Molecular Psychiatry (2009) 14, 912–945 & 2009 Nature Publishing Group All rights reserved 1359-4184/09 $32.00 www.nature.com/mp FEATURE REVIEW Differential contribution of genetic variation in multiple brain nicotinic cholinergic receptors to dependence: recent progress and emerging open questions L Greenbaum and B Lerer Biological Psychiatry Laboratory, Department of Psychiatry, Hadassah – Hebrew University Medical Center, Jerusalem, Israel

Nicotine dependence (ND), a major public health challenge, is a complex, multifactorial behavior, in which both genetic and environmental factors have a role. Brain nicotinic (nAChR)-encoding are among the most prominent candidate genes studied in the context of ND, because of their biological relevance as binding sites for nicotine. Until recently, most research on the role of nAChRs in ND has focused on two of these genes (encoding the a4- and b2-subunits) and not much attention has been paid to the possible contribution of the other nine brain nAChR subunit genes (a2–a3, a5–a7, a9–a10, b3– b4) to the pathophysiology and genetics of ND. This situation has changed dramatically in the last 2 years during which intensive research had addressed the issue, mainly from the genetics perspective, and has shown the importance of the CHRNA5-CHRNA3-CHRNB4 and CHRNA6- CHRNB3 loci in ND-related phenotypes. In this review, we highlight recent findings regarding the contribution of non-a4/b2-subunit containing nAChRs to ND, based on several lines of evidence: (1) human genetics studies (including linkage analysis, candidate- association studies and whole-genome association studies) of several ND-related phenotypes; (2) differential pharmacological and biochemical properties of receptors containing these subunits; (3) evidence from genetically manipulated mice; and (4) the contribution of nAChR genes to ND-related personality traits and neurocognitive profiles. Combining neurobiological genetic and behavioral perspectives, we suggest that genetic susceptibility to ND is not linked to one or two specific nAChR subtype genes but to several. In particular, the a3, a5–6 and b3–4 nAChR subunit-encoding genes may play a much more pivotal role in the neurobiology and genetics of ND than was appreciated earlier. At the functional level, variants in these subunit genes (most likely regulatory) may have independent as well as interactive contributions to the ND phenotype spectrum. We address methodological challenges in the field, highlight open questions and suggest possible pathways for future research. Molecular Psychiatry (2009) 14, 912–945; doi:10.1038/mp.2009.59; published online 30 June 2009 Keywords: nicotinic cholinergic receptor genes; nicotine dependence; tobacco; cigarettes; smoking; addiction

