxisome proliferator-activated receptor-gamma coactivator 1alpha. J Pharmacol Exp Ther 319(2): 693-702. Iwahashi, K., C. Waga and T. Takimoto (2004). Whole deletion of CYP2A6 gene (CYP2A6AST;4C) and smoking behavior. Neuropsychobiology 49(2): 101-4. Iwatsubo, T., N. Hirota, T. Ooie, H. Suzuki, N. Shimada, K. Chiba, et al. (1997). Prediction of in vivo drug metabolism in the human liver from in vitro metabolism data. Pharmacol Ther 73(2): 147-71. Jacob, R. M., E. C. Johnstone, M. J. Neville and R. T. Walton (2004). Identification of CYP2B6 sequence variants by use of multiplex PCR with allele-specific genotyping. Clin Chem 50(8): 1372-7. Jaffe, C. A., D. K. Turgeon, K. Lown, R. Demott-Friberg and P. B. Watkins (2002). Growth hormone secretion pattern is an independent regulator of growth hormone actions in humans. Am J Physiol Endocrinol Metab 283(5): E1008-15. Jamner, L. D., C. K. Whalen, S. E. Loughlin, R. Mermelstein, J. Audrain-McGovern, S. Krishnan-Sarin, et al. (2003). Tobacco use across the formative years: a road map to developmental vulnerabilities. Nicotine Tob Res 5 Suppl 1: S71-87. Jiang, Z., N. Dragin, L. F. Jorge-Nebert, M. V. Martin, F. P. Guengerich, E. Aklillu, et al. (2006). Search for an association between the human CYP1A2 genotype and CYP1A2 metabolic phenotype. Pharmacogenet Genomics 16(5): 359-67. Jinno, H., T. Tanaka-Kagawa, A. Ohno, Y. Makino, E. Matsushima, N. Hanioka, et al. (2003). Functional characterization of cytochrome P450 2B6 allelic variants. Drug Metab Dispos 31(4): 398-403. Johnson, E. F. and C. D. Stout (2005). Structural diversity of human xenobiotic-metabolizing cytochrome P450 . Biochem Biophys Res Commun 338(1): 331-6. Johnston, A. J., J. Ascher, R. Leadbetter, V. D. Schmith, D. K. Patel, M. Durcan, et al. (2002). Pharmacokinetic optimisation of sustained-release bupropion for smoking cessation. Drugs 62 Suppl 2: 11-24. Johnston, J. A., J. Fiedler-Kelly, E. D. Glover, D. P. Sachs, T. H. Grasela and J. DeVeaugh- Geiss (2001). Relationship between drug exposure and the efficacy and safety of bupropion sustained release for smoking cessation. Nicotine Tob Res 3(2): 131-40.

239 Johnstone, E., N. Benowitz, A. Cargill, R. Jacob, L. Hinks, I. Day, et al. (2006). Determinants of the rate of nicotine metabolism and effects on smoking behavior. Clin Pharmacol Ther 80(4): 319-30. Jones, P. A. and D. Takai (2001). The role of DNA methylation in mammalian epigenetics. Science 293(5532): 1068-70. Jorge-Nebert, L. F., Z. Jiang, R. Chakraborty, J. Watson, L. Jin, S. T. McGarvey, et al. (2010). Analysis of human CYP1A1 and CYP1A2 genes and their shared bidirectional promoter in eight world populations. Hum Mutat 31(1): 27-40. Jover, R., R. Bort, M. J. Gomez-Lechon and J. V. Castell (1998). Re-expression of C/EBP alpha induces CYP2B6, CYP2C9 and CYP2D6 genes in HepG2 cells. FEBS Lett 431(2): 227- 30. Jover, R., R. Bort, M. J. Gomez-Lechon and J. V. Castell (2001). Cytochrome P450 regulation by hepatocyte nuclear factor 4 in human hepatocytes: a study using adenovirus-mediated antisense targeting. Hepatology 33(3): 668-75. Juvonen, R. O., M. Iwasaki and M. Negishi (1991). Structural function of residue-209 in coumarin 7-hydroxylase (P450coh). Enzyme-kinetic studies and site-directed mutagenesis. J Biol Chem 266(25): 16431-5. Kaivosaari, S., P. Toivonen, L. M. Hesse, M. Koskinen, M. H. Court and M. Finel (2007). Nicotine glucuronidation and the human UDP-glucuronosyltransferase UGT2B10. Mol Pharmacol 72(3): 761-8. Kalivas, P. W., L. Churchill and M. A. Klitenick (1993). GABA and enkephalin projection from the nucleus accumbens and ventral pallidum to the ventral tegmental area. Neuroscience 57(4): 1047-60. Kalow, W. (2005). Pharmacogenomics: historical perspective and current status. Methods Mol Biol 311: 3-15. Kandel, D. B., M. C. Hu, C. Schaffran, J. R. Udry and N. L. Benowitz (2007). Urine nicotine metabolites and smoking behavior in a multiracial/multiethnic national sample of young adults. Am J Epidemiol 165(8): 901-10. Kawamoto, T., T. Sueyoshi, I. Zelko, R. Moore, K. Washburn and M. Negishi (1999). Phenobarbital-responsive nuclear translocation of the receptor CAR in induction of the CYP2B gene. Mol Cell Biol 19(9): 6318-22. Kemp, P. M., G. S. Sneed, C. E. George and R. F. Distefano (1997). Postmortem distribution of nicotine and cotinine from a case involving the simultaneous administration of multiple nicotine transdermal systems. J Anal Toxicol 21(4): 310-3. Kerkel, K., A. Spadola, E. Yuan, J. Kosek, L. Jiang, E. Hod, et al. (2008). Genomic surveys by methylation-sensitive SNP analysis identify sequence-dependent allele-specific DNA methylation. Nat Genet 40(7): 904-8. Kiefe, C. I., O. D. Williams, C. E. Lewis, J. J. Allison, P. Sekar and L. E. Wagenknecht (2001). Ten-year changes in smoking among young adults: are racial differences explained by socioeconomic factors in the CARDIA study? Am J Public Health 91(2): 213-8. Kim, M. S., T. Kondo, I. Takada, M. Y. Youn, Y. Yamamoto, S. Takahashi, et al. (2009). DNA demethylation in hormone-induced transcriptional derepression. Nature 461(7266): 1007-12. King, G., A. Polednak, R. B. Bendel, M. C. Vilsaint and S. B. Nahata (2004). Disparities in smoking cessation between African Americans and Whites: 1990-2000. Am J Public Health 94(11): 1965-71. Kinirons, M. T. and M. S. O'Mahony (2004). Drug metabolism and ageing. Br J Clin Pharmacol 57(5): 540-4.

240 Kirchheiner, J., C. Klein, I. Meineke, J. Sasse, U. M. Zanger, T. E. Murdter, et al. (2003). Bupropion and 4-OH-bupropion pharmacokinetics in relation to genetic polymorphisms in CYP2B6. Pharmacogenetics 13(10): 619-26. Kitagawa, K., N. Kunugita, T. Katoh, M. Yang and T. Kawamoto (1999). The significance of the homozygous CYP2A6 deletion on nicotine metabolism: a new genotyping method of CYP2A6 using a single PCR-RFLP. Biochem Biophys Res Commun 262(1): 146-51. Kitagawa, K., N. Kunugita, M. Kitagawa and T. Kawamoto (2001). CYP2A6*6, a novel polymorphism in cytochrome p450 2A6, has a single amino acid substitution (R128Q) that inactivates enzymatic activity. J Biol Chem 276(21): 17830-5. Kitamura, S., K. Sugihara and S. Ohta (2006). Drug-metabolizing ability of molybdenum hydroxylases. Drug Metab Pharmacokinet 21(2): 83-98. Kiyotani, K., M. Fujieda, H. Yamazaki, T. Shimada, F. P. Guengerich, A. Parkinson, et al. (2002). Twenty one novel single nucleotide polymorphisms (SNPs) of the CYP2A6 gene in Japanese and Caucasians. Drug Metab Pharmacokinet 17(5): 482-7. Kiyotani, K., H. Yamazaki, M. Fujieda, S. Iwano, K. Matsumura, S. Satarug, et al. (2003). Decreased coumarin 7-hydroxylase activities and CYP2A6 expression levels in humans caused by genetic polymorphism in CYP2A6 promoter region (CYP2A6*9). Pharmacogenetics 13(11): 689-95. Klein, K., T. Lang, T. Saussele, E. Barbosa-Sicard, W. H. Schunck, M. Eichelbaum, et al. (2005). Genetic variability of CYP2B6 in populations of African and Asian origin: allele frequencies, novel functional variants, and possible implications for anti-HIV therapy with efavirenz. Pharmacogenet Genomics 15(12): 861-73. Klingenberg, M. (1958). Pigments of rat liver microsomes. Arch Biochem Biophys 75(2): 376- 86. Klotz, U. (2009). Pharmacokinetics and drug metabolism in the elderly. Drug Metab Rev 41(2): 67-76. Kobayashi, K., J. Morita, K. Chiba, A. Wanibuchi, M. Kimura, S. Irie, et al. (2004). Pharmacogenetic roles of CYP2C19 and CYP2B6 in the metabolism of R- and S- mephobarbital in humans. Pharmacogenetics 14(8): 549-56. Koenigs, L. L., R. M. Peter, S. J. Thompson, A. E. Rettie and W. F. Trager (1997). Mechanism- based inactivation of human liver cytochrome P450 2A6 by 8-methoxypsoralen. Drug Metab Dispos 25(12): 1407-15. Kong, S. Y., H. S. Lim, B. H. Nam, M. C. Kook, Y. W. Kim, K. W. Ryu, et al. (2009). Association of CYP2A6 polymorphisms with S-1 plus docetaxel therapy outcomes in metastatic gastric cancer. Pharmacogenomics 10(7): 1147-55. Koob, G. F. (2000). Neurobiology of addiction. Toward the development of new therapies. Ann N Y Acad Sci 909: 170-85. Koob, G. F. and M. Le Moal (2006). Neurobiology of Addiction London, Elsevier Academic. Koskela, S., J. Hakkola, J. Hukkanen, O. Pelkonen, M. Sorri, A. Saranen, et al. (1999). Expression of CYP2A genes in human liver and extrahepatic tissues. Biochem Pharmacol 57(12): 1407-13. Kozlowski, L. T., N. Y. Mehta, C. T. Sweeney, S. S. Schwartz, G. P. Vogler, M. J. Jarvis, et al. (1998). Filter ventilation and nicotine content of tobacco in cigarettes from Canada, the United Kingdom, and the United States. Tob Control 7(4): 369-75. Kozlowski, L. T., D. A. Wilkinson, W. Skinner, C. Kent, T. Franklin and M. Pope (1989). Comparing tobacco cigarette dependence with other drug dependencies. Greater or equal 'difficulty quitting' and 'urges to use,' but less 'pleasure' from cigarettes. JAMA 261(6): 898-901.

241 Kreth, K., K. Kovar, M. Schwab and U. M. Zanger (2000). Identification of the human cytochromes P450 involved in the oxidative metabolism of "Ecstasy"-related designer drugs. Biochem Pharmacol 59(12): 1563-71. Krueger, S. K. and D. E. Williams (2005). Mammalian flavin-containing monooxygenases: structure/function, genetic polymorphisms and role in drug metabolism. Pharmacol Ther 106(3): 357-87. Kuehl, G. E. and S. E. Murphy (2003). N-glucuronidation of trans-3'-hydroxycotinine by human liver microsomes. Chem Res Toxicol 16(12): 1502-6. Kuehl, G. E. and S. E. Murphy (2003a). N-glucuronidation of nicotine and cotinine by human liver microsomes and heterologously expressed UDP-glucuronosyltransferases. Drug Metab Dispos 31(11): 1361-8. Kyerematen, G. A., M. Morgan, G. Warner, L. F. Martin and E. S. Vesell (1990). Metabolism of nicotine by hepatocytes. Biochem Pharmacol 40(8): 1747-56. Kyerematen, G. A. and E. S. Vesell (1991). Metabolism of nicotine. Drug Metab Rev 23(1-2): 3-41. Lamba, J. K. (2008). Pharmacogenetics of the constitutive androstane receptor. Pharmacogenomics 9(1): 71-83. Lamba, V., J. Lamba, K. Yasuda, S. Strom, J. Davila, M. L. Hancock, et al. (2003). Hepatic CYP2B6 expression: gender and ethnic differences and relationship to CYP2B6 genotype and CAR (constitutive androstane receptor) expression. J Pharmacol Exp Ther 307(3): 906-22. Lang, T., K. Klein, J. Fischer, A. K. Nussler, P. Neuhaus, U. Hofmann, et al. (2001). Extensive genetic polymorphism in the human CYP2B6 gene with impact on expression and function in human liver. Pharmacogenetics 11(5): 399-415. Larsen, F., G. Gundersen, R. Lopez and H. Prydz (1992). CpG islands as gene markers in the human genome. Genomics 13(4): 1095-107. Le Gal, A., Y. Dreano, D. Lucas and F. Berthou (2003). Diversity of selective environmental substrates for human cytochrome P450 2A6: alkoxyethers, nicotine, coumarin, N- nitrosodiethylamine, and N-nitrosobenzylmethylamine. Toxicol Lett 144(1): 77-91. Lea, R. A., S. Dickson and N. L. Benowitz (2006). Within-subject variation of the salivary 3HC/COT ratio in regular daily smokers: prospects for estimating CYP2A6 enzyme activity in large-scale surveys of nicotine metabolic rate. J Anal Toxicol 30(6): 386-9. Lee, A. M., C. Jepson, E. Hoffmann, L. Epstein, L. W. Hawk, C. Lerman, et al. (2007a). CYP2B6 genotype alters abstinence rates in a bupropion smoking cessation trial. Biol Psychiatry 62(6): 635-41. Lee, A. M., C. Jepson, P. G. Shields, N. Benowitz, C. Lerman and R. F. Tyndale (2007). CYP2B6 genotype does not alter nicotine metabolism, plasma levels, or abstinence with nicotine replacement therapy. Cancer Epidemiol Biomarkers Prev 16(6): 1312-4. Lee, B. L., N. L. Benowitz and P. Jacob, 3rd (1987). Influence of tobacco abstinence on the disposition kinetics and effects of nicotine. Clin Pharmacol Ther 41(4): 474-9. Lee, B. L., P. Jacob, 3rd, M. E. Jarvik and N. L. Benowitz (1989). Food and nicotine metabolism. Pharmacol Biochem Behav 33(3): 621-5. Lemay, S., S. Chouinard, P. Blanchet, H. Masson, V. Soland, A. Beuter, et al. (2004). Lack of efficacy of a nicotine transdermal treatment on motor and cognitive deficits in Parkinson's disease. Prog Neuropsychopharmacol Biol Psychiatry 28(1): 31-9. Lerman, C., C. Jepson, E. P. Wileyto, F. Patterson, R. Schnoll, M. Mroziewicz, et al. (2010). Genetic variation in nicotine metabolism predicts the efficacy of extended-duration transdermal nicotine therapy. Clin Pharmacol Ther 87(5): 553-7.

242 Lerman, C., R. Tyndale, F. Patterson, E. P. Wileyto, P. G. Shields, A. Pinto, et al. (2006). Nicotine metabolite ratio predicts efficacy of transdermal nicotine for smoking cessation. Clin Pharmacol Ther 79(6): 600-8. Levin, E. D., C. K. Conners, E. Sparrow, S. C. Hinton, D. Erhardt, W. H. Meck, et al. (1996). Nicotine effects on adults with attention-deficit/hyperactivity disorder. Psychopharmacology (Berl) 123(1): 55-63. Levin, E. D., S. S. Lawrence, A. Petro, K. Horton, A. H. Rezvani, F. J. Seidler, et al. (2007). Adolescent vs. adult-onset nicotine self-administration in male rats: duration of effect and differential nicotinic receptor correlates. Neurotoxicol Teratol 29(4): 458-65. Lewis, D. F. (2003). Essential requirements for substrate binding affinity and selectivity toward human CYP2 family enzymes. Arch Biochem Biophys 409(1): 32-44. Lewis, D. F., M. Dickins, B. G. Lake, P. J. Eddershaw, M. H. Tarbit and P. S. Goldfarb (1999). Molecular modelling of the human cytochrome P450 isoform CYP2A6 and investigations of CYP2A substrate selectivity. Toxicology 133(1): 1-33. Lewis, D. F. and P. Hlavica (2000). Interactions between redox partners in various cytochrome P450 systems: functional and structural aspects. Biochim Biophys Acta 1460(2-3): 353- 74. Lewis, D. F. and J. M. Pratt (1998). The P450 catalytic cycle and oxygenation mechanism. Drug Metab Rev 30(4): 739-86. Liddle, C., B. J. Goodwin, J. George, M. Tapner and G. C. Farrell (1998). Separate and interactive regulation of cytochrome P450 3A4 by triiodothyronine, dexamethasone, and growth hormone in cultured hepatocytes. J Clin Endocrinol Metab 83(7): 2411-6. Lieber, C. S. (1999). Microsomal ethanol-oxidizing system (MEOS): the first 30 years (1968- 1998)--a review. Alcohol Clin Exp Res 23(6): 991-1007. Ling, G., Y. Wei and X. Ding (2007). Transcriptional regulation of human CYP2A13 expression in the respiratory tract by CCAAT/enhancer binding protein and epigenetic modulation. Mol Pharmacol 71(3): 807-16. Lo, R., L. Burgoon, L. Macpherson, S. Ahmed and J. Matthews (2010). Estrogen receptor- dependent regulation of CYP2B6 in human breast cancer cells. Biochim Biophys Acta 1799(6): 469-79. Loboz, K. K., A. S. Gross, K. M. Williams, W. S. Liauw, R. O. Day, J. K. Blievernicht, et al. (2006). Cytochrome P450 2B6 activity as measured by bupropion hydroxylation: effect of induction by rifampin and ethnicity. Clin Pharmacol Ther 80(1): 75-84. London, S. J., J. R. Idle, A. K. Daly and G. A. Coetzee (1999). Genetic variation of CYP2A6, smoking, and risk of cancer. Lancet 353(9156): 898-9. Loriot, M. A., S. Rebuissou, M. Oscarson, S. Cenee, M. Miyamoto, N. Ariyoshi, et al. (2001). Genetic polymorphisms of cytochrome P450 2A6 in a case-control study on lung cancer in a French population. Pharmacogenetics 11(1): 39-44. Lucchesi, B. R., C. R. Schuster and G. S. Emley (1967). The role of nicotine as a determinant of cigarette smoking frequency in man with observations of certain cardiovascular effects associated with the tobacco alkaloid. Clin Pharmacol Ther 8(6): 789-96. Lyness, W. H., N. M. Friedle and K. E. Moore (1979). Destruction of dopaminergic nerve terminals in nucleus accumbens: effect on d-amphetamine self-administration. Pharmacol Biochem Behav 11(5): 553-6. MacDougall, J. M., K. Fandrick, X. Zhang, S. V. Serafin and J. R. Cashman (2003). Inhibition of human liver microsomal (S)-nicotine oxidation by (-)-menthol and analogues. Chem Res Toxicol 16(8): 988-93. Mackay, J. and M. Eriksen (2002). The Tobacco Atlas. Geneva, World Health Organization.