Introduction severe consequences of smoking, an estimated 45.3 million adults in the United States were current Tobacco addiction is a major public health challenge. smokers in 2006.4 Cigarette smoking declined in the Cigarette smoking is the leading cause of preventable United States from 24.7% in 1997 to 20.8% in 2006, death in the United States (around 440 000 premature and is declining in most developed countries but deaths annually during 1997–2001), mostly because remains common in the developing world.4,5 Health- of cancer, cardiovascular diseases and respiratory related economic losses because of smoking were 1,2 diseases. Globally, it is assumed that tobacco- estimated to be $167 billion a year during the period related diseases currently kill 4.2 million people per 1997–2001.1 3 year worldwide. Despite increasing awareness of the Although cigarette smoke contains more than 4000 ingredients,6,7 the pharmacological effects of tobacco Correspondence: Professor B Lerer, Biological Psychiatry Labora- addiction are produced by nicotine.8 Nicotine reaches tory, Department of Psychiatry, Hadassah – Hebrew University the brain in 10–60 s, with an initial concentration of Medical Center, Ein Karem, Jerusalem 91120, Israel. 100–500 nM (depending on how the cigarette is E-mail: [email protected] Received 8 January 2009; revised 30 April 2009; accepted 27 May smoked) and binds to nicotinic acetylcholine recep- 2009; published online 30 June 2009 tors (nAChRs).9 Nicotine is extensively metabolized Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 913 (mainly by the liver), but its average elimination half- that the recent discoveries from several different time of 2 h, allows nicotine to accumulate in the body research disciplines are actually complementary and of the ongoing smoker for hours.2,10 As nicotine support each other. They highlight the involvement of reaches the brain so rapidly after a puff, cigarettes genetic variants in different nAChR subunits, as- are an ideal drug delivery system, enabling smokers to sembled into a variety of receptors subtypes, in titrate brain nicotine levels with each episode of different ND-related phenotypes and move the field smoking.8,9 away from a more limited focus on the widespread, Nicotine dependence (ND) is a complex, multi- high-affinity a4b2-containing receptor. factorial behavior with genetic and environmental In this paper, we review the evidence supporting underpinnings.11 The evidence for a significant role of the role of non-a4- and b2-nAChR subunits in genetic factors in ND and other smoking-related ND-related phenotypes. On the basis of recent phenotypes is substantial. Heritability estimates for population genetic and pharmacological studies and smoking in twin studies range from 46 to 84%.11 genetically manipulated animal models, it appears Sullivan and Kendler12 estimated the heritability of that non-a4 and b2-nAChRs are substantially in- smoking initiation (SI) to be 0.56, with the remaining volved in the pathophysiology of this complex trait, variance from shared environmental (0.24) and and variants within them are important risk or unique environmental effects (0.2). The data for protection factors for ND phenotypes. A focus on smoking persistence or ND are 0.67, 0.02 and 0.31, a4b2-subunit containing receptors cannot explain the respectively.12 Taking gender into account in a meta- multiple facets of human ND behavior such as analysis, Li et al.13 estimated the mean heritability of positive and negative reinforcement, withdrawal, smoking persistence to be 0.59 in male adults and tolerance, subjective response to nicotine and smok- 0.46 in female adults; shared environmental effects as ing cessation treatment failure. Some of these 0.08 in males and 0.28 in females; and unique many facets are probably mediated by additional environmental effects as 0.37 in males and 0.24 in nAChR subunits and receptor subtypes containing females.13 them; genetic variants that influence their Different approaches to understanding the genetic function could plausibly be implicated in different contribution to ND have been used by researchers ND phenotypes. worldwide. These include population genetic studies (linkage, candidate-gene and whole-genome associa- tion studies), pharmacological and molecular biology Neuronal nAChRs approaches, neuroimaging studies and behavioral tests performed on genetically manipulated mice. nAChR subunits Genes postulated to be related to smoking behavior Neuronal nAChRs are pentameric, ligand-gated ion phenotypes are implicated in nicotine metabolism channels composed of five membrane-spanning sub- (for example, CYP2A6 or CYP2D6) or encode ele- units arranged around a central pore that forms a ments of the dopaminergic system (such as the cation channel.18–20 In contrast to nicotinic neuro- transporter and the dopamine D2 receptor), muscular junction receptors, which are composed of serotonin receptors, nAChRs11,14 and other candidate a1-, b1-, g-, d- and e-subunits, neuronal nAChRs are genes involved in cell adhesion, signal transduction composed of a-orb-subunit only. They may be and transport.15 heteropentamers composed of combinations of a- and A large body of research has sought to link tobacco b-subunits in different ratios or homopentamers of a- addiction to one or more specific nAChR subunits.16 subunits. Each subunit is encoded by a single gene, This research has traditionally focused on a4b2- nine alpha (a2–a10) and 3 beta (b2–b4), which is subunit containing nAChRs, based on their high expressed in neuronal (central and peripheral ner- affinity for nicotine, widespread distribution in the vous system) and non-neuronal tissues. The a8- brain and the encouraging success of varenicline, a subunit identified in the avian brain has not been partial agonist of this receptor, as a new smoking identified in mammals.21,22 cessation therapy.17 Less attention has been paid to The 11 nAChR subunit-encoding genes are located the role of other nAChR receptor subunits in ND- on 1, 4, 8, 11, 15 and 20, as shown in related phenotypes (See review by Rose16). Recently, Table 1. CHRNA5, CHRNA3 and CHRNB4 are grouped whole-genome association studies using high- in a gene cluster on 15q24. Human throughput techniques have found a plethora of CHRNA5 and CHRNA3 are positioned in a tail-to-tail genetic variants located in or near nAChR subunit- configuration on opposite DNA strands, and share encoding genes to be associated with ND-related some of their 30 UTR.23 CHRNB3 and CHRNA6 phenotypes. In fact, variants in almost all the 11 are also grouped in a gene cluster, located on 8p11. nAChR subunit-encoding genes have been reported to The clustered arrangement of CHRNA3/CHRNA5/ be associated with some aspect of smoking behavior, CHRNB4 and CHRNB3/CHRNA6 could be related to at different levels of statistical significance. At first control of the expression of these genes. Interestingly, glance, this flood of genetic findings may seem the human a7-nAChR gene was found to be partially difficult to integrate with current concepts of ND duplicated, with both loci mapping to the chromo- neurobiology and pharmacology; however, we believe some 15q13 region.24

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 914 Table 1 Human nicotinic acetylcholine receptor genes: chromosomal location, 50 and 30 gene position and expression profile relevant to ND

Gene Gene cluster Chromosome 50 Position 30 Position Expressiona

CHRNA2 NA 8p21 27392730 27373195 Yes CHRNA4 NA 20q13.2–q13.3 61463192 61445109 Yes CHRNA7 NA 15q14 30110018 30248541 Yes CHRNA9 NA 4p14 40032226 40051730 No CHRNA10 NA 11p15.5 3643393 3649190 No CHRNB2 NA 1q21.3 152806881 152818978 Yes CHRNA3 CHRA3-A5-B4 15q24 76700377 76674706 Yes CHRNA5 15q24 76644961 76673515 Yes CHRNB4 15q24 76720642 76703691 Yes CHRNB3 CHRNB3-A6 8p11.2 42671719 42711366 Yes CHRNA6 8p11.2 42742776 42726920 Yes

Abbreviations: NA, not applicable; ND, nicotine dependence. aIn brain region relevant to ND.

Pharmacological properties of neuronal nAChRs and subunits may contribute to receptor targeting and their involvement in human disorders localization in the neuronal membrane.30–32 Two main classes of neuronal nAChRs are known: Functionally, the nAChR can exist in three a-bungarotoxin (aBgtx)-sensitive receptors are com- conformational states, which are dynamically regu- posed of a7–a10-subunits and can form homomeric lated by agonist exposure: closed, open and desensi- (such as a7) or heteromeric receptors. aBgtx-insensi- tized.33 After nicotine binds to the closed, non- tive receptors, are heteromeric receptors, consist of conducting receptor, it becomes activated a2-a6 and b2–b4-subunits and bind nicotine and (open state). Following this, the receptor channel other agonists but not aBgtx (such as a4b2* and a3b4* conformation becomes closed, and the receptor is receptors, the * indicating that other, as yet unidenti- desensitized. Recovery from desensitization takes fied subunits may also be incorporated in the receptor place only after the nicotine molecule dissociates structure).22,25,26 The different subunit compositions from the receptor and the activation-desensitization and stoichiometry of each receptor type is responsible cycle can then restart. Importantly, the receptors are for its unique electrophysiological and pharmacolo- highly sensitive to modulation by many molecules, gical properties. Binding affinity for nicotine differs including divalent cations, steroids, kinases between receptors, ranging from ‘low’ to ‘high and more.19,22,34 affinity.’ The predominant nAChR in the human The brain nAChRs change considerably during the brain, a4b2*, has high affinity for nicotine, but development and aging in all species22,25 and are differences in the a4/b2-subunits ratio can lead to involved in the pathophysiology of several central subtypes with different pharmacological and func- nervous system (CNS) diseases other than nicotine tional properties.19,26 The brain nAChRs are activated addiction. Mutations in the channel region of a4- or by endogenous (acetylcholine) and exogenous ligands b2-subunits have been described in families suffering (nicotine). from autosomal dominant frontal lobe epilepsy,22 and Homopentameric receptors have five identical rare microdeletions of the 15q13.3 segment, which acetylcholine-binding sites in each molecule, includes the CHRNA7 gene, were reported recently in whereas the heteropentameric receptors have only idiopathic generalized epilepsy.35 A decrease in the two. Two a-subunits (a2–a4ora6) carry the principal density of nAChRs has been reported in brain component of the acetylcholine-binding site, and two disorders such as schizophrenia (in the hippocampus, non-a-subunits (b2orb4) carry the binding site thalamus and some cortical areas)36 and Alzheimer’s complementary component. The fifth subunit does disease (mainly in cortical areas and thalamus),37 not participate in acetylcholine binding (a5, b3 but whereas an increase has been reported in the also b2orb4).25,27 Each subunit protein has four cerebellar cortex and thalamus in autism.38 Decreased transmembrane segments, a long extracellular N- striatal a6b3-receptors have been reported in Parkin- terminal domain, an