243 Madan, A., R. A. Graham, K. M. Carroll, D. R. Mudra, L. A. Burton, L. A. Krueger, et al. (2003). Effects of prototypical microsomal enzyme inducers on cytochrome P450 expression in cultured human hepatocytes. Drug Metab Dispos 31(4): 421-31. Maglich, J. M., C. M. Stoltz, B. Goodwin, D. Hawkins-Brown, J. T. Moore and S. A. Kliewer (2002). Nuclear pregnane x receptor and constitutive androstane receptor regulate overlapping but distinct sets of genes involved in xenobiotic detoxification. Mol Pharmacol 62(3): 638-46. Makinen, J., C. Frank, J. Jyrkkarinne, J. Gynther, C. Carlberg and P. Honkakoski (2002). Modulation of mouse and human phenobarbital-responsive enhancer module by nuclear receptors. Mol Pharmacol 62(2): 366-78. Malaiyandi, V., S. D. Goodz, E. M. Sellers and R. F. Tyndale (2006a). CYP2A6 genotype, phenotype, and the use of nicotine metabolites as biomarkers during ad libitum smoking. Cancer Epidemiol Biomarkers Prev 15(10): 1812-9. Malaiyandi, V., C. Lerman, N. L. Benowitz, C. Jepson, F. Patterson and R. F. Tyndale (2006). Impact of CYP2A6 genotype on pretreatment smoking behaviour and nicotine levels from and usage of nicotine replacement therapy. Mol Psychiatry 11(4): 400-9. Malaiyandi, V., E. M. Sellers and R. F. Tyndale (2005). Implications of CYP2A6 genetic variation for smoking behaviors and nicotine dependence. Clin Pharmacol Ther 77(3): 145-58. Marks, M. J., J. R. Pauly, S. D. Gross, E. S. Deneris, I. Hermans-Borgmeyer, S. F. Heinemann, et al. (1992). Nicotine binding and nicotinic receptor subunit RNA after chronic nicotine treatment. J Neurosci 12(7): 2765-84. Marzolini, C., A. Telenti, L. A. Decosterd, G. Greub, J. Biollaz and T. Buclin (2001). Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS 15(1): 71-5. Matzke, M. A., M. F. Mette, W. Aufsatz, J. Jakowitsch and A. J. Matzke (1999). Host defenses to parasitic sequences and the evolution of epigenetic control mechanisms. Genetica 107(1-3): 271-87. Maurice, M., S. Emiliani, I. Dalet-Beluche, J. Derancourt and R. Lange (1991). Isolation and characterization of a cytochrome P450 of the IIA subfamily from human liver microsomes. Eur J Biochem 200(2): 511-7. McDonald, O. G. and W. E. Evans (2005). Haplotype structure and pharmacogenomics. Pharmacogenomics Second Edition, pg.441-57, Taylor & Francis Group. McGehee, D. S., M. J. Heath, S. Gelber, P. Devay and L. W. Role (1995). Nicotine enhancement of fast excitatory synaptic transmission in CNS by presynaptic receptors. Science 269(5231): 1692-6. McGrath, J., J. W. McDonald and J. K. Macdonald (2004). Transdermal nicotine for induction of remission in ulcerative colitis. Cochrane Database Syst Rev(4): CD004722. Mehlotra, R. K., M. J. Bockarie and P. A. Zimmerman (2007). CYP2B6 983T>C polymorphism is prevalent in West Africa but absent in Papua New Guinea: implications for HIV/AIDS treatment. Br J Clin Pharmacol 64(3): 391-5. Mehlotra, R. K., M. N. Ziats, M. J. Bockarie and P. A. Zimmerman (2006). Prevalence of CYP2B6 alleles in malaria-endemic populations of West Africa and Papua New Guinea. Eur J Clin Pharmacol 62(4): 267-75. Meier, P. J., H. K. Mueller, B. Dick and U. A. Meyer (1983). Hepatic monooxygenase activities in subjects with a genetic defect in drug oxidation. Gastroenterology 85(3): 682-92. Mendelson, J. H., N. Goletiani, M. B. Sholar, A. J. Siegel and N. K. Mello (2008). Effects of smoking successive low- and high-nicotine cigarettes on hypothalamic-pituitary-adrenal axis hormones and mood in men. Neuropsychopharmacology 33(4): 749-60. 244 Messina, E. S., R. F. Tyndale and E. M. Sellers (1997). A major role for CYP2A6 in nicotine C- oxidation by human liver microsomes. J Pharmacol Exp Ther 282(3): 1608-14. Meunier, V., M. Bourrie, B. Julian, E. Marti, F. Guillou, Y. Berger, et al. (2000). Expression and induction of CYP1A1/1A2, CYP2A6 and CYP3A4 in primary cultures of human hepatocytes: a 10-year follow-up. Xenobiotica 30(6): 589-607. Miksys, S., C. Lerman, P. G. Shields, D. C. Mash and R. F. Tyndale (2003). Smoking, alcoholism and genetic polymorphisms alter CYP2B6 levels in human brain. Neuropharmacology 45(1): 122-32. Miles, J. S., W. Bickmore, J. D. Brook, A. W. McLaren, R. Meehan and C. R. Wolf (1989). Close linkage of the human cytochrome P450IIA and P450IIB gene subfamilies: implications for the assignment of substrate specificity. Nucleic Acids Res 17(8): 2907- 17. Miles, J. S., A. W. McLaren, L. M. Forrester, M. J. Glancey, M. A. Lang and C. R. Wolf (1990). Identification of the human liver cytochrome P-450 responsible for coumarin 7- hydroxylase activity. Biochem J 267(2): 365-71. Mimura, M., T. Baba, H. Yamazaki, S. Ohmori, Y. Inui, F. J. Gonzalez, et al. (1993). Characterization of cytochrome P-450 2B6 in human liver microsomes. Drug Metab Dispos 21(6): 1048-56. Minematsu, N., H. Nakamura, M. Furuuchi, T. Nakajima, S. Takahashi, H. Tateno, et al. (2006). Limitation of cigarette consumption by CYP2A6*4, *7 and *9 polymorphisms. Eur Respir J 27(2): 289-92. Miyajima, A., T. Furihata and K. Chiba (2009). Functional analysis of GC Box and its CpG methylation in the regulation of CYP1A2 gene expression. Drug Metab Pharmacokinet 24(3): 269-76. Mo, S. L., Y. H. Liu, W. Duan, M. Q. Wei, J. R. Kanwar and S. F. Zhou (2009). Substrate Specificity, Regulation, and Polymorphism of Human Cytochrome P450 2B6. Curr Drug Metab. Molander, L., A. Hansson and E. Lunell (2001). Pharmacokinetics of nicotine in healthy elderly people. Clin Pharmacol Ther 69(1): 57-65. Molander, L., A. Hansson, E. Lunell, L. Alainentalo, M. Hoffmann and R. Larsson (2000). Pharmacokinetics of nicotine in kidney failure. Clin Pharmacol Ther 68(3): 250-60. Moolchan, E. T., F. H. Franken and M. Jaszyna-Gasior (2006). Adolescent nicotine metabolism: ethnoracial differences among dependent smokers. Ethn Dis 16(1): 239-43. Muangmoonchai, R., D. Smirlis, S. C. Wong, M. Edwards, I. R. Phillips and E. A. Shephard (2001). Xenobiotic induction of cytochrome P450 2B1 (CYP2B1) is mediated by the orphan nuclear receptor constitutive androstane receptor (CAR) and requires steroid co- activator 1 (SRC-1) and the transcription factor Sp1. Biochem J 355(Pt 1): 71-8. Munzel, P. A., S. Schmohl, H. Heel, K. Kalberer, B. S. Bock-Hennig and K. W. Bock (1999). Induction of human UDP glucuronosyltransferases (UGT1A6, UGT1A9, and UGT2B7) by t-butylhydroquinone and 2,3,7,8-tetrachlorodibenzo-p-dioxin in Caco-2 cells. Drug Metab Dispos 27(5): 569-73. Murayama, N., M. Shimada, Y. Yamazoe and R. Kato (1991). Difference in the susceptibility of two phenobarbital-inducible forms, P450IIB1 and P450IIB2, to thyroid hormone- and growth hormone-induced suppression in rat liver: phenobarbital-inducible P450IIB2 suppression by thyroid hormone acting directly, but not through the pituitary system. Mol Pharmacol 39(6): 811-7. Murphy, P. J. (1973). Enzymatic oxidation of nicotine to nicotine 1'(5') iminium ion. A newly discovered intermediate in the metabolism of nicotine. J Biol Chem 248(8): 2796-800.

245 Murphy, S. E., L. M. Johnson, L. M. Losey, S. G. Carmella and S. S. Hecht (2001). Consumption of watercress fails to alter coumarin metabolism in humans. Drug Metab Dispos 29(6): 786-8. Murphy, S. E., V. Raulinaitis and K. M. Brown (2005). Nicotine 5'-oxidation and methyl oxidation by P450 2A enzymes. Drug Metab Dispos 33(8): 1166-73. Mwenifumbo, J. C., N. Al Koudsi, M. K. Ho, Q. Zhou, E. B. Hoffmann, E. M. Sellers, et al. (2008). Novel and established CYP2A6 alleles impair in vivo nicotine metabolism in a population of Black African descent. Hum Mutat 29(5): 679-88. Mwenifumbo, J. C., C. N. Lessov-Schlaggar, Q. Zhou, R. E. Krasnow, G. E. Swan, N. L. Benowitz, et al. (2008a). Identification of novel CYP2A6*1B variants: the CYP2A6*1B allele is associated with faster in vivo nicotine metabolism. Clin Pharmacol Ther 83(1): 115-21. Mwenifumbo, J. C., M. G. Myers, T. L. Wall, S. K. Lin, E. M. Sellers and R. F. Tyndale (2005). Ethnic variation in CYP2A6*7, CYP2A6*8 and CYP2A6*10 as assessed with a novel haplotyping method. Pharmacogenet Genomics 15(3): 189-92. Mwenifumbo, J. C., E. M. Sellers and R. F. Tyndale (2007a). Nicotine metabolism and CYP2A6 activity in a population of black African descent: impact of gender and light smoking. Drug Alcohol Depend 89(1): 24-33. Mwenifumbo, J. C. and R. F. Tyndale (2007). Genetic variability in CYP2A6 and the pharmacokinetics of nicotine. Pharmacogenomics 8(10): 1385-402. Mwenifumbo, J. C. and R. F. Tyndale (2009). Molecular genetics of nicotine metabolism. Handb Exp Pharmacol(192): 235-59. Mwenifumbo, J. C., Q. Zhou, N. L. Benowitz, E. M. Sellers and R. F. Tyndale (2010). New CYP2A6 gene deletion and conversion variants in a population of Black African descent. Pharmacogenomics 11(2): 189-98. Nakajima, M., T. Fukami, H. Yamanaka, E. Higashi, H. Sakai, R. Yoshida, et al. (2006). Comprehensive evaluation of variability in nicotine metabolism and CYP2A6 polymorphic alleles in four ethnic populations. Clin Pharmacol Ther 80(3): 282-97. Nakajima, M., M. Itoh, H. Yamanaka, T. Fukami, S. Tokudome, Y. Yamamoto, et al. (2006a). Isoflavones inhibit nicotine C-oxidation catalyzed by human CYP2A6. J Clin Pharmacol 46(3): 337-44. Nakajima, M., S. Komagata, Y. Fujiki, Y. Kanada, H. Ebi, K. Itoh, et al. (2007). Genetic polymorphisms of CYP2B6 affect the pharmacokinetics/pharmacodynamics of cyclophosphamide in Japanese cancer patients. Pharmacogenet Genomics 17(6): 431-45. Nakajima, M., E. Tanaka, J. T. Kwon and T. Yokoi (2002). Characterization of nicotine and cotinine N-glucuronidations in human liver microsomes. Drug Metab Dispos 30(12): 1484-90. Nakajima, M., S. Yamagishi, H. Yamamoto, T. Yamamoto, Y. Kuroiwa and T. Yokoi (2000). Deficient cotinine formation from nicotine is attributed to the whole deletion of the CYP2A6 gene in humans. Clin Pharmacol Ther 67(1): 57-69. Nakajima, M., T. Yamamoto, K. Nunoya, T. Yokoi, K. Nagashima, K. Inoue, et al. (1996). Characterization of CYP2A6 involved in 3'-hydroxylation of cotinine in human liver microsomes. J Pharmacol Exp Ther 277(2): 1010-5. Nakajima, M., T. Yamamoto, K. Nunoya, T. Yokoi, K. Nagashima, K. Inoue, et al. (1996a). Role of human cytochrome P4502A6 in C-oxidation of nicotine. Drug Metab Dispos 24(11): 1212-7. Nakajima, M., R. Yoshida, T. Fukami, H. L. McLeod and T. Yokoi (2004). Novel human CYP2A6 alleles confound gene deletion analysis. FEBS Lett 569(1-3): 75-81.

246 Nakamura, K., N. Ariyoshi, T. Yokoi, S. Ohgiya, M. Chida, K. Nagashima, et al. (2002). CYP2D6.10 present in human liver microsomes shows low catalytic activity and thermal stability. Biochem Biophys Res Commun 293(3): 969-73. Nebert, D. W., D. R. Nelson and R. Feyereisen (1989a). Evolution of the cytochrome P450 genes. Xenobiotica 19(10): 1149-60. Nebert, D. W. and D. W. Russell (2002). Clinical importance of the cytochromes P450. Lancet 360(9340): 1155-62. Nelson, D. R. (2009). The cytochrome p450 homepage. Hum Genomics 4(1): 59-65. Nelson, D. R., T. Kamataki, D. J. Waxman, F. P. Guengerich, R. W. Estabrook, R. Feyereisen, et al. (1993). The P450 superfamily: update on new sequences, gene mapping, accession numbers, early trivial names of enzymes, and nomenclature. DNA Cell Biol 12(1): 1-51. Nelson, D. R., L. Koymans, T. Kamataki, J. J. Stegeman, R. Feyereisen, D. J. Waxman, et al. (1996). P450 superfamily: update on new sequences, gene mapping, accession numbers and nomenclature. Pharmacogenetics 6(1): 1-42. Norris, J., D. Fan, C. Aleman, J. R. Marks, P. A. Futreal, R. W. Wiseman, et al. (1995). Identification of a new subclass of Alu DNA repeats which can function as estrogen receptor-dependent transcriptional enhancers. J Biol Chem 270(39): 22777-82. Novotny, T. E., K. E. Warner, J. S. Kendrick and P. L. Remington (1988). Smoking by blacks and whites: socioeconomic and demographic differences. Am J Public Health 78(9): 1187-9. Nunoya, K., T. Yokoi, Y. Takahashi, K. Kimura, M. Kinoshita and T. Kamataki (1999). Homologous unequal cross-over within the human CYP2A gene cluster as a mechanism for the deletion of the entire CYP2A6 gene associated with the poor metabolizer phenotype. J Biochem 126(2): 402-7. O'Dell, L. E., A. W. Bruijnzeel, S. Ghozland, A. Markou and G. F. Koob (2004). Nicotine withdrawal in adolescent and adult rats. Ann N Y Acad Sci 1021: 167-74. O'Loughlin, J., G. Paradis, W. Kim, J. DiFranza, G. Meshefedjian, E. McMillan-Davey, et al. (2004). Genetically decreased CYP2A6 and the risk of tobacco dependence: a prospective study of novice smokers. Tob Control 13(4): 422-8. Obach, R. S., J. G. Baxter, T. E. Liston, B. M. Silber, B. C. Jones, F. MacIntyre, et al. (1997). The prediction of human pharmacokinetic parameters from preclinical and in vitro metabolism data. J Pharmacol Exp Ther 283(1): 46-58. Obach, R. S., P. Huynh, M. C. Allen and C. Beedham (2004). Human liver aldehyde oxidase: inhibition by 239 drugs. J Clin Pharmacol 44(1): 7-19. Oesch-Bartlmowicz, B. and F. Oesch (2004). Modulation of mutagenicity by phosphorylation of mutagen-metabolizing enzymes. Arch Biochem Biophys 423(1): 31-6. Oesch-Bartlomowicz, B., B. Richter, R. Becker, S. Vogel, P. R. Padma, J. G. Hengstler, et al. (2001). cAMP-dependent phosphorylation of CYP2B1 as a functional switch for cyclophosphamide activation and its hormonal control in vitro and in vivo. Int J Cancer 94(5): 733-42. Office of Applied Studies. (2006). "Past month cigarette use among racial and ethnic groups. ." and mental health services administration, Rockville, Maryland. , from http://www.oas.samhsa.gov/2002k2006/raceCigs/raceCigs.pdf. Okino, S. T., D. Pookot, L. C. Li, H. Zhao, S. Urakami, H. Shiina, et al. (2006). Epigenetic inactivation of the dioxin-responsive cytochrome P4501A1 gene in human prostate cancer. Cancer Res 66(15): 7420-8. Omura, T. and R. Sato (1962). A new cytochrome in liver microsomes. J Biol Chem 237: 1375- 6.

247 Omura, T. and R. Sato (1964). The Carbon Monoxide-Binding Pigment of Liver Microsomes. Ii. Solubilization, Purification, and Properties. J Biol Chem 239: 2379-85. Oneda, B., S. Crettol, E. Jaquenoud Sirot, M. Bochud, N. Ansermot and C. B. Eap (2009). The P450 oxidoreductase genotype is associated with CYP3A activity in vivo as measured by the midazolam phenotyping test. Pharmacogenet Genomics 19(11): 877-83. Onica, T., K. Nichols, M. Larin, L. Ng, A. Maslen, Z. Dvorak, et al. (2008). Dexamethasone- mediated up-regulation of human CYP2A6 involves the glucocorticoid receptor and increased binding of hepatic nuclear factor 4 alpha to the proximal promoter. Mol Pharmacol 73(2): 451-60. Orth, M., B. Amann, M. M. Robertson and J. C. Rothwell (2005). Excitability of motor cortex inhibitory circuits in Tourette syndrome before and after single dose nicotine. Brain 128(Pt 6): 1292-300. Oscarson, M., H. Gullsten, A. Rautio, M. L. Bernal, B. Sinues, M. L. Dahl, et al. (1998). Genotyping of human cytochrome P450 2A6 (CYP2A6), a nicotine C-oxidase. FEBS Lett 438(3): 201-5. Oscarson, M., R. A. McLellan, V. Asp, M. Ledesma, M. L. Bernal Ruiz, B. Sinues, et al. (2002). Characterization of a novel CYP2A7/CYP2A6 hybrid allele (CYP2A6*12) that causes reduced CYP2A6 activity. Hum Mutat 20(4): 275-83. Oscarson, M., R. A. McLellan, H. Gullsten, J. A. Agundez, J. Benitez, A. Rautio, et al. (1999). Identification and characterisation of novel polymorphisms in the CYP2A locus: implications for nicotine metabolism. FEBS Lett 460(2): 321-7. Oscarson, M., R. A. McLellan, H. Gullsten, Q. Y. Yue, M. A. Lang, M. L. Bernal, et al. (1999a). Characterisation and PCR-based detection of a CYP2A6 gene deletion found at a high frequency in a Chinese population. FEBS Lett 448(1): 105-10. Overby, L. H., G. C. Carver and R. M. Philpot (1997). Quantitation and kinetic properties of hepatic microsomal and recombinant flavin-containing monooxygenases 3 and 5 from humans. Chem Biol Interact 106(1): 29-45. Palovaara, S., O. Pelkonen, J. Uusitalo, S. Lundgren and K. Laine (2003). Inhibition of cytochrome P450 2B6 activity by hormone replacement therapy and oral contraceptive as measured by bupropion hydroxylation. Clin Pharmacol Ther 74(4): 326-33. Parkinson, A., D. R. Mudra, C. Johnson, A. Dwyer and K. M. Carroll (2004). The effects of gender, age, ethnicity, and liver cirrhosis on cytochrome P450 enzyme activity in human liver microsomes and inducibility in cultured human hepatocytes. Toxicol Appl Pharmacol 199(3): 193-209. Pasanen, M., Z. Rannala, A. Tooming, E. A. Sotaniemi, O. Pelkonen and A. Rautio (1997). Hepatitis A impairs the function of human hepatic CYP2A6 in vivo. Toxicology 123(3): 177-84. Pascussi, J. M., L. Drocourt, J. M. Fabre, P. Maurel and M. J. Vilarem (2000a). Dexamethasone induces pregnane X receptor and retinoid X receptor-alpha expression in human hepatocytes: synergistic increase of CYP3A4 induction by pregnane X receptor activators. Mol Pharmacol 58(2): 361-72. Pascussi, J. M., S. Gerbal-Chaloin, J. M. Fabre, P. Maurel and M. J. Vilarem (2000b). Dexamethasone enhances constitutive androstane receptor expression in human hepatocytes: consequences on cytochrome P450 gene regulation. Mol Pharmacol 58(6): 1441-50. Patten, C. J., T. J. Smith, M. J. Friesen, R. E. Tynes, C. S. Yang and S. E. Murphy (1997). Evidence for cytochrome P450 2A6 and 3A4 as major catalysts for N'-nitrosonornicotine alpha-hydroxylation by human liver microsomes. Carcinogenesis 18(8): 1623-30.

248 Patten, C. J., T. J. Smith, S. E. Murphy, M. H. Wang, J. Lee, R. E. Tynes, et al. (1996). Kinetic analysis of the activation of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone by heterologously expressed human P450 enzymes and the effect of P450-specific chemical inhibitors on this activation in human liver microsomes. Arch Biochem Biophys 333(1): 127-38. Patterson, F., R. A. Schnoll, E. P. Wileyto, A. Pinto, L. H. Epstein, P. G. Shields, et al. (2008). Toward personalized therapy for smoking cessation: a randomized placebo-controlled trial of bupropion. Clin Pharmacol Ther 84(3): 320-5. Peamkrasatam, S., K. Sriwatanakul, K. Kiyotani, M. Fujieda, H. Yamazaki, T. Kamataki, et al. (2006). In vivo evaluation of coumarin and nicotine as probe drugs to predict the metabolic capacity of CYP2A6 due to genetic polymorphism in Thais. Drug Metab Pharmacokinet 21(6): 475-84. Pearce, R., D. Greenway and A. Parkinson (1992). Species differences and interindividual variation in liver microsomal cytochrome P450 2A enzymes: effects on coumarin, dicumarol, and oxidation. Arch Biochem Biophys 298(1): 211-25. Pearce, R. E., C. J. McIntyre, A. Madan, U. Sanzgiri, A. J. Draper, P. L. Bullock, et al. (1996). Effects of freezing, thawing, and storing human liver microsomes on cytochrome P450 activity. Arch Biochem Biophys 331(2): 145-69. Pelkonen, O., A. Rautio, H. Raunio and M. Pasanen (2000). CYP2A6: a human coumarin 7- hydroxylase. Toxicology 144(1-3): 139-47. Pelkonen, O., E. A. Sotaniemi and J. T. Ahokas (1985). Coumarin 7-hydroxylase activity in human liver microsomes. Properties of the enzyme and interspecies comparisons. Br J Clin Pharmacol 19(1): 59-66. Pelkonen, O., M. Turpeinen, J. Hakkola, P. Honkakoski, J. Hukkanen and H. Raunio (2008). Inhibition and induction of human cytochrome P450 enzymes: current status. Arch Toxicol 82(10): 667-715. Perez-Stable, E. J., B. Herrera, P. Jacob, 3rd and N. L. Benowitz (1998). Nicotine metabolism and intake in black and white smokers. JAMA 280(2): 152-6. Perkins, K. A. (2001b). Smoking cessation in women. Special considerations. CNS Drugs 15(5): 391-411. Perkins, K. A. (2009). Discriminative stimulus effects of nicotine in humans. Handb Exp Pharmacol(192): 369-400. Perkins, K. A., E. Donny and A. R. Caggiula (1999). Sex differences in nicotine effects and self- administration: review of human and animal evidence. Nicotine Tob Res 1(4): 301-15. Perkins, K. A., L. H. Epstein, R. L. Stiller, M. H. Fernstrom, J. E. Sexton and R. G. Jacob (1990a). Perception and hedonics of sweet and fat taste in smokers and nonsmokers following nicotine intake. Pharmacol Biochem Behav 35(3): 671-6. Perkins, K. A., L. H. Epstein, R. L. Stiller, J. E. Sexton, M. H. Fernstrom, R. G. Jacob, et al. (1990b). Metabolic effects of nicotine after consumption of a meal in smokers and nonsmokers. Am J Clin Nutr 52(2): 228-33. Perkins, K. A., D. Gerlach, M. Broge, C. Fonte and A. Wilson (2001). Reinforcing effects of nicotine as a function of smoking status. Exp Clin Psychopharmacol 9(3): 243-50. Perkins, K. A., D. Gerlach, J. Vender, J. Grobe, J. Meeker and S. Hutchison (2001a). Sex differences in the subjective and reinforcing effects of visual and olfactory cigarette smoke stimuli. Nicotine Tob Res 3(2): 141-50. Perkins, K. A., M. Sanders, D. D'Amico and A. Wilson (1997). Nicotine discrimination and self-administration in humans as a function of smoking status. Psychopharmacology (Berl) 131(4): 361-70.

249 Perkins, K. A. and J. Scott (2008). Sex differences in long-term smoking cessation rates due to nicotine patch. Nicotine Tob Res 10(7): 1245-50. Perry, D. C., M. I. Davila-Garcia, C. A. Stockmeier and K. J. Kellar (1999). Increased nicotinic receptors in brains from smokers: membrane binding and autoradiography studies. J Pharmacol Exp Ther 289(3): 1545-52. Peterson, L. A., A. Trevor and N. Castagnoli, Jr. (1987). Stereochemical studies on the cytochrome P-450 catalyzed oxidation of (S)-nicotine to the (S)-nicotine delta 1'(5')- iminium species. J Med Chem 30(2): 249-54. Pichard-Garcia, L., R. Hyland, J. Baulieu, J. M. Fabre, A. Milton and P. Maurel (2000). Human hepatocytes in primary culture predict lack of cytochrome P-450 3A4 induction by eletriptan in vivo. Drug Metab Dispos 28(1): 51-7. Pitarque, M., C. Rodriguez-Antona, M. Oscarson and M. Ingelman-Sundberg (2005). Transcriptional regulation of the human CYP2A6 gene. J Pharmacol Exp Ther 313(2): 814-22. Pitarque, M., O. von Richter, B. Oke, H. Berkkan, M. Oscarson and M. Ingelman-Sundberg (2001). Identification of a single nucleotide polymorphism in the TATA box of the CYP2A6 gene: impairment of its promoter activity. Biochem Biophys Res Commun 284(2): 455-60. Pitarque, M., O. von Richter, C. Rodriguez-Antona, J. Wang, M. Oscarson and M. Ingelman- Sundberg (2004). A nicotine C-oxidase gene (CYP2A6) polymorphism important for promoter activity. Hum Mutat 23(3): 258-66. Poland, R. E., R. N. Pechnick, C. C. Cloak, Y. J. Wan, I. Nuccio and K. M. Lin (2000). Effect of cigarette smoking on coumarin metabolism in humans. Nicotine Tob Res 2(4): 351-4. Porter, T. D. and M. J. Coon (1991). Cytochrome P-450. Multiplicity of isoforms, substrates, and catalytic and regulatory mechanisms. J Biol Chem 266(21): 13469-72. Potter, A. S. and P. A. Newhouse (2004). Effects of acute nicotine administration on behavioral inhibition in adolescents with attention-deficit/hyperactivity disorder. Psychopharmacology (Berl) 176(2): 182-94. Pritchard, M. P., R. Ossetian, D. N. Li, C. J. Henderson, B. Burchell, C. R. Wolf, et al. (1997). A general strategy for the expression of recombinant human cytochrome P450s in Escherichia coli using bacterial signal peptides: expression of CYP3A4, CYP2A6, and CYP2E1. Arch Biochem Biophys 345(2): 342-54. Rae, J. M., M. D. Johnson, M. E. Lippman and D. A. Flockhart (2001). Rifampin is a selective, pleiotropic inducer of drug metabolism genes in human hepatocytes: studies with cDNA and oligonucleotide expression arrays. J Pharmacol Exp Ther 299(3): 849-57. Rahmioglu, N. and K. R. Ahmadi (2010). Classical twin design in modern pharmacogenomics studies. Pharmacogenomics 11(2): 215-26. Ramji, S. (1998). Bioactivation of doxorubicin by human NADPH-cytochrome P450 reductase. Department of Pharmacology. Toronto, University of Toronto. Masters of Science. Ramji, S., C. Lee, T. Inaba, A. V. Patterson and D. S. Riddick (2003). Human NADPH- cytochrome p450 reductase overexpression does not enhance the aerobic cytotoxicity of doxorubicin in human breast cancer cell lines. Cancer Res 63(20): 6914-9. Rao, Y., E. Hoffmann, M. Zia, L. Bodin, M. Zeman, E. M. Sellers, et al. (2000). Duplications and defects in the CYP2A6 gene: identification, genotyping, and in vivo effects on smoking. Mol Pharmacol 58(4): 747-55. Rauma, A. L., A. Rautio, M. Pasanen, O. Pelkonen, R. Torronen and H. Mykkanen (1996). Coumarin 7-hydroxylation in long-term adherents of a strict uncooked vegan diet. Eur J Clin Pharmacol 50(1-2): 133-7.

250 Redlich, G., U. M. Zanger, S. Riedmaier, N. Bache, A. B. Giessing, M. Eisenacher, et al. (2008). Distinction between human cytochrome P450 (CYP) isoforms and identification of new phosphorylation sites by mass spectrometry. J Proteome Res 7(11): 4678-88. Redon, R., S. Ishikawa, K. R. Fitch, L. Feuk, G. H. Perry, T. D. Andrews, et al. (2006). Global variation in copy number in the human genome. Nature 444(7118): 444-54. Reich, D. E., M. Cargill, S. Bolk, J. Ireland, P. C. Sabeti, D. J. Richter, et al. (2001). Linkage disequilibrium in the human genome. Nature 411(6834): 199-204. Reich, D. E., S. B. Gabriel and D. Altshuler (2003). Quality and completeness of SNP databases. Nat Genet 33(4): 457-8. Ring, H. Z., A. M. Valdes, D. M. Nishita, S. Prasad, P. Jacob, 3rd, R. F. Tyndale, et al. (2007). Gene-gene interactions between CYP2B6 and CYP2A6 in nicotine metabolism. Pharmacogenet Genomics 17(12): 1007-15. Roberts, D. C. and G. F. Koob (1982). Disruption of cocaine self-administration following 6- hydroxydopamine lesions of the ventral tegmental area in rats. Pharmacol Biochem Behav 17(5): 901-4. Roberts, R. L., S. E. Luty, R. T. Mulder, P. R. Joyce and M. A. Kennedy (2004). Association between cytochrome P450 2D6 genotype and harm avoidance. Am J Med Genet B Neuropsychiatr Genet 127B(1): 90-3. Robertson, P., H. H. DeCory, A. Madan and A. Parkinson (2000). In vitro inhibition and induction of human hepatic cytochrome P450 enzymes by . Drug Metab Dispos 28(6): 664-71. Rodrigues, A. D. (1994). Comparison of levels of aldehyde oxidase with cytochrome P450 activities in human liver in vitro. Biochem Pharmacol 48(1): 197-200. Rodriguez-Antona, C., M. T. Donato, E. Pareja, M. J. Gomez-Lechon and J. V. Castell (2001). Cytochrome P-450 mRNA expression in human liver and its relationship with enzyme activity. Arch Biochem Biophys 393(2): 308-15. Rodriguez-Antona, C., A. Gomez, M. Karlgren, S. C. Sim and M. Ingelman-Sundberg (2009). Molecular genetics and epigenetics of the cytochrome P450 gene family and its relevance for cancer risk and treatment. Hum Genet. Rosenberg, J., N. L. Benowitz, P. Jacob and K. M. Wilson (1980). Disposition kinetics and effects of intravenous nicotine. Clin Pharmacol Ther 28(4): 517-22. Rossini, A., T. de Almeida Simao, R. M. Albano and L. F. Pinto (2008). CYP2A6 polymorphisms and risk for tobacco-related cancers. Pharmacogenomics 9(11): 1737-52. Rotger, M., M. Saumoy, K. Zhang, M. Flepp, R. Sahli, L. Decosterd, et al. (2007a). Partial deletion of CYP2B6 owing to unequal crossover with CYP2B7. Pharmacogenet Genomics 17(10): 885-90. Rotger, M., H. Tegude, S. Colombo, M. Cavassini, H. Furrer, L. Decosterd, et al. (2007). Predictive value of known and novel alleles of CYP2B6 for efavirenz plasma concentrations in HIV-infected individuals. Clin Pharmacol Ther 81(4): 557-66. Rowe, J. W., R. Andres, J. D. Tobin, A. H. Norris and N. W. Shock (1976). The effect of age on creatinine clearance in men: a cross-sectional and longitudinal study. J Gerontol 31(2): 155-63. Rowold, D. J. and R. J. Herrera (2000). Alu elements and the human genome. Genetica 108(1): 57-72. Roy, P., L. J. Yu, C. L. Crespi and D. J. Waxman (1999). Development of a substrate-activity based approach to identify the major human liver P-450 catalysts of cyclophosphamide and ifosfamide activation based on cDNA-expressed activities and liver microsomal P- 450 profiles. Drug Metab Dispos 27(6): 655-66. Russell, M. A. (1980). Nicotine intake and its regulation. J Psychosom Res 24(5): 253-64. 251 Russell, M. A. and C. Feyerabend (1978). Cigarette smoking: a dependence on high-nicotine boli. Drug Metab Rev 8(1): 29-57. Russell, M. A., C. Wilson, C. Feyerabend and P. V. Cole (1976). Effect of nicotine chewing gum on smoking behaviour and as an aid to cigarette withdrawal. Br Med J 2(6032): 391-3. Russell, M. A., C. Wilson, U. A. Patel, C. Feyerabend and P. V. Cole (1975). Plasma nicotine levels after smoking cigarettes with high, medium, and low nicotine yields. Br Med J 2(5968): 414-6. Saito, S., A. Iida, A. Sekine, S. Kawauchi, S. Higuchi, C. Ogawa, et al. (2003). Catalog of 680 variations among eight cytochrome p450 ( CYP) genes, nine esterase genes, and two other genes in the Japanese population. J Hum Genet 48(5): 249-70. Salavaggione, O. E., L. Wang, M. Wiepert, V. C. Yee and R. M. Weinshilboum (2005). Thiopurine S-methyltransferase pharmacogenetics: variant allele functional and comparative genomics. Pharmacogenet Genomics 15(11): 801-15. Salisbury, B. A., M. Pungliya, J. Y. Choi, R. Jiang, X. J. Sun and J. C. Stephens (2003). SNP and haplotype variation in the human genome. Mutat Res 526(1-2): 53-61. Satarug, S., M. A. Lang, P. Yongvanit, P. Sithithaworn, E. Mairiang, P. Mairiang, et al. (1996). Induction of cytochrome P450 2A6 expression in humans by the carcinogenic parasite infection, opisthorchiasis viverrini. Cancer Epidemiol Biomarkers Prev 5(10): 795-800. Satarug, S., M. Nishijo, P. Ujjin, Y. Vanavanitkun, J. R. Baker and M. R. Moore (2004). Effects of chronic exposure to low-level cadmium on renal tubular function and CYP2A6- mediated coumarin metabolism in healthy human subjects. Toxicol Lett 148(3): 187-97. Satarug, S., P. Ujjin, Y. Vanavanitkun, M. Nishijo, J. R. Baker and M. R. Moore (2004a). Effects of cigarette smoking and exposure to cadmium and lead on phenotypic variability of hepatic CYP2A6 and renal function biomarkers in men. Toxicology 204(2- 3): 161-73. Schachter, S., L. T. Kozlo-ski and B. Silverstein (1977a). 2. Effects of urinary pH on cigarette smoking. J Exp Psychol Gen 106(1): 13-9. Schachter, S., B. Silverstein, L. T. Kozlowski, D. Perlick, C. P. Herman and B. Liebling (1977). Studies of the interaction of psychological and pharmacological determinants of smoking. J Exp Psychol Gen 106(1): 3-12. Schneider, N. G., P. Popek, M. E. Jarvik and E. R. Gritz (1977). The use of nicotine gum during cessation of smoking. Am J Psychiatry 134(4): 439-40. Schnoll, R. A., F. Patterson, E. P. Wileyto, R. F. Tyndale, N. Benowitz and C. Lerman (2009). Nicotine metabolic rate predicts successful smoking cessation with transdermal nicotine: a validation study. Pharmacol Biochem Behav 92(1): 6-11. Schoedel, K. A., E. B. Hoffmann, Y. Rao, E. M. Sellers and R. F. Tyndale (2004). Ethnic variation in CYP2A6 and association of genetically slow nicotine metabolism and smoking in adult Caucasians. Pharmacogenetics 14(9): 615-26. Schoedel, K. A., E. M. Sellers, R. Palmour and R. F. Tyndale (2003). Down-regulation of hepatic nicotine metabolism and a CYP2A6-like enzyme in African green monkeys after long-term nicotine administration. Mol Pharmacol 63(1): 96-104. Schumacher, A. and A. Petronis (2006). Epigenetics of complex diseases: from general theory to laboratory experiments. Curr Top Microbiol Immunol 310: 81-115. Sebat, J., B. Lakshmi, J. Troge, J. Alexander, J. Young, P. Lundin, et al. (2004). Large-scale copy number polymorphism in the human genome. Science 305(5683): 525-8. Sellers, E. M., H. L. Kaplan and R. F. Tyndale (2000). Inhibition of cytochrome P450 2A6 increases nicotine's oral bioavailability and decreases smoking. Clin Pharmacol Ther 68(1): 35-43. 252 Sellers, E. M., Y. Ramamoorthy, M. V. Zeman, M. V. Djordjevic and R. F. Tyndale (2003). The effect of methoxsalen on nicotine and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK) metabolism in vivo. Nicotine Tob Res 5(6): 891-9. Sellers, E. M., R. F. Tyndale and L. C. Fernandes (2003a). Decreasing smoking behaviour and risk through CYP2A6 inhibition. Drug Discov Today 8(11): 487-93. Shiffman, S., J. A. Johnston, M. Khayrallah, C. A. Elash, C. J. Gwaltney, J. A. Paty, et al. (2000). The effect of bupropion on nicotine craving and withdrawal. Psychopharmacology (Berl) 148(1): 33-40. Shimada, T., H. Yamazaki and F. P. Guengerich (1996). Ethnic-related differences in coumarin 7-hydroxylation activities catalyzed by cytochrome P4502A6 in liver microsomes of Japanese and Caucasian populations. Xenobiotica 26(4): 395-403. Shimada, T., H. Yamazaki, M. Mimura, Y. Inui and F. P. Guengerich (1994). Interindividual variations in human liver cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: studies with liver microsomes of 30 Japanese and 30 Caucasians. J Pharmacol Exp Ther 270(1): 414-23. Silagy, C., T. Lancaster, L. Stead, D. Mant and G. Fowler (2004). Nicotine replacement therapy for smoking cessation. Cochrane Database Syst Rev(3): CD000146. Sinues, B., A. Fanlo, E. Mayayo, C. Carcas, J. Vicente, I. Arenaz, et al. (2008). CYP2A6 activity in a healthy Spanish population: effect of age, sex, smoking, and oral contraceptives. Hum Exp Toxicol 27(5): 367-72. Sinues, B., A. Fanlo, J. Vicente, E. Mayayo-Sinues, E. Mayayo, J. I. Labarta, et al. (2008a). Growth hormone does not alter CYP2A6 activity in growth hormone-deficient children. Basic Clin Pharmacol Toxicol 102(1): 45-9. Siu, E. C. and R. F. Tyndale (2007). Non-nicotinic therapies for smoking cessation. Annu Rev Pharmacol Toxicol 47: 541-64. Siu, E. C. and R. F. Tyndale (2008). Selegiline is a mechanism-based inactivator of CYP2A6 inhibiting nicotine metabolism in humans and mice. J Pharmacol Exp Ther 324(3): 992- 9. Siu, E. C., D. B. Wildenauer and R. F. Tyndale (2006). Nicotine self-administration in mice is associated with rates of nicotine inactivation by CYP2A5. Psychopharmacology (Berl) 184(3-4): 401-8. Slemmer, J. E., B. R. Martin and M. I. Damaj (2000). Bupropion is a nicotinic antagonist. J Pharmacol Exp Ther 295(1): 321-7. Soloway, S. B. (1976). Naturally occurring insecticides. Environ Health Perspect 14: 109-17. Solus, J. F., B. J. Arietta, J. R. Harris, D. P. Sexton, J. Q. Steward, C. McMunn, et al. (2004). Genetic variation in eleven phase I drug metabolism genes in an ethnically diverse population. Pharmacogenomics 5(7): 895-931. Sotaniemi, E. A., A. Rautio, M. Backstrom, P. Arvela and O. Pelkonen (1995). CYP3A4 and CYP2A6 activities marked by the metabolism of lignocaine and coumarin in patients with liver and kidney diseases and epileptic patients. Br J Clin Pharmacol 39(1): 71-6. Speed, W. C., S. P. Kang, D. P. Tuck, L. N. Harris and K. K. Kidd (2009). Global variation in CYP2C8-CYP2C9 functional haplotypes. Pharmacogenomics J 9(4): 283-90. Stahl, S. M., J. F. Pradko, B. R. Haight, J. G. Modell, C. B. Rockett and S. Learned-Coughlin (2004). A Review of the Neuropharmacology of Bupropion, a Dual Norepinephrine and Dopamine Reuptake Inhibitor. Prim Care Companion J Clin Psychiatry 6(4): 159-166. Staszewski, S., J. Morales-Ramirez, K. T. Tashima, A. Rachlis, D. Skiest, J. Stanford, et al. (1999). Efavirenz plus zidovudine and lamivudine, efavirenz plus indinavir, and indinavir plus zidovudine and lamivudine in the treatment of HIV-1 infection in adults. Study 006 Team. N Engl J Med 341(25): 1865-73. 253 Stead, L. F., R. Perera, C. Bullen, D. Mant and T. Lancaster (2008). Nicotine replacement therapy for smoking cessation. Cochrane Database Syst Rev(1): CD000146. Steinberg, L. (2004). Risk taking in adolescence: what changes, and why? Ann N Y Acad Sci 1021: 51-8. Steppuhn, A., K. Gase, B. Krock, R. Halitschke and I. T. Baldwin (2004). Nicotine's defensive function in nature. PLoS Biol 2(8): E217. Stewart, J. J., H. J. Berkel, R. C. Parish, M. R. Simar, A. Syed, J. A. Bocchini, Jr., et al. (2001). Single-dose pharmacokinetics of bupropion in adolescents: effects of smoking status and gender. J Clin Pharmacol 41(7): 770-8. Stier, A. (1976). Lipid structure and drug metabolizing enzymes. Biochem Pharmacol 25(2): 109-13. Stolerman, I. P. and M. J. Jarvis (1995). The scientific case that nicotine is addictive. Psychopharmacology (Berl) 117(1): 2-10; discussion 14-20. Stranger, B. E., M. S. Forrest, M. Dunning, C. E. Ingle, C. Beazley, N. Thorne, et al. (2007). Relative impact of nucleotide and copy number variation on gene expression phenotypes. Science 315(5813): 848-53. Strasser, A. A., V. Malaiyandi, E. Hoffmann, R. F. Tyndale and C. Lerman (2007). An association of CYP2A6 genotype and smoking topography. Nicotine Tob Res 9(4): 511- 8. Stresser, D. M. and D. Kupfer (1999). Monospecific antipeptide antibody to cytochrome P-450 2B6. Drug Metab Dispos 27(4): 517-25. Strom, S. C., L. A. Pisarov, K. Dorko, M. T. Thompson, J. D. Schuetz and E. G. Schuetz (1996). Use of human hepatocytes to study P450 gene induction. Methods Enzymol 272: 388- 401. Su, T., Z. Bao, Q. Y. Zhang, T. J. Smith, J. Y. Hong and X. Ding (2000). Human cytochrome P450 CYP2A13: predominant expression in the respiratory tract and its high efficiency metabolic activation of a tobacco-specific carcinogen, 4-(methylnitrosamino)-1-(3- pyridyl)-1-butanone. Cancer Res 60(18): 5074-9. Su, T., J. J. Sheng, T. W. Lipinskas and X. Ding (1996). Expression of CYP2A genes in rodent and human nasal mucosa. Drug Metab Dispos 24(8): 884-90. Sueyoshi, T., T. Kawamoto, I. Zelko, P. Honkakoski and M. Negishi (1999). The repressed nuclear receptor CAR responds to phenobarbital in activating the human CYP2B6 gene. J Biol Chem 274(10): 6043-6. Sugihara, K., S. Kitamura, K. Tatsumi, T. Asahara and K. Dohi (1997). Differences in aldehyde oxidase activity in cytosolic preparations of human and monkey liver. Biochem Mol Biol Int 41(6): 1153-60. Swan, G. E., N. L. Benowitz, C. N. Lessov, P. Jacob, 3rd, R. F. Tyndale and K. Wilhelmsen (2005). Nicotine metabolism: the impact of CYP2A6 on estimates of additive genetic influence. Pharmacogenet Genomics 15(2): 115-25. Swan, G. E., C. N. Lessov-Schlaggar, A. W. Bergen, Y. He, R. F. Tyndale and N. L. Benowitz (2009). Genetic and environmental influences on the ratio of 3'hydroxycotinine to cotinine in plasma and urine. Pharmacogenet Genomics 19(5): 388-98. Taavitsainen, P., R. Juvonen and O. Pelkonen (2001). In vitro inhibition of cytochrome P450 enzymes in human liver microsomes by a potent CYP2A6 inhibitor, trans-2- phenylcyclopropylamine (tranylcypromine), and its nonamine analog, cyclopropylbenzene. Drug Metab Dispos 29(3): 217-22. Tan, W., G. F. Chen, D. Y. Xing, C. Y. Song, F. F. Kadlubar and D. X. Lin (2001). Frequency of CYP2A6 gene deletion and its relation to risk of lung and esophageal cancer in the Chinese population. Int J Cancer 95(2): 96-101. 254 Tateishi, T., H. Nakura, M. Asoh, M. Watanabe, M. Tanaka, T. Kumai, et al. (1997). A comparison of hepatic cytochrome P450 protein expression between infancy and postinfancy. Life Sci 61(26): 2567-74. Tayama, Y., K. Miyake, K. Sugihara, S. Kitamura, M. Kobayashi, S. Morita, et al. (2007). Developmental changes of aldehyde oxidase activity in young Japanese children. Clin Pharmacol Ther 81(4): 567-72. Telenti, A. and U. M. Zanger (2008). Pharmacogenetics of anti-HIV drugs. Annu Rev Pharmacol Toxicol 48: 227-56. Teneggi, V., S. T. Tiffany, L. Squassante, S. Milleri, L. Ziviani and A. Bye (2002). Smokers deprived of cigarettes for 72 h: effect of nicotine patches on craving and withdrawal. Psychopharmacology (Berl) 164(2): 177-87. Thorgeirsson, T. E., D. F. Gudbjartsson, I. Surakka, J. M. Vink, N. Amin, F. Geller, et al. (2010). Sequence variants at CHRNB3-CHRNA6 and CYP2A6 affect smoking behavior. Nat Genet Epub. Tiihonen, J., U. Pesonen, J. Kauhanen, M. Koulu, T. Hallikainen, L. Leskinen, et al. (2000). CYP2A6 genotype and smoking. Mol Psychiatry 5(4): 347-8. Timofeeva, M. N., S. Kropp, W. Sauter, L. Beckmann, A. Rosenberger, T. Illig, et al. (2009). CYP450 polymorphisms as risk factors for early-onset lung cancer: gender-specific differences. Carcinogenesis 30(7): 1161-9. Tokizane, T., H. Shiina, M. Igawa, H. Enokida, S. Urakami, T. Kawakami, et al. (2005). Cytochrome P450 1B1 is overexpressed and regulated by hypomethylation in prostate cancer. Clin Cancer Res 11(16): 5793-801. Tozlu, S., I. Girault, S. Vacher, J. Vendrell, C. Andrieu, F. Spyratos, et al. (2006). Identification of novel genes that co-cluster with estrogen receptor alpha in breast tumor biopsy specimens, using a large-scale real-time reverse transcription-PCR approach. Endocr Relat Cancer 13(4): 1109-20. Tsuchiya, K., H. Gatanaga, N. Tachikawa, K. Teruya, Y. Kikuchi, M. Yoshino, et al. (2004). Homozygous CYP2B6 *6 (Q172H and K262R) correlates with high plasma efavirenz concentrations in HIV-1 patients treated with standard efavirenz-containing regimens. Biochem Biophys Res Commun 319(4): 1322-6. Turner, D. M., A. K. Armitage, R. H. Briant and C. T. Dollery (1975). Metabolism of nicotine by the isolated perfused dog lung. Xenobiotica 5(9): 539-51. Turpeinen, M., H. Raunio and O. Pelkonen (2006). The functional role of CYP2B6 in human drug metabolism: substrates and inhibitors in vitro, in vivo and in silico. Curr Drug Metab 7(7): 705-14. Tyndale, R. F. and E. M. Sellers (2001). Variable CYP2A6-mediated nicotine metabolism alters smoking behavior and risk. Drug Metab Dispos 29(4 Pt 2): 548-52. U.S. Department of Health and HumanServices (1998). Tobacco Use Among U.S. Racial/Ethnic Minority Groups: African Americans, American Indians and Alaska Natives, Asian Americans and Pacific Islanders, and Hispanics: A Report of the Surgeon General. U.S. Department of Health and Human Services, Centres for Disease Control and Prevention, National Cancer for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta, Georgia. Retreived January 2010 from http://www.cdc.gov/tobacco/data_statistics/sgr/sgr_1998/index.htm. Ueda, A., H. K. Hamadeh, H. K. Webb, Y. Yamamoto, T. Sueyoshi, C. A. Afshari, et al. (2002). Diverse roles of the nuclear orphan receptor CAR in regulating hepatic genes in response to phenobarbital. Mol Pharmacol 61(1): 1-6. Ujjin, P., S. Satarug, Y. Vanavanitkun, S. Daigo, N. Ariyoshi, H. Yamazaki, et al. (2002). Variation in coumarin 7-hydroxylase activity associated with genetic polymorphism of 255 cytochrome P450 2A6 and the body status of iron stores in adult Thai males and females. Pharmacogenetics 12(3): 241-9. Urakawa, N., T. Nagata, K. Kudo, K. Kimura and T. Imamura (1994). Simultaneous determination of nicotine and cotinine in various human tissues using capillary chromatography/mass spectrometry. Int J Legal Med 106(5): 232-6. Usami, O., Y. Ashino, Y. Komaki, M. Tomaki, T. Irokawa, T. Tamada, et al. (2007). Efavirenz- induced neurological symptoms in rare homozygote CYP2B6 *2/*2 (C64T). Int J STD AIDS 18(8): 575-6. Venkatakrishnan, K., L. L. Von Moltke and D. J. Greenblatt (2001). Human drug metabolism and the cytochromes P450: application and relevance of in vitro models. J Clin Pharmacol 41(11): 1149-79. Vieregge, A., M. Sieberer, H. Jacobs, J. M. Hagenah and P. Vieregge (2001). Transdermal nicotine in PD: a randomized, double-blind, placebo-controlled study. Neurology 57(6): 1032-5. von Richter, O., M. Pitarque, C. Rodriguez-Antona, A. Testa, R. Mantovani, M. Oscarson, et al. (2004). Polymorphic NF-Y dependent regulation of human nicotine C-oxidase (CYP2A6). Pharmacogenetics 14(6): 369-79. von Weymarn, L. B., K. M. Brown and S. E. Murphy (2006). Inactivation of CYP2A6 and CYP2A13 during nicotine metabolism. J Pharmacol Exp Ther 316(1): 295-303. von Weymarn, L. B., Q. Y. Zhang, X. Ding and P. F. Hollenberg (2005). Effects of 8- methoxypsoralen on cytochrome P450 2A13. Carcinogenesis 26(3): 621-9. Waeber, B., M. D. Schaller, J. Nussberger, J. P. Bussien, K. G. Hofbauer and H. R. Brunner (1984). Skin blood flow and cigarette smoking: the role of vasopressin. Clin Exp Hypertens A 6(10-11): 2003-6. Wain, L. V., J. A. Armour and M. D. Tobin (2009). Genomic copy number variation, human health, and disease. Lancet 374(9686): 340-50. Walaas, I. and F. Fonnum (1980a). Biochemical evidence for gamma-aminobutyrate containing fibres from the nucleus accumbens to the substantia nigra and ventral tegmental area in the rat. Neuroscience 5(1): 63-72. Walaas, I. and F. Fonnum (1980b). Biochemical evidence for glutamate as a transmitter in hippocampal efferents to the basal forebrain and hypothalamus in the rat brain. Neuroscience 5(10): 1691-8. Wall, M. A., J. Johnson, P. Jacob and N. L. Benowitz (1988). Cotinine in the serum, saliva, and urine of nonsmokers, passive smokers, and active smokers. Am J Public Health 78(6): 699-701. Walsky, R. L., A. V. Astuccio and R. S. Obach (2006). Evaluation of 227 drugs for in vitro inhibition of cytochrome P450 2B6. J Clin Pharmacol 46(12): 1426-38. Wang, H., S. Faucette, T. Sueyoshi, R. Moore, S. Ferguson, M. Negishi, et al. (2003). A novel distal enhancer module regulated by pregnane X receptor/constitutive androstane receptor is essential for the maximal induction of CYP2B6 gene expression. J Biol Chem 278(16): 14146-52. Wang, H., S. R. Faucette, D. Gilbert, S. L. Jolley, T. Sueyoshi, M. Negishi, et al. (2003a). Glucocorticoid receptor enhancement of pregnane X receptor-mediated CYP2B6 regulation in primary human hepatocytes. Drug Metab Dispos 31(5): 620-30. Wang, H., W. Tan, B. Hao, X. Miao, G. Zhou, F. He, et al. (2003). Substantial reduction in risk of lung adenocarcinoma associated with genetic polymorphism in CYP2A13, the most active cytochrome P450 for the metabolic activation of tobacco-specific carcinogen NNK. Cancer Res 63(22): 8057-61.