intracellular loop between son’s disease.39 a7-nAChRs may have a role in transmembranes 3 and 4 and a short C-terminal cognitive deficits in schizophrenia; in a small trial domain.28 a5- and b3-subunits do not play a direct the a7-nicotinic agonist, anabaseine, appeared to have role in receptor–ligand binding, and function only as positive effects on neurocognition in patients with auxiliary subunits when expressed with other a- and schizophrenia.40 The use of nicotinic agonists or b-subunits, forming neuronal nAChRs with stoichio- antagonists in these and other CNS disorders is under 27,29,30 metries such as (a4b2)2a5or(a4b2)2b3. These intense research.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 915 The neurobiology of ND Other studies suggest that additional brain regions The neurobiology of dependence on nicotine and and pathways are important in ND, including the other drugs of abuse is related to the brain limbic thalamus, insula, habenula and cortical areas. Some system.41 A shared property of many addictive drugs, of these areas contain a high density of nAChR.16,45 including nicotine, is that they increase dopamine (DA) concentration in the nucleus accumbens Localization pattern of neuronal nAChRs genes (NAc).42 The mesolimbic DA pathway, which in- Nicotinic acetylcholine receptors are expressed in both cludes dopaminergic neurons in the midbrain ventral the central (CNS) and peripheral (PNS) nervous tegmental area (VTA) and their target areas in the systems.34.nAChRs are extensively expressed through- limbic forebrain, particularly the NAc, is the focus of out the CNS (cortex, hippocampus, striatum, amygda- much research in the field of addiction neurobiol- la, thalamus, VTA, substantia nigra, cerebellum, ogy.41 It seems that the VTA–NAc pathway is critical medial habenula, locus coeruleus, olfactory bulb, in drug-induced rewards.2 interpeduncular nucleus (IPN) and pineal gland),18 Modulation of DA transmission in the VTA–NAc allowing nicotine to influence cellular events in many pathway by nicotine is complex, and mediated by brain regions. The location of nAChRs may be nAChRs expressed on several neuronal sub-popula- presynaptic, postsynaptic or extrasynaptic but most tions. According to a simplified model, nicotine attention has been given to their presynaptic role.46 activates VTA DA neurons and increases their firing Located presynaptically at cholinergic or non-choli- rate directly by stimulation of nAChRs expressed on nergic terminals, nAChRs modulate several neuro- them and indirectly by stimulation of other nAChR transmission pathways, such as the cholinergic, receptors on glutamatergic nerve terminals arising from dopaminergic, GABAergic, serotonergic pathways and the prefrontal cortex, as well as from the pedunculo- others.21 As presynaptic receptors, nAChRs modulate pontine tegmentum and the laterodorsal tegmentum.41 exocytosis and neurotransmitter release either by These glutamatergic terminals, mainly express a7- activation of voltage-operated calcium channels fol- containing receptors (a7*), innervate midbrain DA lowing membrane depolarization, or directly because neurons, and activate them. As a result, a similar state of calcium influx through the receptor pore, altering to long-term potentiation is induced in the excitatory the intracellular calcium concentration.19 glutamatergic input to the VTA DA neurons.9,42 The a4b2* receptors (the predominant receptor On the other hand, inhibitory GABAergic inter- subtype in the brain of humans and other mammalian neurons located in the VTA decrease DA neuron species) and homomeric a7-receptors are widely activity. They are excited by cholinergic inputs from expressed within the CNS,47 while the expression other parts of the brain.33,42 A solid background of pattern of other subunits is much more restricted to experimental literature emphasizes the role of the particular brain regions.18,25,20 a3b4* is the major mesolimbic DA system in nicotine addiction, includ- receptor in the autonomic nervous system and