256 Wang, H., W. Tan, B. Hao, X. Miao, G. Zhou, F. He, et al. (2003b). Substantial reduction in risk of lung adenocarcinoma associated with genetic polymorphism in CYP2A13, the most active cytochrome P450 for the metabolic activation of tobacco-specific carcinogen NNK. Cancer Res 63(22): 8057-61. Wang, J., M. Pitarque and M. Ingelman-Sundberg (2006). 3'-UTR polymorphism in the human CYP2A6 gene affects mRNA stability and enzyme expression. Biochem Biophys Res Commun 340(2): 491-7. Wang, J., A. Sonnerborg, A. Rane, F. Josephson, S. Lundgren, L. Stahle, et al. (2006a). Identification of a novel specific CYP2B6 allele in Africans causing impaired metabolism of the HIV drug efavirenz. Pharmacogenet Genomics 16(3): 191-8. Wang, L., T. V. Nguyen, R. W. McLaughlin, L. A. Sikkink, M. Ramirez-Alvarado and R. M. Weinshilboum (2005). Human thiopurine S-methyltransferase pharmacogenetics: variant allozyme misfolding and aggresome formation. Proc Natl Acad Sci U S A 102(26): 9394-9. Wang, L., W. Sullivan, D. Toft and R. Weinshilboum (2003c). Thiopurine S-methyltransferase pharmacogenetics: chaperone protein association and allozyme degradation. Pharmacogenetics 13(9): 555-64. Ward, B. A., J. C. Gorski, D. R. Jones, S. D. Hall, D. A. Flockhart and Z. Desta (2003). The cytochrome P450 2B6 (CYP2B6) is the main catalyst of efavirenz primary and secondary metabolism: implication for HIV/AIDS therapy and utility of efavirenz as a substrate marker of CYP2B6 catalytic activity. J Pharmacol Exp Ther 306(1): 287-300. Waxman, D. J. and C. O'Connor (2006). Growth hormone regulation of sex-dependent liver gene expression. Mol Endocrinol 20(11): 2613-29. Wei, P., J. Zhang, M. Egan-Hafley, S. Liang and D. D. Moore (2000). The nuclear receptor CAR mediates specific xenobiotic induction of drug metabolism. Nature 407(6806): 920-3. Weinshilboum, R. (2003). Inheritance and drug response. N Engl J Med 348(6): 529-37. Weinshilboum, R. and L. Wang (2004). Pharmacogenetics: inherited variation in amino acid sequence and altered protein quantity. Clin Pharmacol Ther 75(4): 253-8. Wen, X., J. S. Wang, K. T. Kivisto, P. J. Neuvonen and J. T. Backman (2001). In vitro evaluation of valproic acid as an inhibitor of human cytochrome P450 isoforms: preferential inhibition of cytochrome P450 2C9 (CYP2C9). Br J Clin Pharmacol 52(5): 547-53. Wen, X., J. S. Wang, P. J. Neuvonen and J. T. Backman (2002). Isoniazid is a mechanism-based inhibitor of cytochrome P450 1A2, 2A6, 2C19 and 3A4 isoforms in human liver microsomes. Eur J Clin Pharmacol 57(11): 799-804. White, H. K. and E. D. Levin (2004). Chronic transdermal nicotine patch treatment effects on cognitive performance in age-associated memory impairment. Psychopharmacology (Berl) 171(4): 465-71. Wilkins, B., H. Johansen, M. P. Beaudet and C. I. Neutel. (2000). "Oral contraceptive use." Statistics Canada, Catalogue 82-003, Health Reports, Spring 2000 Vol. 11, No.4, from http://www.statcan.gc.ca/studies-etudes/82-003/feature-caracteristique/5022583-eng.pdf. Wilkinson, G. R. and D. G. Shand (1975). Commentary: a physiological approach to hepatic drug clearance. Clin Pharmacol Ther 18(4): 377-90. World Health Organization (2008). WHO Report on the global tobacco epidemic, 2008. The MPOWER package. Wortham, M., M. Czerwinski, L. He, A. Parkinson and Y. J. Wan (2007). Expression of constitutive androstane receptor, hepatic nuclear factor 4 alpha, and P450 oxidoreductase genes determines interindividual variability in basal expression and activity of a broad 257 scope of xenobiotic metabolism genes in the human liver. Drug Metab Dispos 35(9): 1700-10. Wrighton, S. A. and J. C. Stevens (1992). The human hepatic cytochromes P450 involved in drug metabolism. Crit Rev Toxicol 22(1): 1-21. Wyen, C., H. Hendra, M. Vogel, C. Hoffmann, H. Knechten, N. H. Brockmeyer, et al. (2008). Impact of CYP2B6 983T>C polymorphism on non-nucleoside reverse transcriptase inhibitor plasma concentrations in HIV-infected patients. J Antimicrob Chemother 61(4): 914-8. Wynne, H. A., L. H. Cope, E. Mutch, M. D. Rawlins, K. W. Woodhouse and O. F. James (1989). The effect of age upon liver volume and apparent liver blood flow in healthy man. Hepatology 9(2): 297-301. Xia, X. Y., R. X. Peng, J. P. Yu, H. Wang and J. Wang (2002). In vitro metabolic characteristics of cytochrome P-450 2A6 in Chinese liver microsomes. Acta Pharmacol Sin 23(5): 471- 6. Xie, H. J., U. Yasar, S. Lundgren, L. Griskevicius, Y. Terelius, M. Hassan, et al. (2003). Role of polymorphic human CYP2B6 in cyclophosphamide bioactivation. Pharmacogenomics J 3(1): 53-61. Xu, C., Y. S. Rao, B. Xu, E. Hoffmann, J. Jones, E. M. Sellers, et al. (2002a). An in vivo pilot study characterizing the new CYP2A6*7, *8, and *10 alleles. Biochem Biophys Res Commun 290(1): 318-24. Xu, J., A. Mendrek, M. S. Cohen, J. Monterosso, S. Simon, M. Jarvik, et al. (2007). Effect of cigarette smoking on prefrontal cortical function in nondeprived smokers performing the Stroop Task. Neuropsychopharmacology 32(6): 1421-8. Xu, P., S. L. Huang, R. H. Zhu, X. M. Han and H. H. Zhou (2002). Phenotypic polymorphism of CYP2A6 activity in a Chinese population. Eur J Clin Pharmacol 58(5): 333-7. Yamanaka, H., M. Nakajima, T. Fukami, H. Sakai, A. Nakamura, M. Katoh, et al. (2005a). CYP2A6 AND CYP2B6 are involved in nornicotine formation from nicotine in humans: interindividual differences in these contributions. Drug Metab Dispos 33(12): 1811-8. Yamanaka, H., M. Nakajima, M. Katoh, A. Kanoh, O. Tamura, H. Ishibashi, et al. (2005). Trans-3'-hydroxycotinine O- and N-glucuronidations in human liver microsomes. Drug Metab Dispos 33(1): 23-30. Yamanaka, H., M. Nakajima, K. Nishimura, R. Yoshida, T. Fukami, M. Katoh, et al. (2004). Metabolic profile of nicotine in subjects whose CYP2A6 gene is deleted. Eur J Pharm Sci 22(5): 419-25. Yamano, S., P. T. Nhamburo, T. Aoyama, U. A. Meyer, T. Inaba, W. Kalow, et al. (1989). cDNA cloning and sequence and cDNA-directed expression of human P450 IIB1: identification of a normal and two variant cDNAs derived from the CYP2B locus on chromosome 19 and differential expression of the IIB mRNAs in human liver. Biochemistry 28(18): 7340-8. Yamano, S., J. Tatsuno and F. J. Gonzalez (1990). The CYP2A3 gene product catalyzes coumarin 7-hydroxylation in human liver microsomes. Biochemistry 29(5): 1322-9. Yamazaki, H., K. Inoue, M. Hashimoto and T. Shimada (1999). Roles of CYP2A6 and CYP2B6 in nicotine C-oxidation by human liver microsomes. Arch Toxicol 73(2): 65-70. Yamazaki, H., K. Inoue, C. G. Turvy, F. P. Guengerich and T. Shimada (1997). Effects of freezing, thawing, and storage of human liver samples on the microsomal contents and activities of cytochrome P450 enzymes. Drug Metab Dispos 25(2): 168-74. Yanagihara, Y., S. Kariya, M. Ohtani, K. Uchino, T. Aoyama, Y. Yamamura, et al. (2001). Involvement of CYP2B6 in n-demethylation of ketamine in human liver microsomes. Drug Metab Dispos 29(6): 887-90. 258 Yang, M., N. Kunugita, K. Kitagawa, S. H. Kang, B. Coles, F. F. Kadlubar, et al. (2001). Individual differences in urinary cotinine levels in Japanese smokers: relation to genetic polymorphism of drug-metabolizing enzymes. Cancer Epidemiol Biomarkers Prev 10(6): 589-93. Yano, J. K., M. H. Hsu, K. J. Griffin, C. D. Stout and E. F. Johnson (2005). Structures of human microsomal cytochrome P450 2A6 complexed with coumarin and methoxsalen. Nat Struct Mol Biol 12(9): 822-3. Yoshida, R., M. Nakajima, K. Nishimura, S. Tokudome, J. T. Kwon and T. Yokoi (2003). Effects of polymorphism in promoter region of human CYP2A6 gene (CYP2A6*9) on expression level of messenger ribonucleic acid and enzymatic activity in vivo and in vitro. Clin Pharmacol Ther 74(1): 69-76. Yoshida, R., M. Nakajima, Y. Watanabe, J. T. Kwon and T. Yokoi (2002). Genetic polymorphisms in human CYP2A6 gene causing impaired nicotine metabolism. Br J Clin Pharmacol 54(5): 511-7. Yun, C. H., T. Shimada and F. P. Guengerich (1991). Purification and characterization of human liver microsomal cytochrome P-450 2A6. Mol Pharmacol 40(5): 679-85. Zanger, U. M., K. Klein, T. Saussele, J. Blievernicht, M. H. Hofmann and M. Schwab (2007). Polymorphic CYP2B6: molecular mechanisms and emerging clinical significance. Pharmacogenomics 8(7): 743-59. Zevin, S. and N. L. Benowitz (1999). Drug interactions with tobacco smoking. An update. Clin Pharmacokinet 36(6): 425-38. Zevin, S., P. Jacob, 3rd and N. Benowitz (1997). Cotinine effects on nicotine metabolism. Clin Pharmacol Ther 61(6): 649-54. Zhang, W., T. Kilicarslan, R. F. Tyndale and E. M. Sellers (2001). Evaluation of methoxsalen, tranylcypromine, and tryptamine as specific and selective CYP2A6 inhibitors in vitro. Drug Metab Dispos 29(6): 897-902. Zhang, X., K. Amemo, S. Ameno, K. Iwahashi, H. Kinoshita, T. , et al. (2001). Lack of association between smoking and CYP2A6 gene polymorphisms in A Japanese population. Nihon Arukoru Yakubutsu Igakkai Zasshi 36(5): 486-90. Zhang, X., K. Ameno, S. Ameno, H. Kinoshita, T. Kubota, M. Kumihashi, et al. (2002). Effects of whole deletion of CYP2A6 on nicotine metabolism in humans. Drug Chem Toxicol 25(2): 203-13. Zhang, X., J. D'Agostino, H. Wu, Q. Y. Zhang, L. von Weymarn, S. E. Murphy, et al. (2007). CYP2A13: variable expression and role in human lung microsomal metabolic activation of the tobacco-specific carcinogen 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone. J Pharmacol Exp Ther 323(2): 570-8. Zhou, S. F., J. P. Liu and B. Chowbay (2009). Polymorphism of human cytochrome P450 enzymes and its clinical impact. Drug Metab Rev 41(2): 89-295. Zukunft, J., T. Lang, T. Richter, K. I. Hirsch-Ernst, A. K. Nussler, K. Klein, et al. (2005). A natural CYP2B6 TATA box polymorphism (-82T--> C) leading to enhanced transcription and relocation of the transcriptional start site. Mol Pharmacol 67(5): 1772- 82.

259 List of publications and abstracts

Research Articles

Al Koudsi N, Tyndale RF. Hepatic CYP2B6 is altered by genetic, physiologic, and environmental factors but plays little role in nicotine metabolism. Xenobiotica 2010 Jun;40(6):381-92.

Al Koudsi N, Hoffmann EB, Assadzadeh A, Tyndale RF. Hepatic CYP2A6 levels and nicotine metabolism: impact of genetic, physiological, environmental, and epigenetic factors. Eur J Clin Pharmacol. 2010 Mar;66(3):239-51.

Al Koudsi N, Ahluwalia JS, Lin SK, Sellers EM, Tyndale RF. A novel CYP2A6 allele (CYP2A6*35) resulting in an amino-acid substitution (Asn438Tyr) is associated with lower CYP2A6 activity in vivo. Pharmacogenomics J. 2009 Aug;9(4):274-82.

Breen DM, Chan KK, Dhaliwall JK, Ward MR, Al Koudsi N, Lam L, De Souza M, Ghanim H, Dandona P, Stewart DJ, Bendeck MP, Giacca A. Insulin increases reendothelialization and inhibits cell migration and neointimal growth after arterial injury. Arterioscler Thromb Vasc Biol. 2009 Jul;29(7):1060-6.

Ho MK, Mwenifumbo JC, Al Koudsi N, Okuyemi KS, Ahluwalia JS, Benowitz NL, Tyndale RF. Association of nicotine metabolite ratio and CYP2A6 genotype with smoking cessation treatment in African-American light smokers. Clin Pharmacol Ther. 2009 Jun;85(6):635-43.

Mwenifumbo JC, Al Koudsi N, Ho MK, Zhou Q, Hoffmann EB, Sellers EM, Tyndale RF. Novel and established CYP2A6 alleles impair in vivo nicotine metabolism in a population of Black African descent. Hum Mutat. 2008 May;29(5):679-88.

Audrain-McGovern J, Al Koudsi N, Rodriguez D, Wileyto EP, Shields PG, Tyndale RF. The role of CYP2A6 in the emergence of nicotine dependence in adolescents. Pediatrics. 2007 Jan;119(1):264-74.

Al Koudsi N, Mwenifumbo JC, Sellers EM, Benowitz NL, Swan GE, Tyndale RF. Characterization of the novel CYP2A6*21 allele using in vivo nicotine kinetics. Eur J Clin Pharmacol. 2006 Jun;62(6):481-4.

Review Article

Al Koudsi N, O’Loughlin J, Rodriguez D, Audrain-McGovern J, Tyndale RF. The genetic aspects of nicotine metabolism and their impact on adolescent nicotine dependence. J of Ped Biochem. In press.

Al Koudsi N, Tyndale RF. Genetic influences on smoking: a brief review. Ther Drug Monit. 2005 Dec;27(6):704-9.

260 Online Publications

Siu E.C., Al Koudsi N., Ho M.K., and Tyndale R.F., 2006. New Frontiers in the Treatment and Management of Smoking Cessation. Cyberounds Psychiatry/Neuroscience. (http://www.cyberounds.com/conf/psychiatryneuroscience/2006-08-11/index.html) Siu E.C., Al Koudsi N., Ho M.K., and Tyndale R.F., 2006 Smoking, quitting and genetics. The doctor will see you now. (http://www.thedoctorwillseeyounow.com/articles/behavior/smoking_14)

Conference Presentations

Al Koudsi N., Sellers E.M., and Tyndale R.F. Hepatic CYP2A6 levels and nicotine metabolism are influenced by CYP2A6 genetic variation and sex, but not age. Winner of 2nd best trainee oral presentation at the 3rd Annual Canadian Joint Therapeutics Congress, 2006, Toronto, Canada.

Al Koudsi N., Rodriguez D., Audrain-McGovern J, and Tyndale RF. Risk of acquiring nicotine dependence among adolescent Caucasians: Influence of CYP2A6 genetic variation. Annual meeting of Society for Research on Nicotine and Tobacco, Orlando 2006.

Tyndale R.F., O’Loughlin J., Audrain-McGovern J., Al Koudsi N. Smoking and genetics: Genetically variable nicotine metabolism is differentially associated with smoking depending on phenotypes. 8th Annual Conference of the Society for Research on Nicotine and Tobacco Europe (Turkey, 2006).

Abstracts

1. Al Koudsi N., Assadzadeh A., Hoffmann E., and Tyndale R.F. A Pilot Study Investigating the Role of Epigenetics in Regulating CYP2A6. Poster presentation at the 17th International Symposium on Microsomes and Drug Addiction. Saratoga Springs, NY, USA 2008 2. Al Koudsi N., Mwenifumbo J.C., Hoffmann E., Sellers E.M., and Tyndale R.F. A novel CYP2A6 variant 6458 A>T (Asp438Tyr) is associated with lower nicotine metabolism. Poster presentation at the annual meeting of the Society for Research on Nicotine and Tobacco, Portland 2008. 3. Mwenifumbo J.C., Al Koudsi N., Ho M.K., Zhou Q., Hoffmann E., Sellers E.M., and Tyndale R.F. Novel and established CYP2A6 alleles impair in vivo nicotine metabolism in a population of black African descent. Poster presentation at the annual meeting of the Society for Research on Nicotine and Tobacco, Portland 2008. 4. Al Koudsi N., Mwenifumbo J.C., Ho M.K, Hoffmann E., Sellers E.M., and Tyndale R.F. A novel CYP2A6 variant 594 G>C (Val110Leu) is associated with lower CYP2A6 activity. Poster presentation at the annual meeting of the Society for Research on Nicotine and Tobacco, Austin 2007. 5. Al Koudsi N., Sellers E.M., and Tyndale R.F. Genetic variation, age, and sex: Impacts on hepatic CYP2A6 levels and nicotine metabolizing activity. Winner of 2nd best poster presentation at Visions of Pharmacology (VIP) 2006,University of Toronto, Canada. 6. Al Koudsi N., Sellers E.M., and Tyndale R.F. Genetic variation, age, and sex: Impacts on hepatic CYP2A6 levels and nicotine metabolizing activity. Winner of 2nd best poster presentation at Visions of Pharmacology (VIP) 2006,University of Toronto, Canada. 7. Al Koudsi N., Shea S.H., Myers M.G., Wall T.L., and Tyndale R.F. Genetic variation in CYP2A6 and CYP2E1 among Jewish Caucasian college students: Association with alcohol 261 and nicotine dependence (Poster presentation at the 10th annual meeting of Society for Research on Nicotine and Tobacco, Prague 2005). 8. Dhaliwall J., Chan K.K., Al Koudsi N., Lam L., Madadi G., Bendeck M.P., and Giacca A. In vivo effect of insulin to decrease matrix metalloproteinase MMP-2 activity. American Diabetes Association 65th Scientific Sessions, 2005. 9. Chan K.K., Dhaliwall J., Al Koudsi N., Lam L., De Souza M., Bendeck M.P., and Giacca A. Insulin Inhibits Neointimal Growth by Decreasing Vascular Smooth Muscle Cell Migration In Vivo. American Diabetes Association 64th Scientific Sessions, 2004.

262 Appendices

Appendix A

Al Koudsi N, Mwenifumbo JC, Sellers EM, Benowitz NL, Swan GE, Tyndale RF. Characterization of the novel CYP2A6*21 allele using in vivo nicotine kinetics. Eur J Clin Pharmacol. 2006 Jun;62(6):481-4.

263 Eur J Clin Pharmacol (2006) 62: 481–484 DOI 10.1007/s00228-006-0113-3

SHORT COMMUNICATION

Nael Al Koudsi . Jill C. Mwenifumbo . Edward M. Sellers . Neal L. Benowitz . Gary E. Swan . Rachel F. Tyndale Characterization of the novel CYP2A6*21 allele using in vivo nicotine kinetics

Received: 9 December 2005 / Accepted: 8 February 2006 / Published online: 28 April 2006 # Springer-Verlag 2006

Abstract Objective: The impact of CYP2A6*21 (K476R) frequency of the CYP2A6*21 allele was found to be 2.3% in on in vivo nicotine metabolism and disposition was Caucasians (n=5/222 alleles, evaluated in one twin from investigated. Methods: A two-step allele-specific PCR each twin pair). In vivo pharmacokinetic parameters, such as assay was developed to detect the 6573A>G single nucle- nicotine clearance (1.32±0.37 vs. 1.18±0.20 L/min), frac- otide polymorphism (SNP) in CYP2A6*21. Nicotine tional clearance of nicotine to cotinine (1.02±0.36 vs. 0.99± metabolism phenotypes from a previously described intra- 0.23 L/min), nicotine half-life (111±37 vs. 116± 29 min), and venous labeled nicotine and cotinine infusion study [1] was the trans-3′-hydroxycotinine to cotinine ratio (1.92±1.0 vs. used to assess the impact of CYP2A6*21. Genomic DNA 1.55±0.58) indicated no substantial differences in nicotine samples from 222 (111 monozygotic and dizygotic twin metabolism between those without the variant (CYP2A6*1/ pairs) Caucasian subjects were genotyped for CYP2A6 *1, n=163) and those with the variant (CYP2A6*1/*21, n=9), alleles (CYP2A6*1X2, -*1B, -*2, -*4, -*7, -*9, -*10, -*12, respectively. Conclusions: CYP2A6*21 does not have a and -*21). The pharmacokinetic parameters were compared detectable impact on nicotine metabolism in vivo. Our data between individuals with no detected CYP2A6 variants suggest that CYP2A6*21 may not be important for future (CYP2A6*1/*1, n=163) and individuals heterozygous for the studies of nicotine metabolism and the resulting impacts on CYP2A6*21 allele (CYP2A6*1/*21, n=9). Results: The smoking behaviors.

Keywords CYP2A6 . Nicotine metabolism . Nael Al Koudsi and Jill C. Mwenifumbo contributed equally to . . this work. Polymorphism Pharmacogenetics Smoking

N. Al Koudsi . J. C. Mwenifumbo . . E. M. Sellers R. F. Tyndale Introduction Centre for Addiction & Mental Health, and Department of Pharmacology, University of Toronto, Cytochrome P450 2A6 (CYP2A6) is the human hepatic Toronto, ON, Canada enzyme involved in the metabolism of several pharmaceu- tical agents, the most notable being nicotine [2, 3]. Inter- N. L. Benowitz Division of Clinical Pharmacology and Experimental individual and inter-ethnic differences in the efficacy and Therapeutics, Medical Service, toxicity of CYP2A6 substrates can result from variability in San Francisco General Hospital Medical Center, the activity and levels of the enzyme. This variability is due and Departments of Medicine, in part to polymorphisms in the CYP2A6 gene. Psychiatry and Biopharmaceutical Sciences, University of California, In vivo, approximately 80% of absorbed nicotine is San Francisco, CA, USA inactivated to cotinine via C-oxidation, and CYP2A6 mediates 90% of this reaction [2–4]. Because tobacco- G. E. Swan dependent individuals smoke, at least in part, to maintain Center for Health Sciences, SRI International, constant levels of nicotine, variation in the rate and extent Menlo Park, CA, USA of nicotine’s metabolic inactivation by CYP2A6 can affect R. F. Tyndale (*) smoking behaviors such as the risk of being a smoker, the University of Toronto, number of cigarettes smoked, or the likelihood of cessation Rm 4326 Medical Sciences Building, 1 King’s College Circle, [5–8]. Toronto, M5S 1A8, Canada e-mail: [email protected] CYP2A6*21 is a predicted haplotype containing two Tel.: +1-416-9786374 single nucleotide polymorphisms (SNPs), 51G>A and Fax: +1-416-9786395 6573A>G. The 51G>A is a synonymous SNP (V17V), 482 while 6573A>G is a nonsynonymous SNP that results in PCR Buffer (10 mM Tris pH 8.8, 50 mM KCl), 0.1 mM of lysine 476 arginine (K476R) amino acid substitution [9]. each dNTP, 1.7 mM MgCl2, 0.25 μM of each primer, The CYP2A6*21 allele is of interest for several reasons: (1) 1.25 U of Taq polymerase (Fermentas, Life Sciences, the resulting amino acid substitution K476R is located in Burlington, Canada), 50 ng of genomic DNA, and H2O. An substrate recognition site 6 (SRS6) [10]; (2) the amino acid initial denaturation was performed at 95°C for 1 min change is in close proximity to the nonsynonymous SNPs followed by 30 cycles each consisting of denaturation at 6558T>C (I471T) and 6582G>T (G479V), which are 95°C for 15 s, annealing at 52°C for 20 s, and extension at found in the decrease-of-function and loss-of-function 72°C for 3 min (PTC-200 Peltier Thermal Cyclers). In the alleles CYP2A6*7 and CYP2A6*5, respectively [11, 12]; second PCR reaction (PCR-II) an allele specific region from (3) in vitro, the substitution of lysine to arginine (K476R) exon 9 to 3′-flanking region, was amplified utilizing either of results in decreased coumarin 7-hydroxylation [13]; (4) in two forward primers 2A6ex9WF 5′-TGACGTGTCCC our liver bank, one liver sample heterozygous for CCAA-3′, or 2A6ex9VF 5′-TGACGTGTCCCCCAG-3′ CYP2A6*1/*21 had approximately 90% lower CYP2A6 in combination with one reverse primer 2A6R4 5′- levels and 80% lower nicotine metabolism activity than GCTTTTTAAGAATCTGTCTAGAA-3′. The 25-μl reac- homozygous CYP2A6*1/*1 liver samples (n=18); (5) tion mixture consisted of: 1 x PCR Buffer (10 mM Tris CYP2A6*21 has been found at relatively high allele pH 8.8, 50 mM KCl), 0.1 mM of each dNTP, frequencies, up to 7.0% in Caucasians and 3.4% in Chinese 1.5 mM MgCl2, 0.25 μM of each primer, 0.3 U of Taq [9, 14]; (6) genetic variation in nicotine metabolism can polymerase (Fermentas, Life Sciences), 0.8-μl of undiluted alter smoking behaviors. Consequently, the charactization PCR-I template and H2O. Initial denaturation was performed of each new, relatively high frequency allele for its in vivo at 95°C for 1 min followed by 20 cycles each consisting of impact on nicotine metabolism is important [8]. denaturation at 95°C for 15 s, annealing at 56°C for 20 s, and extension at 72°C for 45 s. An aliquot (20-μl) of the PCR-II product (537 bp) was analyzed by electrophoresis Materials and methods using 1.2% agarose gel (OnBio, Richmond Hill, Canada) stained with ethidium bromide. The assay was developed An allele-specific PCR assay was developed to detect the using cosmids containing genomic clones of CYP2A6 as a 6573A>G SNP (K476R) based on the genomic DNA positive control and CYP2A7 and CYP2A13 as negative sequence (GenBank Accession number NG_000008.5). In controls [7]. The assay was verified by sub-cloning the the first PCR reaction (PCR-I), a region from intron 6 to the PCR-I product into a pCR 4 TOPO vector (TOPO TA 3′-flanking region of the CYP2A6 gene was specifically Cloning Kit for Sequencing, Invitrogen Canada Inc) and amplified (2213 bp) utilizing the forward and reverse the sequencing of the insert using M13For and M13Rev primers 2A6In6F1 5′-ATTTCCTGCTCTGAGACC-3′ and primers (The Centre for Applied Genomics, Toronto, 2A6R3 5′-GGAATAGGTGCTTTTTAAGAATC-3′, re- Canada). spectively. The 25-μl reaction mixture consisted of: 1×

Table 1 In vivo pharmacokinetic parameters suggest a negligible impact of CYP2A6*21 on CYP2A6 nicotine metabolizing activity Phenotypea (mean±S.D.) CYP2A6 genotype p-valueb Difference (95% CI) CYP2A6*1/*1 (n=163) CYP2A6*1/*21 (n=9)