adrenal ing attenuation of the rewarding benefit of nicotine in medulla.26 However, it is important to note that the animal models, after blocking DA release in the NAc same neuronal population may express multiple with antagonists or lesions.9 subtypes of nAChRs.31 a4b2* receptors may co- The nAChRs on DA VTA neurons (as well as on assemble with additional units, to form more complex GABAergic interneurons) are mainly composed of receptors such as a4b2a3, a4b2a6ora4b2a5.18,48 a7 high-affinity a4- and b2-subunits, in combination receptors are also widely expressed in the brain, with other nAChR subunits (a4b2* receptors).43,44 including VTA, the prefrontal and motor cortex, basal The DA VTA and GABAergic interneuron a4b2* ganglia, hippocampus, cerebellum, and lateral and receptors desensitize after few minutes of acute medial geniculate bodies but with lower expression in nicotine administration (to be discussed later). As a the thalamus.49 There is some evidence that a7- consequence, GABAergic inhibitory input onto DA subunits can co-assemble with b-subunits in hetero- receptors is decreased.42 However, in the context of logous systems, but no biochemical evidence of such chronic nicotine administration and addiction, great receptors has been found in neurons.26,50 In contrast to importance is attached to other receptor types with a4b2* and a7-receptors, other neuronal nAChRs relatively low affinity for nicotine, such as the subunits show a much more restricted localization presynaptic a7*, expressed on the glutamatergic pattern and are concentrated in specific brain areas.18 projection to the VTA. These receptors do not strongly In relation to ND and the focus of this review, the fact desensitize at nicotine levels achieved by smokers that adult midbrain neurons express a variety of and glutamatergic afferent excitation onto DA neu- nAChR subunit mRNAs, including a3–a7, b2andb3, rons is enhanced. In parallel, while nicotine level is of particular importance. These results have been elevates during prolonged smoking, GABAergic neu- confirmed (with relatively small discrepancies) by ron a4b2* receptors are desensitized and GABAergic several independent studies of nAChR subunit mRNA inhibition of dopaminergic neurons in the VTA is expression in the midbrain DA and non-DA neu- effectively depressed.9,16 In addition to these mechan- rons.43,44,51,52 Owing to the biological importance of isms, the balance of VTA DA excitation/inhibition is this area in drug dependence, these receptor subunits fine-tuned by other neurotransmitters, such as ser- are plausible candidates for a functional role in otonin or norepinephrine.42 mediating nicotine effects on addiction.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 916 In the next paragraphs, we will concentrate on the restricted expression pattern. The b4-containing CNS expression patterns of some specific nAChRs nAChRs are expressed at relatively low densities in genes. It is important to note that most gene expres- the rat and mouse substantia nigra and VTA.43,70 sion data have been obtained from rodents or Indeed, besides the olfactory bulb and pineal gland, monkeys. Because of species differences in regional the b4-subunit is expressed in the CNS at high levels expression, care should be taken when extrapolating only in the medial habenula and IPN43,71,72 such results to humans.18,34 CHRNA9 and CHRNA10. The expression of a9- and CHRNA2. A study of CHRNA2 expression in rats a10-subunits in inner hair cells and their role in and mice detected a2-mRNA expression in the cochlear innervation in rats and mice are well midbrain regions (including VTA and IPN) as well documented.73 These receptors are also expressed in as in several cortical layers.53 In primate cortex, a2- dorsal root ganglions neurons and pituitary,74 but containing receptors were highly expressed in according to current knowledge, they are not found in monkeys and humans, making this subunit other regions of the brain.18 expression pattern in the brain species specific.18,48,54