CLnic-d2 (L/min) 1.32±0.37 1.18±0.20 0.13 0.14 (−0.10 to 0.38) c t1/2 nicotine (min) 111±37 116±29 0.27 −5(−30 to −20) c CLcot-d4 (L/min) 53.2±22.3 51.4±18.4 0.41 1.8 (−13.1 to 16.7) t1/2 cotinine (min) 1043±352 1026±232 0.45 17 (−216 to 250) c CLnic→cot (L/min) 1.02±0.36 0.99±0.23 0.49 0.03 (−0.21 to 0.27) f 0.77±0.13 0.83±0.08 0.07 −0.06 (−0.15 to 0.03) c,d 3HC/COT-d2 1.92±1.0 1.55±0.58 0.17 0.37 (−0.30 to 1.03) c,d 3HC/COT-d4 1.64±0.83 1.48±0.62 0.32 0.16 (−0.39 to 0.71) a CLnic-d2: Total plasma clearance of deuterium-labelled nicotine (nicotine-d2) t1/2 nicotine: Half-life of nicotine CLcot-d4: Total plasma clearance of deuterium-labelled cotinine (cotinine-d4) t1/2 cotinine: Half-life of cotinine CLnic→cot: Fractional clearance of nicotine to cotinine f: Fractional conversion of nicotine to cotinine 3HC/COT-d2: Ratio of urine trans-3′-hydroxycotinine to cotinine obtained from nicotine-d2 infusion 3HC/COT-d4: Ratio of urine trans-3′-hydroxycotinine to cotinine obtained from cotinine-d4 infusion bp-values are derived from one tailed unpaired t-tests. All measures between the two groups had equal variances. As none of the pharmacokinetic variables were significantly different between genotype groups, estimation of more conservative p-values to account for pairwise dependence of observations was not conducted cDue to non-normal distribution (determined by skewness) values were log normalized for statistical analysis dThree wildtype individuals were excluded from the 3HC/COT data analyses due to missing data 483 The in vivo impact of CYP2A6*21 on the disposition forms, for example mouse Cyp2a5, which metabolizes kinetics and metabolism of nicotine after intravenous nicotine and is the orthologue of human CYP2A6, has labeled nicotine dosing was assessed. This nicotine phar- arginine (R) at residue 476 [16]. Recently, the CYP2A6 macokinetic study has been previously described and was crystal structure was solved facilitating our understanding approved by the institutional review boards of SRI of the impact that this amino acid substitution (K476R) International (SRI; Menlo Park, California) and the may have [17]. The structure indicates that the wild type University of California, San Francisco [1]. Genomic lysine (K) 476 resides on the surface of the enzyme and DNA samples from 222 (111 monozygotic and dizygotic does not make direct contact with nicotine. However, twin pairs) Caucasian subjects were genotyped for CYP2A6 lysine (K) 476 may be involved in long-range interactions alleles (CYP2A6 *1X2,-*1B,-*2,-*4,-*7,-*9,-*10 and that are transmitted to residues that contact the substrate. It -*12) with demonstrated in vivo impact on nicotine is important, however, to note that the lysine 476 to metabolism [15] using previously described genotyping arginine substitution (K476R) is fairly conservative, as protocols [8]. Individuals (n=50) with any of the tested both amino acids have similar physical characteristics (e.g. CYP2A6 variants (other than CYP2A6*21) were not mass and volume) and have fully protonated, charged side included in these analyses in order to avoid confounding chains at pH 7.4 with amino groups. Thus, it appears that the interpretation of the impact of the CYP2A6*21 allele. this conservative substitution (K476R) in CYP2A6*21 has The pharmacokinetic parameters were compared between little effect on CYP2A6 nicotine metabolism. individuals with no detected CYP2A6 variants (CYP2A6*1/ In conclusion, CYP2A6*21 does not have a detectable *1, n=163) and individuals with CYP2A6*21 (CYP2A6*1/ impact on nicotine metabolism in vivo. However its impact *21, n=9) using the statistics package SPSS 13.0 (SPSS on other substrates may be worth assessing because Incorporated, Chicago, Illinois, USA). CYP2A6 variant alleles can affect substrates differentially (e.g. CYP2A6*7)[18]. Our in vivo data suggests that CYP2A6*21 may not be important for future studies of Results and discussion nicotine metabolism and the resulting impacts on smoking behaviors. Of the 222 participants, 163 had no detected variants (CYP2A6*1/*1), 50 had at least one tested variant, and 9 Acknowledgements The authors wish to thank Dr. L. Ashworth had no other variants except for the CYP2A6*21 allele (Human genome center, Liverpool) for generously providing us with Cosmid DNA clones 19296, 19019, and 27292 that contain (CYP2A6*1/*21). The use of twin pairs to assess the allele CYP2A6, CYP2A7, and CYP2A13, respectively. We sincerely frequency could lead to a biased estimate, therefore the first thank Dr. Jason Yano for his helpful comments on the position and twin of each twin pair was selected arbitrarily and used to potential impact of CYP2A6*21 using the CYP2A6 crystal structure. assess the CYP2A6*21 allele frequency (total of 111 We thank Drs. T. Inaba and E. Roberts for providing the human livers that provided data used as part of the rationale for this study. We are individual twins). The resulting CYP2A6*21 allele fre- grateful for the participation of the twins without whom this work quency was 2.3% (n=5/222 alleles) and the observed would not have been possible. CIHR, The Centre for Addiction & genotype distribution was in Hardy-Weinberg equilibrium Mental Health and US Public Health Service grants awarded by the as assessed by the chi-square test. National Institutes on Drug Abuse, and carried out in part at the General Clinical Research Center at San Francisco General Hospital CYP2A6*21 had no apparent impact on the clearance with support of the Division of Research Resources MOP-53248, and half-life of nicotine or cotinine (Table 1). In addition, DA11170, DA02277, DA12393 and NIH RR00083 supported this more specific measures of CYP2A6 activity such as the study. NA receives funding from CIHR-STPTR and OGS, JCM receives funding from CIHR-STPTR and SPICE, and RFT holds a fractional clearance of nicotine to cotinine (CLnic→cot) and the ratio of trans-3′-hydroxycotinine to cotinine (3HC/ Canadian Research Chair in Pharmacogenetics. All experiments conducted are in compliance with the current laws of U.S.A. and COT) were not significantly different (Table 1). Other Canada inclusive of ethics approval. variants (CYP2A6*9 and *12 heterozygotes) that result in a moderate loss (∼25%) of CYP2A6 nicotine metabolism activity have been shown to alter nicotine kinetics in this References dataset even in the heterozygous form [15] indicating that CYP2A6*21 has an impact smaller than this. These data 1. Swan GE, Benowitz NL, Jacob P 3rd, Lessov CN, Tyndale RF, suggest that alternative genetic variants may exist in our Wilhelmsen K, Krasnow RE, McElroy MR, Moore SE, one CYP2A6*1/*21 liver sample accounting for its lower Wambach M (2004) Pharmacogenetics of nicotine metabolism in twins: methods and procedures. Twin Res 7:435 nicotine metabolizing activity and protein levels; this is 2. Messina ES, Tyndale RF, Sellers EM (1997) A major role for currently under investigation. Alternatively, this specific CYP2A6 in nicotine C-oxidation by human liver microsomes. liver sample also had a low capacity to metabolize J Pharmacol Exp Ther 282:1608 (a CYP2D6 substrate) indicating other 3. Nakajima M, Yamamoto T, Nunoya K, Yokoi T, Nagashima K, Inoue K, Funae Y, Shimada N, Kamataki T, Kuroiwa Y (1996) factors such as disease and/or medications might have Role of human cytochrome P4502A6 in C-oxidation of contributed to this liver’s low CYP2A6 activity. nicotine. Drug Metab Dispos 24:1212 Although there was a solid rationale to suggest that the 4. Benowitz NL, Jacob P 3rd (1994) Metabolism of nicotine to K476R substitution in CYP2A6*21 would result in a cotinine studied by a dual stable isotope method. Clin significant reduction in activity, our in vivo data do not Pharmacol Ther 56:483 support this. Interestingly, other functional CYP2A iso- 484 5. Gu DF, Hinks LJ, Morton NE, Day IN (2000) The use of long 12. Oscarson M, McLellan RA, Gullsten H, Agundez JA, Benitez PCR to confirm three common alleles at the CYP2A6 locus and J, Rautio A, Raunio H, Pelkonen O, Ingelman-Sundberg M the relationship between genotype and smoking habit. Ann (1999) Identification and characterisation of novel polymor- Hum Genet 64:383 phisms in the CYP2A locus: implications for nicotine metab- 6. O’Loughlin J, Paradis G, Kim W, DiFranza J, Meshefedjian G, olism. FEBS Lett 460:321 McMillan-Davey E, Wong S, Hanley J, Tyndale RF (2004) 13. Kim D, Wu ZL, Guengerich FP (2005) Analysis of Coumarin Genetically decreased CYP2A6 and the risk of tobacco 7-Hydroxylation Activity of Cytochrome P450 2A6 using dependence: a prospective study of novice smokers. Tob Random Mutagenesis. J Biol Chem 280:40319 Control 13:422 14. Solus JF, Arietta BJ, Harris JR, Sexton DP, Steward JQ, 7. Rao Y, Hoffmann E, Zia M, Bodin L, Zeman M, Sellers EM, McMunn C, Ihrie P, Mehall JM, Edwards TL, Dawson EP Tyndale RF (2000) Duplications and defects in the CYP2A6 (2004) Genetic variation in eleven phase I drug metabolism gene: identification, genotyping, and in vivo effects on genes in an ethnically diverse population. Pharmacogenomics smoking. Mol Pharmacol 58:747 5:895 8. Schoedel KA, Hoffmann EB, Rao Y, Sellers EM, Tyndale RF 15. Swan GE, Benowitz NL, Lessov CN, Jacob P 3rd, Tyndale RF, (2004) Ethnic variation in CYP2A6 and association of Wilhelmsen K (2005) Nicotine metabolism: the impact of genetically slow nicotine metabolism and smoking in adult CYP2A6 on estimates of additive genetic influence. Caucasians. Pharmacogenetics 14:615 Pharmacogenet Genomics 15:115 9. Haberl M, Anwald B, Klein K, Weil R, Fuss C, Gepdiremen A, 16. Lewis DF, Dickins M, Lake BG, Eddershaw PJ, Tarbit MH, Zanger UM, Meyer UA, Wojnowski L (2005) Three haplotypes Goldfarb PS (1999) Molecular modelling of the human associated with CYP2A6 phenotypes in Caucasians. Pharmacogenet cytochrome P450 isoform CYP2A6 and investigations of Genomics 15:609 CYP2A substrate selectivity. Toxicology 133:1 10. Lewis DF, Lake BG, Dickins M, Goldfarb PS (2003) Homol- 17. Yano JK, Hsu MH, Griffin KJ, Stout CD, Johnson EF (2005) ogy modelling of CYP2A6 based on the CYP2C5 crystal- Structures of human microsomal cytochrome P450 2A6 lographic template: enzyme-substrate interactions and QSARs complexed with coumarin and methoxsalen. Nat Struct Mol for binding affinity and inhibition. Toxicol In Vitro 17:179 Biol 12:822 11. Mwenifumbo JC, Myers MG, Wall TL, Lin SK, Sellers EM, 18. Xu C, Rao YS, Xu B, Hoffmann E, Jones J, Sellers EM, Tyndale RF (2005) Ethnic variation in CYP2A6*7, CYP2A6*8 Tyndale RF (2002) An in vivo pilot study characterizing the and CYP2A6*10 as assessed with a novel haplotyping method. new CYP2A6*7, *8, and *10 alleles. Biochem Biophys Res Pharmacogenet Genomics 15:189 Commun 290:318

Appendix B

Audrain-McGovern J, Al Koudsi N, Rodriguez D, Wileyto EP, Shields PG, Tyndale RF. The role of CYP2A6 in the emergence of nicotine dependence in adolescents. Pediatrics. 2007 Jan;119(1):264-74.

268 The Role of CYP2A6 in the Emergence of Nicotine Dependence in Adolescents Janet Audrain-McGovern, Nael Al Koudsi, Daniel Rodriguez, E. Paul Wileyto, Peter G. Shields and Rachel F. Tyndale Pediatrics 2007;119;e264-e274; originally published online Nov 27, 2006; DOI: 10.1542/peds.2006-1583

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://www.pediatrics.org/cgi/content/full/119/1/e264

PEDIATRICS is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. PEDIATRICS is owned, published, and trademarked by the American Academy of Pediatrics, 141 Northwest Point Boulevard, Elk Grove Village, Illinois, 60007. Copyright © 2007 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 0031-4005. Online ISSN: 1098-4275.

Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 ARTICLE

The Role of CYP2A6 in the Emergence of Nicotine Dependence in Adolescents

Janet Audrain-McGovern, PhDa, Nael Al Koudsi, BScb, Daniel Rodriguez, PhDa, E. Paul Wileyto, PhDa, Peter G. Shields, MDc, Rachel F. Tyndale, PhDb aTobacco Use Research Center, Department of Psychiatry, University of Pennsylvania, Philadelphia, Pennsylvania; bCentre for Addiction and Mental Health, Department of Pharmacology, University of Toronto, Toronto, Ontario, Canada; cLombardi Comprehensive Cancer Center, Georgetown University Medical Center, Washington, DC

Financial Disclosure: Dr Tyndale holds shares in Nicogen Inc, a company focused on creating novel smoking-cessation treatments; no funding for this study was received from Nicogen, and no benefit to the company was obtained. The other authors have indicated they have no financial relationships relevant to this article to disclose.

ABSTRACT

OBJECTIVES. The objectives of our study were to evaluate whether genetic variation in nicotine metabolic inactivation accounted for the emergence of nicotine depen- dence from mid- to late adolescence and whether initial smoking experiences www.pediatrics.org/cgi/doi/10.1542/ peds.2006-1583 mediated this effect. doi:10.1542/peds.2006-1583 METHODS. Participants were 222 adolescents of European ancestry who participated Key Words in a longitudinal cohort study of the biobehavioral determinants of adolescent adolescent smoking, nicotine metabolism, CYP2A6 smoking. Survey data were collected annually from grade 9 to the end of grade 12. Abbreviations Self-report measures included nicotine dependence, smoking, age first smoked, ISE—initial smoking experience initial smoking experiences, peer and household member smoking, and alcohol SM—slower metabolizer of nicotine NM—normal metabolizer of nicotine and marijuana use. DNA collected via buccal swabs was assessed for CYP2A6 alleles mFTQ—modified Fagerstrom Tolerance that are common in white people and are demonstrated to decrease enzymatic Questionnaire YRBS—Youth Risk Behavior Survey function (CYP2A6*2,*4,*9,*12). LGM—latent growth-curve modeling CFI—comparative-fit index RESULTS. Latent growth-curve modeling indicated that normal metabolizers (indi- RMSEA—root-mean-square error of viduals with no detected CYP2A6 variants) progressed in nicotine dependence at a approximation faster rate and that these increases in nicotine dependence leveled off more slowly SRMR—standardized root-mean residual HW—Hardy-Weinberg compared with slower metabolizers (individuals with CYP2A6 variants). Initial Accepted for publication Aug 9, 2006 smoking experiences did not account for how CYP2A6 genetic variation impacts Address correspondence to Janet Audrain- nicotine dependence. McGovern, PhD, Department of Psychiatry, University of Pennsylvania, 3535 Market St, CONCLUSIONS. These findings may help to promote a better understanding of the Suite 4100, Philadelphia, PA 19104. E-mail: [email protected] biology of smoking behavior and the emergence of nicotine dependence in ado- PEDIATRICS (ISSN Numbers: Print, 0031-4005; lescents and inform future work aimed at understanding the complex interplay Online, 1098-4275). Copyright © 2007 by the between genetic, social, and psychological factors in adolescent smoking behavior. American Academy of Pediatrics

e264 AUDRAIN-MCGOVERN et al Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 DOLESCENTS DIFFER IN the initial responsivity to variation in nicotine metabolism played a role in the Aboth the rewarding and aversive effects of cigarette emergence of nicotine dependence from mid- to late smoking. Adolescents who become nicotine dependent adolescence. Specifically, we hypothesized that adoles- may be more responsive to the rewarding effects of cents with the wild-type CYP2A6 genotype (NMs) would smoking. Research indicates that pleasant emotional and progress in nicotine dependence faster than adolescents physiologic effects (eg, enjoyed it, felt high, dizzy versus with a CYP2A6 genetic variant (SMs). We further hy- coughing, feeling sick) of the initial smoking experiences pothesized that ISEs (pleasant and unpleasant initial ex- (ISEs) discriminated adolescents who continued to ex- periences) would mediate this effect. periment with cigarettes and those who did not.1–3 A recent study also showed that these initial smoking re- METHODS actions can predict the development of nicotine depen- dence.4 Participants and Procedures Individual differences in response to smoking and the Participants were 222 9th-grade high school students of emergence of nicotine dependence may be partially ex- European ancestry who were enrolled in 1 of 5 public plained by genetic factors. The heritability of nicotine high schools in Virginia. These adolescents participated dependence has been well documented.5–7 Genetic sus- in a longitudinal cohort study of biobehavioral determi- ceptibility to drug dependence is thought to reflect, in nants of adolescent smoking. Of these 222 adolescents, part, variability in drug metabolism.8 Thus, genes that 113 (51%) were male and 109 (49%) were female. are involved in the metabolic inactivation of nicotine, This sample is a subset of a larger cohort that was such as CYP2A6, might be important in understanding drawn from 2393 students identified through class ros- which adolescents progress in nicotine dependence. Ap- ters at the beginning of 9th grade and followed until the proximately 80% of nicotine consumed via cigarette end of 12th grade. Figure 1 provides a summary of the smoking is removed from the body via inactivation to sample derivation for the larger cohort study as well as cotinine; the CYP2A6 gene encodes a hepatic enzyme the subset of participants that comprised the present that mediates ϳ90% to 100% of this metabolism to study. Students were ineligible to participate if they had cotinine.9–12 The CYP2A6 gene is highly polymorphic, a special classroom placement (eg, were learning-dis- with many genetic variants identified to date. However, abled or English was their second language). On the only a small number of these variants have been char- basis of the cohort selection criteria, a total of 2120 acterized for their impact on enzymatic activity in vivo (89%) students were eligible to participate. Of the 2120 or their frequencies among different ethnic groups eligible students, 1533 (72%) parents provided a re- (www.imm.ki.SE/CYPalleles/cyp2a6.htm).12,13 Studies sponse. Of these 1533 students, 1151 (75%) parents have linked smoking rate and risk of Diagnostic and Sta- consented to their teen’s participation in the study, tistical Manual of Mental Disorders, Fourth Edition–defined yielding an overall consent rate of 54%. An analysis of nicotine dependence in adults with polymorphisms in differences between parents who consented and those the CYP2A6 gene.14,15 Slower metabolizers (SMs), those who did not consent to their teen’s participation in the with genetic variants predicting Յ50% of the activity of study revealed a race-by-education interaction. The in- normal metabolizers (NMs), smoke fewer cigarettes and teraction indicated that the likelihood of consent was are less likely to be current smokers.14–16 significantly greater for white parents with more than a There has been little research to evaluate the role of high school education than for those with a high school CYP2A6 in the etiology of adolescent nicotine depen- education or less (89% vs 77%).18 dence. Adolescents who metabolize nicotine faster com- Participation in the study also required student as- pared with those who metabolize nicotine slower might sent. Fifteen students declined participation. Another 13 experience more pleasurable effects of smoking and students failed to participate in the baseline administra- fewer aversive effects, which, in turn, may increase the tion because of absence. The final baseline sample size likelihood of subsequent smoking and nicotine depen- (year 2000) was 1123 of the 2120 eligible students. dence. One recent study assessed whether the CYP2A6 Approximately 65% of the adolescents enrolled were genotype predicted risk of nicotine dependence defined white (of European ancestry), and ϳ35% were non- by International Classification of Diseases, 10th Revision from white (black, Asian, Hispanic, or “other”). The rates of early to midadolescence. O’Loughlin et al17 found no participation at the 3 spring follow-ups in the 10th association between CYP2A6 and initial responses to (2001), 11th (2002), and 12th (2003) grades were smoking, and contrary to expectation, the risk of becom- ϳ96% (1081), 93% (1043), and 89% (1005), respec- ing nicotine dependent was almost 3 times higher tively. University institutional review board approval of among adolescents with at least 1 fully inactive CYP2A6 the study protocol was obtained. variant (SMs, Ͻ50% of the activity of NMs) than ado- To limit potential bias resulting from ethnic admix- lescents with the wild-type genotype (NMs).17 ture (ie, allelic frequencies differing as a result of race In this study we sought to evaluate whether genetic and not the phenotype or outcome under investigation),

PEDIATRICS Volume 119, Number 1, January 2007 e265 Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 FIGURE 1 Adolescent cohort study.

the analyses were limited to adolescents of European distributed the survey. The surveys comprised fre- ancestry (n ϭ 714). Of the 714 adolescents, 326 adoles- quently used, valid, and reliable measures of adolescent cents smoked at least 1 whole cigarette either before the smoking history, household and peer smoking, and al- baseline assessment (9th grade) or during the follow-up cohol and marijuana use. The surveys were completed in period (end of 12th grade). We only included those the classroom. The survey contained a front page with adolescents who smoked at least 1 cigarette, because the student’s name. The front page was removed when never-smokers would not have had the opportunity for the survey was given to the student. The completed the genetic predisposition involving genetically variable survey only contained an identification number. A nicotine metabolism to be expressed.14,19–21 Separating member of the research team read aloud a set of instruc- never-smokers from those who have smoked has been tions, emphasizing confidentiality to promote honest re- considered an important step in refining smoking phe- sponding, and encouraged questions if survey items notypes.22 If genetic variation in nicotine metabolic in- were not clear. Teachers or school administrators were activation accounts for the emergence of nicotine depen- not involved in the data collection (to promote honest dence, then biological exposure is necessary for the responding).23 Research supports the validity of self-re- genetic effects to be expressed. Never-smokers may dif- port measures of smoking behavior and substance use in fer in numerous ways from those who have been ex- adolescents, particularly in nontreatment contexts in posed to nicotine through smoking. Approximately 31 which confidentiality is emphasized.24,25 Although a spe- adolescents had missing data on at least 1 covariate, and cific reading level was not determined, as indicated 62 adolescents had insufficient DNA for genotyping. above, adolescents with a special classroom placement Eleven adolescents who had higher nicotine-depen- were ineligible to participate. The surveys took ϳ30 dence scores at baseline (scores greater than the median minutes to complete. of 2) were removed. The primary variables of interest Buccal cells were collected as described previously,26,27 were nicotine dependence, CYP2A6 genotype, and pleas- and DNA was extracted with standard phenol-chloro- ant and unpleasant initial smoking reactions. Age first form techniques. Genotyping was performed by using smoked, baseline smoking, alcohol use, marijuana use, previously described 2-step allele-specific polymerase peer and household member smoking, and gender chain reaction assays.14 The CYP2A6 alleles investigated served as controlling variables. The data presented lead to either a decrease (CYP2A6*9 and CYP2A6*12) or herein are based on 222 adolescents of European ances- loss (CYP2A6*2 and CYP2A6*4) of CYP2A6 function and try. occur at relatively high frequencies in white people. Survey data were collected on-site during a classroom Positive controls included heterozygote and homozygote common to all students. A member of the research team samples for each variant, and negative controls included e266 AUDRAIN-MCGOVERN et al Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 instead of DNA. Assays were previously validat- been shown to be predictive of a more regular smoking ed,14 and 20% were repeated indicating a negligible dis- habit and subsequent nicotine dependence in adoles- cordance rate. cents.4,36 Negative sensations associated with ISEs have been shown to protect against subsequent dependence.4 Measures Covariates Nicotine Dependence Nicotine dependence was measured with a modified Baseline Smoking version of the Fagerstrom Tolerance Questionnaire An ordered categorical variable was generated from re- (mFTQ) for adolescents.28,29 This 7-item measure has sponses to a series of standard epidemiologic questions been used frequently in studies of adolescent smok- regarding smoking.24,25,37,38 On the basis of participant ing.23,30–32 Because nicotine dependence is a continuous responses to these items, adolescents were categorized as variable, adolescents progressed in nicotine dependence a (1) never-smoker (never having smoked a cigarette, when they reached a score of 1 (low level of nicotine not even a puff), (2) puffer (not ever having smoked a dependence) on the mFTQ and could progress to a score whole cigarette), (3) experimenter (smoked at least 1 of 9 (high level of nicotine dependence). Nicotine de- whole cigarette but Ͻ100 cigarettes total in a lifetime), pendence was measured at every data-collection wave. or (4) current smoker (smoked on at least 1 of the past Nicotine dependence was conceived of as a process 30 days and Ͼ100 cigarettes in a lifetime).39,40 existing on a continuum and not a state whereby an adolescent was placed in a category reflecting a static end Age First Smoked product of regular smoking.33 Thus, our statistical model Adolescents were asked, “How old were you when you evaluated the rate at which an adolescent progressed to smoked your first whole cigarette?” The question was a score of 1 on the mFTQ and the rate at which the based on an item from the Youth Risk Behavior Survey mFTQ score increased to a score of 9 (acceleration) and (YRBS).37 decreased (deceleration) across time. Friends Smoking Genotype Groupings Adolescents were asked if their best friend smokes and Individuals were categorized initially into 3 main groups how many of their other 4 best male and 4 best female (normal, intermediate, and slowest metabolizers) ac- friends currently smoke, which yielded an estimate of cording to the impact of the CYP2A6 alleles on nicotine smoking among their 9 best friends.41,42 metabolism. NMs (100% activity) included adolescents with no detected CYP2A6 variants. Intermediate metabo- Household Member Smoking lizers (75% activity) included adolescents who had 1 Adolescents were asked if any member of their house- copy of either CYP2A6*9 or CYP2A6*12. The SMs (Յ50% hold smokes cigarettes, such as mother, father, and/or activity) included adolescents with 1 or 2 copies of the siblings. This variable was dichotomized because of non- inactive variants (CYP2A6*2 and CYP2A6*4) or 2 copies normality of the responses (0, no; 1, yes). of decreased activity variants CYP2A6*9 and/or CYP2A6*12. Because the average levels of nicotine de- Alcohol and Marijuana Use pendence for intermediate metabolizers fell between the Lifetime alcohol and marijuana use was assessed with average values for the NMs and the SMs across time, the items that asked, “During you life, on how many days intermediate and slowest metabolizers were combined have you had at least one drink (not just a sip) of into 1 group (SMs, Յ75% activity) for sample-size pur- alcohol?” and “During your life, how many times have poses. you used marijuana?”37 The response options were 0 (0 days or times), 1 (1 day or time), and 2 (Ͼ1 to Ն100 days Initial Smoking Experiences or times). ISEs were measured by the 7-item Early Smoking Expe- riences Scale.34 Pleasurable and unpleasurable sensations Statistical Analysis were rated on a 4-point scale (1, none; 4, intense), Statistical analysis used latent growth-curve modeling including rush or buzz, relaxation, nausea, and cough. (LGM).43 LGM is a multivariate method that models Nicotine-dependent individuals tend to have more repeated measures of an observed variable on latent pleasant effects associated with their initial exposure to variables (factors) representing baseline level and devel- smoking.34 Retrospective reports of pleasurable sensa- opmental trends (eg, linear, quadratic).43,44 The factors tions measured by this scale have been validated.35 Con- are random effects. Therefore, LGM permits the estima- vergent and discriminant validity has been demon- tion of developmental heterogeneity in initial status and strated for adolescent populations.30 Pleasurable ISEs the rate of change from baseline across time43 and the adapted from the Early Smoking Experiences Scale have regression of factors on select covariates. In this study,