CHRNA3. The a3-subunit is highly expressed in the Genetics of ND-related phenotypes 55 PNS mainly in the a3b4* receptor. In addition, a3- In the following section, we provide an overview containing nAChRs are found in several regions of of recent findings on the association of nAChR rodent brain, including the VTA, pineal gland, medial gene variants with ND-related phenotypes (Table 2), habenula, IPN, brainstem nuclei, hippocampus and 56–58 highlighting issues to consider in the context of cerebellum. genetic research. Three research methods have been used: linkage analysis, candidate-gene association CHRNA5. In humans, mRNA encoding the a5- studies and genome-wide association studies nicotinic subunit has been detected in several (GWAS). structures of the nervous system (including cortex) According to a recent review by Li,3 more than 20 but appears to be mainly expressed in the cerebellum genome-wide linkage analysis studies for various ND- and thalamus (where it assembles with a4b2* and related behaviors have been conducted in different a3b2* receptor subtypes) and the autonomic populations. Thirteen regions, located on chromo- ganglia.59 In particular, a4b2a5-containing receptors somes 3–7, 9–11, 17, 20 and 22 (two different regions are expressed on dopaminergic neurons in rat and on chromosomes 5 and chromosome 9) were found to mouse striatum and are involved in DA release due to show suggestive or significant results in at least two nicotine stimulation.60,61 In addition, a5-containing independent samples, with regions on chromosomes receptors are expressed on GABAergic neurons in the 9, 10, 11 and 17 receiving the strongest support.3 striatum and VTA. Three of the nAChRs are located in these suggestive or CHRNA6. In mouse and rat CNS, a6- and b3- significant regions: CHRNA 2(8p21–22), CHRNA4 subunits colocalize in dopaminergic neurons and (20q13.2–q13.3) and CHRNA10 (11p15). retina62,63 as a6b2b3* or a4a6b2b3 nAChRs.64–66 a6 The second approach, candidate-gene studies, has nAChRs are expressed in higher concentration on been used for several years to study the nAChR gene nerve terminals than on cell bodies or dendrites, association with ND phenotypes, primarily using suggesting that a6 or another subunit (such as b3) single-nucleotide polymorphism (SNPs). The studies plays a targeting role.32 The a6-subunit is found at have been performed in family-based (for example: Feng et al.75) or case–control designs (for example: high levels in catecholaminergic nuclei, including the 76 locus coeruleus and superior colliculus and VTA. In Saccone et al. ). In general, many different genes of rat and mouse striatum, a6b2* and a4(non-a6)b2* are various biological functions have been associated the major nAChR subtype populations, differently with ND-related phenotypes, but prominent associa- expressed by dopaminergic and non-dopaminergic tion was found with several neuronal nAChRs genes, as discussed below (for recent review, see Portugal neurons, and involved in mediating the release of 77 DA.32,60,64,67 and Gould ). The third and newest technique is GWAS. Several CHRNB3. Several reports show that b3-containing GWAS finding have been reported in the last 2 years, nAChRs have a physiologically significant role using sets of 300–500 000 SNPs and including large in dopaminergic neurotransmission and many numbers of participants. Impressive associations were nAChRs expressed on mice substantia nigra-VTA reported for the CHRNB3/CHRNA6 gene cluster78 and dopaminergic neurons are b3-dependent.68 In for variants in the CHRNA3/CHRNA5 cluster.79,80 addition to midbrain and striatum, b3-containing However, other GWASs do not support strong asso- receptors are also found in the mouse superior ciation of nAChRs with ND phenotypes.15 colliculus, lateral geniculate nucleus and retina.69 ND is a complex phenomenon encompassing several aspects. Thus, it is not surprising that many CHRNB4. The nicotinic b4-subunit is widely different ND-related phenotypes have been used by expressed in the mouse PNS, but in the CNS it has a researchers in genetic studies. Four major phenotype