PEDIATRICS Volume 119, Number 1, January 2007 e267 Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 there were 4 annual repeated measurements of nicotine TABLE 1 Proportions for Categorical Covariates dependence, spanning ages 14 to 18 years. This ap- % proach considered individual growth and, thus, did not Gender assume that all adolescents start at the same level of Female 49 nicotine dependence at baseline and progress in nicotine Male 51 dependence at the same rate. We used Mplus 4.1 (Mu- CYP2A6 the´n & Muthe´n, Los Angeles, CA) for all multivariate NMs 74 SMs 26 modeling. Mplus is a statistical software package for con- 9th-grade smoking ducting growth modeling from a latent variable frame- Current/frequent 3 work. Never/puffer/experimenter 97 The multivariate modeling used all available data, a Household members smoke missing-data strategy used when data are missing at No 69 Yes 31 random and capitalizes on the data that are available for 9th-grade marijuana use each wave for each participant. Mplus provides this op- Used more than once 17 tion for latent variable modeling with missing data with Used once 12 maximum-likelihood estimation of the mean, variance, Never used 71 and covariance parameters, when requested, using the 9th-grade alcohol use 45 Had Ͼ1 drink 53 expectation maximization algorithm. Those with miss- Had 1 drink 20 ing data did not differ from those without missing data Never had a drink 27 on the covariates and on the dependent variable of Age first smoked nicotine dependence (P Ͼ .05). We log-transformed nic- Ͼ13 y 39 otine dependence to correct for univariate nonnormal- 13 y 25 Յ12 y 36 ity. Model fit was evaluated with model ␹2, comparative- fit index (CFI), root-mean-square error of approxima- (␹2 ϭ 8.5, P ϭ .07; CFI ϭ 0.97; RMSEA ϭ 0.07 tion (RMSEA), and standardized root-mean residual 4(n ϭ 222) [95% confidence limits: 0, 0.14], P ϭ .24; SRMR ϭ .05), (SRMR). Suggested criteria for model fit are nonsignifi- although the lower RMSEA 95% confidence limit was 0 cant model ␹2, CFI Ͼ 0.95, RMSEA Ͻ 0.05 to 0.08, and and the upper limit was 0.14, indicating the possibility of SRMR Ͻ 0.08.46–48 An RMSEA value of 0 represents fit from perfect to less than adequate. The baseline level exact model fit.48 Mplus provides a 95% RMSEA confi- was significant (␩ ϭ 0.11; z ϭ 5.17; P Ͻ .0001). The dence interval, and for single-group models it provides a 0 linear trend was also significant (␩ ϭ 0.16; z ϭ 4.51; P P value for the probability that the RMSEA value is 0 Ͻ .0001), although the quadratic trend was not (P Ͼ Ͻ.05.46 .05). The variances for baseline level and the linear trend were both significant (P Ͻ .05). The quadratic trend was RESULTS fixed to 0 to eliminate a nonsignificant negative vari- Descriptive Statistics ance. Distributions for the categorical covariates appear in Ta- ble 1. Means and SDs and bivariate correlations for the 4 Full Model repeated measures of nicotine dependence (log-trans- The full LGM with covariates fit the data well with linear 2 formed) and covariates appear in Table 2. and quadratic trends (␹17(n ϭ 222) ϭ 18.24, P ϭ .374; CFI ϭ 1.00; RMSEA ϭ 0.02 [95% confidence limits: 0, 0.07], CYP2A6 Allele Frequencies and Genotype Groupings P ϭ .84; SRMR ϭ 0.03). In addition, the upper RMSEA The allele frequencies of CYP2A6*2 (5.3% [490 alleles]), 95% confidence limit decreased to an adequate level CYP2A6*4 (0.6% [478 alleles]), CYP2A6*9 (6.1% [494 (0.07). Figure 2 is the structural model including covari- alleles]), and CYP2A6*12 (1.9% [482 alleles]) were sim- ates and standardized path coefficients for the significant ilar to previously reported allele frequencies in an ado- paths. lescent and adult white populations.14,17 The CYP2A6 ge- notype distributions did not deviate significantly from The effect of CYP2A6 Genotype on Nicotine Dependence Hardy-Weinberg (HW) equilibrium. Of the 222 individ- Baseline Nicotine Dependence uals, 164 (74%) were NMs and 58 (26%) were SMs. Parameter estimates, SEs, and z values appear in Table 3. Parameter estimates reflect a change in the dependent Model Fit variable for a unit change in the predictor variable, and Measurement Model the z value indicates the likelihood that the change is The single-group measurement model, absent covari- significant. Four predictor variables had significant ef- ates, fit reasonably well with linear and quadratic trends fects on baseline nicotine dependence. Greater lifetime e268 AUDRAIN-MCGOVERN et al Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 TABLE 2 Bivariate Correlations for All Measured Variables in the Model 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 Pleasant ISE 1.00 2 Negative ISE Ϫ0.16 1.00 3 Log ND 9 0.25 0.03 1.00 4 Log ND 10 0.17 0.01 0.26 1.00 5 Log ND 11 0.40 Ϫ0.15 0.25 0.45 1.00 6 Log ND 12 0.17 Ϫ0.08 0.17 0.41 0.53 1.00 7 Female Ϫ0.11 0.03 0.01 Ϫ0.12 Ϫ0.06 Ϫ0.07 1.00 8 Household smoking Ϫ0.09 Ϫ0.01 0.06 0.11 0.08 0.17 0.02 1.00 9 9th-grade smoking 0.16 0.08 0.14 0.23 0.27 0.15 Ϫ0.11 0.13 1.00 10 Age first smoked Ϫ0.19 Ϫ0.04 Ϫ0.25 Ϫ0.18 Ϫ0.15 Ϫ0.14 0.13 Ϫ0.21 Ϫ0.20 1.00 11 9th-grade marijuana use 0.13 0.09 0.29 0.28 0.22 0.20 Ϫ0.15 0.08 Ϫ0.23 Ϫ0.28 1.00 12 9th-grade alcohol use 0.14 0.05 0.18 0.09 0.12 0.05 0.00 0.05 Ϫ0.02 Ϫ0.22 0.30 1.00 13 No. of friends smoking 0.10 0.03 0.36 0.27 0.22 0.13 Ϫ0.02 0.06 0.22 Ϫ0.16 0.29 0.14 1.00 14 CYP2A6 0.07 0.02 0.01 0.13 0.00 Ϫ0.06 0.05 0.00 Ϫ0.09 Ϫ0.03 Ϫ0.09 Ϫ0.06 0.09 1.00

Mean 6.09 7.21 0.11 0.24 0.41 0.49 0.49 0.31 0.03 1.02 0.46 1.26 1.75 0.74 SD 2.24 2.53 0.30 0.48 0.59 0.58 0.50 0.46 0.16 0.87 0.77 0.86 2.02 0.44 Variable scores: female (male, 0); household smoking (0, no household member smokes; 1, at least 1 household member smokes); 9th-grade smoking (1, never-smoker, puffer, or experimenter; 0, current or frequent smoker); age first smoked (0, Յ12 years old; 1, 13 years old; 2, Ͼ13 years old); lifetime marijuana use (2, used more than once; 1, used once; 0, never used); lifetime alcohol use (2, used more than once; 1, used once; 0 , never used); and CYP2A6 (0, SM/intermediate metabolizer; 1, NM). ND indicates nicotine dependence.

FIGURE 2 Latent growth-curve model of the role of CYP2A6 in the emergence of nicotine dependence in adolescents. Note that the repeated observed measure of nicotine dependence was log-transformed to correct for univariate nonnormality. The values represent standardized regression coefficients. Circles represent latent variables (factors), and rectangles represent observed (measured) variables. The arrow representing the factor loading (0) from the 2 trend factors to 9th-grade log nicotine dependence is omitted for simplicity. Only significant paths are shown; thus, the nonsignificant covariates (gender, household smoking, lifetime alcohol use, and negative ISEs) are not shown. a P Ͻ .06; b P Ͻ .05; c P Ͻ .01.

marijuana use at 9th grade was associated with higher baseline nicotine dependence (␤ ϭ .02; z ϭ 1.98; P ϭ nicotine dependence at baseline (␤ ϭ .06; z ϭ 2.16; P ϭ .048). Finally, the effect of age first smoking approached .031). The more friends one had in 9th grade that significance (␤ ϭϪ.04; z ϭϪ1.76; P ϭ .078), suggest- smoked, the higher the level of nicotine dependence ing that the younger an adolescent was at smoking at baseline (␤ ϭ .04; z ϭ 4.37; P Ͻ .0001). In addi- onset, the higher the baseline level of nicotine depen- tion, having a pleasant ISE was associated with higher dence.

PEDIATRICS Volume 119, Number 1, January 2007 e269 Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 (222 ؍ TABLE 3 Linear Regression Coefficients, SEs, and z-Test Statistics (N Predictors Dependent Latent and Measured Variables Baseline ND ND Linear Trend ND Quadratic Trend Pleasant ISE Negative ISE ␥ SE z ␥ SE z ␥ SE z ␥ SE z ␥ SE z Female 0.03 0.04 0.82 Ϫ0.08 0.07 Ϫ1.12 0.02 0.02 0.92 Household smoking 0.01 0.04 0.22 0.02 0.08 0.28 0.01 0.03 0.44 Smoking 0.00 0.12 Ϫ0.02 0.70 0.24 2.91a Ϫ0.21 0.08 Ϫ2.70b 1.42 0.95 1.50 0.71 1.12 0.63 Age first smoked Ϫ0.04 0.02 Ϫ1.70c 0.05 0.05 0.93 Ϫ0.01 0.02 Ϫ0.92 Ϫ0.38 0.32 Ϫ1.21 Ϫ0.14 0.38 Ϫ0.38 Marijuana use 0.06 0.03 2.16b 0.04 0.05 0.73 Ϫ0.01 0.02 Ϫ0.54 0.20 0.25 0.81 0.23 0.29 0.78 Alcohol use 0.01 0.02 0.51 0.01 0.05 0.16 Ϫ0.01 0.02 Ϫ0.41 0.45 0.29 1.57 0.18 0.35 0.53 Friends smoking 0.04 0.01 4.37a 0.00 0.02 0.10 0.00 0.01 Ϫ0.50 Pleasant ISE 0.02 0.01 2.00b 0.03 0.02 1.33 Ϫ0.01 0.01 Ϫ1.08 Negative ISE 0.01 0.01 0.58 Ϫ0.02 0.02 Ϫ1.03 0.00 0.01 0.58 CYP2A6 0.00 0.04 0.02 0.17 0.09 1.96b Ϫ0.07 0.03 Ϫ2.42b 0.25 0.50 0.51 0.09 0.60 0.14 ND indicates nicotine dependence. a P Ͻ .01. c P Ͻ .08. b P Ͻ .05.

Nicotine-Dependence Linear Trend off in their nicotine dependence at a slower rate than There was a significant effect for the CYP2A6 genotype SMs from 11th to 12th grade. ISEs, pleasant or unpleas- on linear trend (␤ ϭ .17; z ϭ 1.97; P ϭ .048), such that ant, did not explain how CYP2A6 genetic variation im- NMs had faster acceleration in nicotine dependence than pacts nicotine dependence. SMs. Current smoking had a significant and positive effect on linear trend (␤ ϭ .71; z ϭ 2.96; P ϭ .0031), Statistical Power to Detect Effects indicating that having smoked in the past month at To test the statistical power of these results, we ran a baseline (9th grade) resulted in an increased acceleration Monte Carlo analysis based on the results of the LGM. in nicotine dependence across the 4 waves. Monte Carlo analyses assess the power of a sample to detect specific effects on the basis of repeated samplings Nicotine-Dependence Quadratic Trend from a population with known parameters.45 In the There was a significant negative effect for CYP2A6 geno- present case, the population parameters were those re- type on the quadratic trend (␤ ϭϪ.07; z ϭϪ2.41; P ϭ sulting from our analysis, and the population size was N .016), such that NMs had slower deceleration in nicotine ϭ 222. For the effect of the CYP2A6 genotype on nicotine dependence after 10th grade than SMs. The effect of the dependence, the power was .60 for the linear trend, and quadratic trend materializes, independent of the linear .80 for the quadratic trend. trend, only after the second wave (see the factor loadings in Fig 2). Current smokers at grade 9 (smoking at least 1 cigarette in the past month) had slower deceleration in Analysis of Population Substructure nicotine dependence (␤ ϭϪ.21; z ϭϪ2.67; P ϭ .008) The sample was examined for evidence of population than adolescents who had not yet smoked or had not stratification by using the Structure clustering program smoked a cigarette in the past month. (University of Chicago, Chicago IL [http://pritch.bsd. uchicago.edu/software.html]), which uses genotypes Testing for Mediation Effects that may be out of HW equilibrium overall and attempts We tested whether pleasant or unpleasant ISEs mediated to identify subpopulations that are at HW equilibrium the relationship between the CYP2A6 genotype and nic- internally.49 On the basis of the hypothesis that the otine dependence. CYP2A6 did not have a significant sample population was not 1 population but 2 subpopu- effect on either pleasant or unpleasant ISEs, negating the lations, the program attempted to classify individuals as possibility of mediation. belonging to one population or the other by using class In summary, there was a significant increase (accel- probabilities. Data for the analysis were genotypes of 42 eration) in nicotine dependence from 9th to 11th grade randomly selected biallelic single-nucleotide polymor- (linear trend), which then was followed by a leveling off phisms (a list of single-nucleotide polymorphisms is (deceleration) of nicotine dependence scores from 11th available on request). Our 42 random single-nucleotide to 12th grade (quadratic trend). The term trend is equiv- polymorphisms were at HW equilibrium according to alent to slope or rate of growth across time. NMs in- the GENHW routine in Stata (Stata Corp, College Sta- creased in their nicotine dependence scores at a faster tion, TX). Structure results indicated a single population. rate than SMs from 9th to 11th grade. NMs also leveled The average probability of assignment to subpopulation e270 AUDRAIN-MCGOVERN et al Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 1 was .50, with the entire range of assignment probabil- otine dependence was not supported. There are several ities from .48 to .53. plausible reasons why a mediated effect was not found. Quite simply, these ISEs may not account for the rela- DISCUSSION tionship between CYP2A6 genetic variation and emer- In this study we sought to evaluate whether genetic gence of nicotine dependence, or the mediated relation- variation in nicotine metabolism played a role in the ship is more complex than modeled. It is also possible emergence of nicotine dependence from mid- to late that the context of initial use of cigarettes influences an adolescence. We hypothesized that NMs would progress adolescent’s reactions to the physiologic and emotional in nicotine dependence faster than SMs and that this reactions to smoking. Research indicates that others, effect would be mediated by ISEs. Consistent with our usually of the same gender who have smoked previ- hypotheses, NMs did show a faster rate of progression in ously, are present for 90% of the first opportunities to nicotine dependence (significant linear trend), and these smoke cigarettes.3 Peer presence may prompt adoles- increases in nicotine dependence leveled off more slowly cents to experiment further despite initial negative reac- compared with SMs (significant quadratic trend). Con- tions to cigarette smoking. In addition, if the ISE also trary to our hypothesis, ISEs did not account for how involved other substance use such as alcohol or mari- CYP2A6 genetic variation impacts nicotine dependence. juana, the likelihood of continued experimentation may The finding that adolescent NMs progress in nicotine have been influenced irrespective of the reactions to dependence at a faster rate than SMs can be discussed smoking.2,53 Friedman et al2 found that experimenters within the context of animal self-administration studies, who continued in their smoking did not experience the role of learning in the etiology of drug dependence, fewer unpleasant reactions to smoking. Thus, nonphar- and research on the relationship between adult smoking macologic and pharmacologic factors associated with the practices and CYP2A6 variation. A faster rate of acquisi- initial smoking episode may be important in explaining tion might be associated with stronger dependence on smoking progression and the emergence of nicotine de- nicotine. Animal research indicates that addiction-prone pendence.54 Although we controlled for peer smoking, rat strains have a faster rate of drug self-administration household smoking, and alcohol and marijuana use in acquisition than addiction-resistant rat strains.50 In addi- the present model, we did not measure the context (eg, tion, models of drug dependence assume that repetitive presence of others, use of another substance) of the ISE. drug use is a learned behavior, strengthened over time Finally, it is possible that recall of ISEs is compromised by and over repeated exposure to the drug (eg, number of current smoking status.55 Although we prospectively cigarettes).33 In the present study, among those adoles- captured the first episode in over half of the sample, the cents with even low levels of nicotine dependence, NMs model did show that pleasant ISEs were positively asso- smoked significantly more cigarettes than SMs at grade ciated with nicotine dependence at baseline. Thus, those 12 (73 vs 32 cigarettes per week; P ϭ .04). NMs inacti- adolescents who were smoking more regularly at base- vate nicotine faster and may smoke more to titrate nic- line retrospectively reported more pleasurable experi- otine to a preferred level.14,15 Thus, faster metabolism is ences. compensated for by smoking more cigarettes, which, in Our findings contrast with a previous report of the turn, is associated with more learning trials. Therefore, relationship between the CYP2A6 genotype and the odds NMs not only accelerated in nicotine dependence at a of becoming nicotine dependent from early to midado- faster rate, but the habit may be more ingrained because lescence.17 This may be related to different measures of they also smoked more cigarettes. This process may ac- nicotine dependence (mFTQ versus International Classifi- count for the path from smoking experimentation to a cation of Diseases, 10th Revision criteria), which may cap- nicotine-dependent smoking habit among NMs. ture differing aspects of nicotine dependence, particu- These findings might clarify why both SMs and NMs larly among adolescents who have low levels of can become nicotine dependent, yet the SMs represent a dependence and are smoking at low rates.32 It may also smaller portion of those adults who present for formal have been influenced by the age of the cohort partici- smoking-cessation treatment.14,16 SMs may be better able pants (14–18 vs 12–16 years) and the fact that we com- to quit successfully, resulting in shorter durations of bined those with reduced nicotine inactivation (less than smoking.14,51 In late adolescence, SMs level off in nico- ϳ75%) into 1 group versus 2 groups because of the tine dependence faster than NMs. This could explain small sample size of those with the slowest inactivation why SMs are half as likely to be smokers in adulthood, of nicotine (Յ50%). Although our studies had similar and if they do smoke, they smoke fewer cigarettes.14,15 numbers of adolescents with the slowest inactivation of Data also suggest that SMs are more successful than NMs nicotine (Յ50%), power analyses indicated that power at quitting when using the nicotine patch, likely because was insufficient to conduct analyses on this group sep- of their higher levels of plasma nicotine.52 arately. Nonetheless, similar trends were observed when The hypothesis that ISEs would mediate the relation- slow and intermediate metabolizers were grouped to- ship between CYP2A6 genotype and progression in nic- gether or analyzed separately. In this study, combining

PEDIATRICS Volume 119, Number 1, January 2007 e271 Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 the slowest and the intermediate groups together was percent of those parents who responded did provide possible because of the gene-dose effect; however, this consent, and the difference between those who provided was not observed for nicotine dependence in the previ- consent and those who declined was small.18 However, ous study.17 Consistent with our findings, O’Loughlin et some caution is warranted in generalizing the results of al17 did not find that ISEs mediated the effect between this study. Although the sample may not be representa- CYP2A6 and the odds of becoming nicotine dependent. tive of all adolescents in the United States, the sample is One other recent study of smoking in English youth nationally and locally representative on basic demo- found no significant impact of the CYP2A6 genotype on graphic characteristics, and the sample smoking rates are risk for being a current or ex-smoker relative to being a regionally and locally comparable to those found in na- never-smoker at 13 to 15 years of age and at 18 years of tional surveys.67–69 For example, data from our 2003 age.56 However, the interpretation of these data are un- survey indicated that 10% were daily smokers compared clear; one might argue that never-smokers are a poor with ϳ9% in the 2003 YRBS and ϳ15% in the 2003 comparison group because there is no chance for the Monitoring the Future Survey.70,71 In addition, 15% of impact of nicotine metabolism to affect risk in individu- the adolescents in our sample were current smokers als with no smoking and, therefore, no nicotine expo- compared with 13% in the 2003 YRBS survey. sure.14,19–21 Another potential limitation is that our measure of Consistent with our findings, O’Loughlin et al17 found nicotine dependence, the mFTQ, was adapted from an a trend for higher levels of smoking among dependent adult assessment of nicotine dependence.28 This tradi- NMs compared with SM groups. Similarly, another tional approach for the assessment of nicotine depen- study in white and black adolescents found that a sig- dence has limitations with respect to capturing the nificant relationship between the ratio of 3-hydroxyco- emergence of nicotine dependence in adolescents.72,73 tinine to cotinine, a validated measure of CYP2A6 activ- However, at present, the limited research on the acqui- ity,57 and levels of smoking indicating that SMs smoked sition of or the changes in nicotine dependence across fewer cigarettes per day.58 In contrast, Huang et al56 did time has not highlighted an epidemiologic instrument not find a significant effect of the CYP2A6 genotype on that adequately captures the process of nicotine depen- levels of smoking, although this was assessed in all dence.33,74 Finally, there were insufficient numbers of smokers rather than in those who were dependent. As adolescents in other racial groups to conduct analyses previously shown in adults and again here, the genotype stratified by race, and the sample size of our study sug- only alters smoking levels in those who are dependent gests that this investigation may be considered a pilot smokers,14 which was not assessed in the Huang et al56 study. study. Despite these potential limitations, our findings help As one of the first investigations of the impact of explain variability in adolescent nicotine dependence CYP2A6 genetic variation on the emergence of adoles- and may provide clues to why some carry a smoking cent nicotine dependence, our study has both strengths habit into adulthood and others do not. Future research and weaknesses. Strengths include the collection of DNA may include investigation of environmental factors that and behavioral data from a large sample of adolescents, modify the effect of CYP2A6 genetic variation on nicotine the use of a more refined longitudinal nicotine-depen- dependence. That is, it would be valuable to identify dence phenotype, the conception of nicotine depen- factors that either promote nicotine dependence in SMs dence as a continuum rather than a category, and the (vulnerability interaction) and protect against nicotine analysis of the potentially biasing effects of ethnic ad- dependence in NMs (buffering interaction).75 In addi- mixture as an alternative explanation for the study find- tion, future research is needed to better understand how ings.33,59,60 differences in nicotine metabolism influence other sys- Although not a limitation of the current study, it is tems involved in nicotine dependence (eg, nicotinic ace- important to note that we did not incorporate biomarker tylcholine receptors desensitization and upregulation). validation of smoking status. Although biochemical ver- This line of inquiry may inform youth smoking-preven- ification of smoking status is important in smoking-ces- tion and intervention efforts and reduce smoking-re- sation intervention studies, such measures are not typi- lated morbidity and mortality. cally implemented in epidemiologic studies, because adolescent self-reports have been determined to be valid ACKNOWLEDGMENTS and sufficient, especially when confidentiality is as- This study was supported by Transdisciplinary Tobacco sured.25,61–63 In addition, the standard cotinine cutoff of Use Research Center grants P50 84718 and NCI RO1 15 ng/mL cannot validate cotinine levels consistent with CA109250 from the National Cancer Institute and the definitions of being an adolescent current smoker (eg, 1 National Institute on Drug Abuse (to Dr Audrain- cigarette in the past 30 days).64–66 McGovern), Canadian Institutes of Health Research One potential limitation of this study is the parental grant MOP-53248 (to Dr Tyndale), the Centre for Ad- consent rate for adolescent participation. Seventy-five diction and Mental Health, Ontario Graduate Scholar- e272 AUDRAIN-MCGOVERN et al Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 ship Program, and Canadian Institutes of Health Re- the human dopamine transporter variable number of tandem search Strategic Training Program in Tobacco Research repeats (VNTR) polymorphism: failure to replicate and finding (Mr Koudsi), and a Canada Research Chair in Pharma- that never-smokers may be different. Nicotine Tob Res. 2002;4: 333–340 cogenetics (Dr Tyndale) 21. Kozlowski LT, Harford MR. On the significance of never using a drug: an example from cigarette smoking. J Abnorm Psychol. 1976;85:433–434 REFERENCES 22. Vandenbergh DJ, Kozlowski LT, Bennett CJ, et al. DAT’s not 1. Hirschman RS, Leventhal H, Glynn K. The development of all, but it may be more than we realize. Nicotine Tob Res. 2002;4: smoking behavior: conceptualization and supportive cross- 251–252 sectional survey data. J Appl Soc Psychol. 1984;14:184–206 23. Rojas N, Killen J, Haydel K, Robinson T. Nicotine dependence 2. Friedman LS, Lichtenstein E, Biglan A. Smoking onset among among adolescent smokers. Arch Pediatr Adolesc Med. 1998;152: teens: an empirical analysis of initial situations. Addict Behav. 151–156 1985;10:1–13 24. Patrick DL, Cheadle A, Thompson DC, et al. The validity of 3. Curry SJ, Peterson AV Jr, Mann SL. Investigation of first op- self-reported smoking: a review and meta-analysis. Am J Public portunities to use cigarettes and smokeless tobacco. Health Educ Health. 1994;84:1086–1093 Res. 1989;4:27–34 25. Botvin GJ, Botvin EM. Adolescent tobacco, alcohol, and drug 4. DiFranza JR, Savageau JA, Fletcher K, et al. Recollections and abuse: prevention strategies, empirical findings, and assess- repercussions of the first inhaled cigarette. Addict Behav. 2004; ment issues. J Dev Behav Pediatr. 1992;13:290–301 29:261–72 26. Lum A, Le Marchand L. A simple mouthwash method for 5. Sullivan P, Kendler K. The genetic epidemiology of smoking. obtaining genomic DNA in molecular epidemiological studies. Nicotine Tob Res. 1999;1(suppl 2):S51–S57; discussion S69–S70 Cancer Epidemiol Biomarkers Prev. 1998;7:719–724 6. Kendler KS, Neale MC, Sullivan P, Corey LA, Gardner CO, 27. Harty LC, Caporaso NE, Hayes RB, et al. Alcohol dehydroge- Prescott CA. A population-based twin study in women of nase 3 genotype and risk of oral cavity and pharyngeal cancers. smoking initiation and nicotine dependence. Psychol Med. 1999; J Natl Cancer Inst. 1997;89:1698–1705 29:299–308 28. Prokhorov A, Pallonen U, Fava J, Ding L, Niaura R. Measuring 7. Sullivan PF, Neale BM, van den Oord E, et al. Candidate genes nicotine dependence among high-risk adolescent smokers. for nicotine dependence via linkage, epistasis, and bioinfor- Addict Behav. 1996;21:117–127 matics. Am J Med Genet B Neuropsychiatr Genet. 2004;126:23–36 29. Prokhorov AV, Koehly LM, Pallonen U, Hudman KS. Adoles- 8. Crabbe JC. Genetic contributions to addiction. Annu Rev Psy- cent nicotine dependence measured by the modified Fager- chol. 2002;53:435–462 strom Tolerance Questionnaire at two time points. J Child 9. Benowitz NL, Jacob P 3rd. Metabolism of nicotine to cotinine Adolesc Subst Abuse. 1998;7:35–46 studied by a dual stable isotope method. Clin Pharmacol Ther. 30. Rodriguez D, Audrain-McGovern J. Construct validity analysis 1994;56:483–493 of the early smoking experience questionnaire for adolescents. 10. Messina ES, Tyndale RF, Sellers EM. A major role for CYP2A6 Addict Behav. 2004;29:1053–1057 in nicotine C-oxidation by human liver microsomes. J Pharma- 31. Wellman RJ, Difranza JR, Pbert L, et al. A comparison of the col Exp Ther. 1997;282:1608–1614 psychometric properties of the Hooked on Nicotine Checklist 11. Nakajima M, Yamamoto T, Nunoya K, et al. Characterization of and the modified Fagerstrom Tolerance Questionnaire. Addict CYP2A6 involved in 3Ј-hydroxylation of cotinine in human Behav. 2006;31:486–495 liver microsomes. J Pharmacol Exp Ther. 1996;277:1010–1015 32. Kandel D, Schaffran C, Griesler P, Samuolis J, Davies M, Ga- 12. Xu C, Rao YS, Xu B, et al. An in vivo pilot study characterizing lanti R. On the measurement of nicotine dependence in the new CYP2A6*7, *8, and *10 alleles. Biochem Biophys Res adolescence: comparisons of the mFTQ and a DSM-IV-based Commun. 2002;290:318–324 scale. J Pediatr Psychol. 2005;30:319–332 13. Malaiyandi V, Sellers EM, Tyndale RF. Implications of CYP2A6 33. Tiffany ST, Conklin CA, Shiffman S, Clayton RR. What can genetic variation for smoking behaviors and nicotine depen- dependence theories tell us about assessing the emergence of dence. Clin Pharmacol Ther. 2005;77:145–158 tobacco dependence? Addiction. 2004;99(suppl 1):78–86 14. Schoedel KA, Hoffmann EB, Rao Y, Sellers EM, Tyndale RF. 34. Pomerleau O, Pomerleau C, Namenek R. Early experiences Ethnic variation in CYP2A6 and association of genetically with tobacco among women smokers, ex-smokers, and never- slow nicotine metabolism and smoking in adult Caucasians. smokers. Addiction. 1998;93:595–599 Pharmacogenetics. 2004;14:615–626 35. Pomerleau OF, Pomerleau CS, Mehringer AM, Snedecor SM, 15. Rao Y, Hoffmann E, Zia M, et al. Duplications and defects in Cameron OG. Validation of retrospective reports of early ex- the CYP2A6 gene: identification, genotyping, and in vivo ef- periences with smoking. Addict Behav. 2005;30:607–611 fects on smoking. Mol Pharmacol. 2000;58:747–755 36. Chen X, Stacy A, Zheng H, et al. Sensations from initial expo- 16. Malaiyandi V, Lerman C, Benowitz NL, Jepson C, Patterson F, sure to nicotine predicting adolescent smoking in China: a Tyndale RF. Impact of CYP2A6 genotype on pretreatment potential measure of vulnerability to nicotine. Nicotine Tob Res. smoking behaviour and nicotine levels from and usage of 2003;5:455–463 nicotine replacement therapy. Mol Psychiatry. 2006;11: 37. Grunbaum JA, Kann L, Kinchen SA, et al. Youth risk behavior 400–409 surveillance: United States, 2003 [published corrections appear 17. O’Loughlin J, Paradis G, Kim W, et al. Genetically decreased in MMWR Morb Mortal Wkly Rep. 2004;53:536 and MMWR Morb CYP2A6 and the risk of tobacco dependence: a prospective Mortal Wkly Rep. 2005;54:608]. MMWR Surveill Summ. 2004;53: study of novice smokers. Tob Control. 2004;13:422–428 1–96 18. Audrain J, Tercyak KP, Goldman P, Bush A. Recruiting ado- 38. Brener ND, Kann L, McManus T, Kinchen SA, Sundberg EC, lescents into genetic studies of smoking behavior. Cancer Epi- Ross JG. Reliability of the 1999 youth risk behavior survey demiol Biomarkers Prev. 2002;11:249–252 questionnaire. J Adolesc Health. 2002;31:336–342 19. Kozlowski L. Rehabilitating a genetic perspective in the study 39. Audrain-McGovern J, Rodriguez D, Moss HB. Smoking pro- of tobacco and alcohol use. Br J Addict. 1991;86:517–520 gression and physical activity. Cancer Epidemiol Biomarkers Prev. 20. Vandenbergh D, Bennett C, Grant M, et al. Smoking status and 2003;12:1121–1129

PEDIATRICS Volume 119, Number 1, January 2007 e273 Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 40. Audrain-McGovern J, Rodriguez D, Tercyak KP, Cuevas J, 58. Moolchan ET, Franken FH, Jaszyna-Gasior M. Adolescent nic- Rodgers K, Patterson F. Identifying and characterizing adoles- otine metabolism: ethnoracial differences among dependent cent smoking trajectories. Cancer Epidemiol Biomarkers Prev. smokers. Ethn Dis. 2006;16:239–243 2004;13:2023–2034 59. Lerman C, Berrettini W. Elucidating the role of genetic factors 41. Distefan JM, Gilpin EA, Choi WS, Pierce JP. Parental influences in smoking behavior and nicotine dependence. Am J Med Genet predict adolescent smoking in the United States, 1989–1993. B Neuropsychiatr Genet. 2003;118:48–54 J Adolesc Health. 1998;22:466–474 60. Sullivan PF, Eaves LJ, Kendler KS, Neale MC. Genetic case- 42. Choi WS, Pierce JP, Gilpin EA, Farkas AJ, Berry CC. Which control association studies in neuropsychiatry. Arch Gen Psychi- adolescent experimenters progress to established smoking in atry. 2001;58:1015–1024 the United States. Am J Prev Med. 1997;13:385–391 61. Williams CL, Eng A, Botvin GJ, Wynder EL. Validation of 43. Muthe´n BO. Beyond SEM: general latent variable modeling. students’ self-reported cigarette smoking status with plasma Behaviormetrika. 2002;29:81–117 cotinine levels. Am J Public Health. 1979;69:1272–1274 44. Duncan TE, Duncan SC, Strycker LA, Li F, Alpert A. An Intro- 62. Velicer WF, Prochaska JO, Rossi JS, Snow MG. Assessing out- duction to Latent Variable Growth Curve Modeling: Concepts, Issues, come in smoking cessation studies. Psychol Bull. 1992;111: and Applications. Mahwah, NJ: Lawrence Erlbaum Associates; 23–41 1999 63. Wills TA, Cleary SD. The validity of self-reports of smoking: 45. Muthe´n BO. Mplus Technical Appendices. Los Angeles, CA: Mu- analyses by race/ethnicity in a school sample of urban adoles- the´n & Muthe´n; 1998–2004 cents. Am J Public Health. 1997;87:56–61 46. Muthe´n LK, Muthe´n BO. Mplus User’s Guide. Los Angeles, CA: 64. SRNT Subcommittee on Biochemical Verification. Biochemical Muthe´n & Muthe´n; 2001 verification of tobacco use and cessation. Nicotine Tob Res. 2002: 47. Loehlin JC. Latent Variable Models: An Introduction to Factor, Path, 4,149–159 and Structural Equation Analysis. Mahwah, NJ: Lawrence Erl- 65. Abroms L, Simons-Morton B, Haynie DL, Chen R. Psychosocial baum Associates; 2004 predictors of smoking trajectories during middle and high 48. Curran PJ, Bollen KA, Chen F, Paxton P, Kirby JB. Finite school. Addiction. 2005;100:852–861 sampling properties of the point estimates and confidence in- 66. Dolcini MM, Adler NE, Lee P, Bauman KE. An assessment of tervals of the RMSEA. Sociol Methods Res. 2003;32:208–252 the validity of adolescent self-reported smoking using three 49. Pritchard JK, Stephens M, Donnelly P. Inference of population biological indicators. Nicotine Tob Res. 2003;5:473–483 structure using multilocus genotype data. Genetics. 2000;155: 67. US Census Bureau. State and county quick facts: Fairfax 945–959 County, Virginia. Available at: http://quickfacts.census.gov/ 50. Ranaldi R, Bauco P, McCormick S, Cools AR, Wise RA. Equal qfd/states/51/51059.html. Accessed February 8, 2005 sensitivity to cocaine reward in addiction-prone and addiction- 68. US Census Bureau. State and county quick facts: Virginia. resistant rat genotypes. Behav Pharmacol. 2001;12:527–534 Available at: http://quickfacts.census.gov/qfd/states/ 51. Gu DF, Hinks LJ, Morton NE, Day IN. The use of long PCR to 51000.html. Accessed August 29, 2003 confirm three common alleles at the CYP2A6 locus and the 69. Developmental Research & Programs: Fairfax County, VA. relationship between genotype and smoking habit. Ann Hum Fairfax county youth survey report, Communities That Care. Genet. 2000;64:383–390 2001. Available at: www.fairfaxcounty.gov/demogrph/pdf/ 52. Lerman C, Tyndale RF, Patterson F, et al. Nicotine metabolite youth2001.pdf. Accessed March 3, 2002 ratio predicts the efficacy of transdermal nicotine for smoking 70. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE. cessation. Clin Pharmacol Ther. 2006;79:600–608 Monitoring the Future National Results on Adolescent Drug Use: 53. Dani JA, Harris RA. Nicotine addiction and comorbidity with Overview of Key Findings, 2003. National Institute of Drug Abuse, alcohol abuse and mental illness. Nat Neurosci. 2005;8: ed. Bethesda, MD: National Institute of Drug Abuse; 2004 1465–1470 71. Centers for Disease Control and Prevention. Surveillance sum- 54. Eissenberg T, Balster R. Initial tobacco use episodes in children maries. MMWR Morb Mortal Wkly Rep. 2004;53(SS-2):1–96 and adolescents: current knowledge, future direction. Drug 72. Colby SM, Tiffany ST, Shiffman S, Niaura RS. Measuring nic- Alcohol Depend. 2000;59(suppl 1):S41–S60 otine dependence among youth: a review of available ap- 55. Riedel BW, Blitstein JL, Robinson LA, Murray DM, Klesges RC. proaches and instruments. Drug Alcohol Depend. 2000;59(suppl The reliability and predictive value of adolescents’ reports of 1):S23–S39 initial reactions to smoking. Nicotine Tob Res. 2003;5:553–559 73. Colby SM, Tiffany ST, Shiffman S, Niaura RS. Are adolescent 56. Huang S, Cook DG, Hinks LJ, et al. CYP2A6, MAOA, DBH, smokers dependent on nicotine? A review of the evidence. DRD4, and 5HT2A genotypes, smoking behaviour and cotinine Drug Alcohol Depend. 2000;59(suppl 1):S83–S95 levels in 1518 UK adolescents. Pharmacogenet Genomics. 2005; 74. Glautier S. Measures and models of nicotine dependence: pos- 15:839–850 itive reinforcement. Addiction. 2004;99(suppl 1):30–50 57. Dempsey D, Tutka P, Jacob P 3rd, et al. Nicotine metabolite 75. Karp I, O’Loughlin J, Hanley J, Tyndale RF, Paradis G. Risk ratio as an index of cytochrome P450 2A6 metabolic activity. factors for tobacco dependence in adolescent smokers. Tob Clin Pharmacol Ther. 2004;76:64–72 Control. 2006;15:199–204

e274 AUDRAIN-MCGOVERN et al Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010 The Role of CYP2A6 in the Emergence of Nicotine Dependence in Adolescents Janet Audrain-McGovern, Nael Al Koudsi, Daniel Rodriguez, E. Paul Wileyto, Peter G. Shields and Rachel F. Tyndale Pediatrics 2007;119;e264-e274; originally published online Nov 27, 2006; DOI: 10.1542/peds.2006-1583 Updated Information including high-resolution figures, can be found at: & Services http://www.pediatrics.org/cgi/content/full/119/1/e264 References This article cites 65 articles, 19 of which you can access for free at: http://www.pediatrics.org/cgi/content/full/119/1/e264#BIBL Citations This article has been cited by 4 HighWire-hosted articles: http://www.pediatrics.org/cgi/content/full/119/1/e264#otherarticl es Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): Therapeutics & Toxicology http://www.pediatrics.org/cgi/collection/therapeutics_and_toxico logy Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.pediatrics.org/misc/Permissions.shtml Reprints Information about ordering reprints can be found online: http://www.pediatrics.org/misc/reprints.shtml

Downloaded from www.pediatrics.org. Provided by GERSTEIN SCIENCE INFO CTR on September 10, 2010