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 917 Table 2 Summary table of nicotine dependence (ND)-related phenotypes used by various candidate-gene association studies and GWAS

Phenotype measures Reports that used the measure

FTND (Fagerstrom test for nicotine dependence)/FTQ-based phenotypes Extremes phenotype: controls (FTND = 0 but smoked at least 100 cigarettes during Bierut et al.78; Saccone et al.76; Thorgeirsson et al.80 lifetime) vs cases (FTND > 4) Extremes phenotype: low (FTQ < = 4) vs high (FTQ > 6) ND Greenbaum et al.81; Weiss et al.82 Extremes phenotype: low (FTQ < = 3) vs high (FTQ > 7) ND Silverman et al.117 Extremes phenotype: low (FTND < = 2) vs high (FTND > 8) ND Feng et al.75 FTND score (0–10 point scale) Li et al.87; Stevens et al.90; Chen et al.96 Age adjusted FTND score Feng et al.75 Separate FTND items Hung et al.100 ND defined by FTND scores and carbon monoxide levels Uhl et al.15

Heaviness of smoking index (HIS) Extremes phenotype: controls (HSI = 0) vs cases (HIS > 3) Hung et al.100 HSI score (0–6 point scale) Li et al.87

Smoking quantity—cigarettes per day (CPD) Extremes phenotype: controls always smoked < 5 CPD; cases smoked > than 25 CPD Berretini et al.79 Extremes phenotype: controls always smoked < 5 CPD; cases smoked > than 30 CPD Stevens et al.90 Extreme phenotype: cases > 5 CPD; control smoked 1–100 cigarette in lifetime, but Sherva et al.89 not regularly Habitual smoking: affected smoked > 20 CPD for 6 months or more; unaffected Bierut et al.88 smoked < 10 CPD. Smoking quantity levels: 0 (1–10 CPD), 1 (11–20 CPD), 2, (21–30 CPD) and 3 (31 CPD Thorgeirsson et al.80; Stevens et al.90 and more) Number of cigarettes consumed per day Amos et al.101;Liet al.87; Voineskos et al.113; Spitz et al.103

Early subjective responses to tobacco (grouped by factor analysis) Adverse (depressed, paranoid, confused, anxious, irritable) Ehringer et al.115; Zeiger et al.108 Negative physical (dizzy, nauseous, lazy, drowsy) Ehringer et al.115; Zeiger et al.108 Positive (mellow, top of the world, energetic, creative, sociable) Ehringer et al.115; Zeiger et al.108 Tobacco sensitivity (0nauseous0 or 0dizzy0 feeling) Ehringer et al.115; Zeiger et al.108; Sherva et al.89 Physical effects (i.e., nausea, buzzing, dizziness, heart poundering, unpleasentness Hutchinson et al.112 and palm sweating) Cognitive effects (i.e., calm, attentive, satisfied, euphoric, energized, stimulated, Hutchinson et al.112 clear headed and mentaly alert) Experience of rush or high Hutchinson et al.112; Sherva et al.89 Reward experience Hutchinson et al.112

DSM-IV criteria based ND phenotype Endorsment of three or more (out of seven) dependence symptoms Hoft et al.109

Smoking initiation Initiated smoking versus did not initiate smoking ( < 1 complete cigarette) Silverman et al.117; Greenbaum et al.81

Age of initiation Ehringer et al.115; Zeiger et al.108; Schlaepfer et al.83

Smoking cessation treatment outcome Self-report of smoking abstinence and low carbon monoxide reading Hutchinson et al.112 Number of unsuccessful quit attempts Hoft et al.109 Smoking cessation probability Breitiling et al.91 Abstinence rates at end of treatment and after a 6 month follow–up period Conti et al.116 Time to relapse Conti et al.116 Withdrawal symptoms at target quit date Conti et al.116 Successful versus unsuccessful quitters (for at least 6 weeks after completion of trail) Uhl et al.15 Time to cessation (in ever heavy smokers) Breitiling et al.91

Other phenotypes Revised tolerance questionnaire (RTQ) Feng et al.75 Number of days smoked in the last 6 months Ehringer et al.115; Zeiger et al.108 Peak tobacco use (maximal CPD) Ehringer et al.115; Zeiger et al.108 Typical pattern of use (0 (none) to 7 (2 or more times per day) score Ehringer et al.115; Zeiger et al.108 Pack-years of exposure Amos et al.101 Current smoker versus current nonsmokers among schizophrenic patients De Luca et al.127 Composite International Diagnostic Interview—Substance Abuse Module Schlaepfer et al.83 (CIDI-SAM) Time to first cigarette Stevens et al.90

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 918 categories are commonly used in the field, each Power considerations are important in evaluating reflecting a different but mostly interrelated aspect findings.14 The sample sizes of the studies mentioned of ND-related behavior: below are very heterogeneous, from B200 to more than 20 000 participants. A further consideration is 1. Severity of ND: Assessed by a scale such as the that spurious findings may occur due to multiple FTND or FTQ (see below) or by DSM-IV criteria. testing that inflates false-positive rates.14 Also, the 2. Smoking quantity: Consumption of cigarettes per ethnic origin of the samples should be taken day (CPD). into consideration when comparing the results of 3. Smoking cessation: Defined as success or failure of studies.87 attempts to quit, not necessarily in the context of specific smoking cessation treatment or outcome of CHRNA5–CHRNA3–CHRNB4 cluster smoking cessation treatment, with comparison Recently, a plethora of evidence has accumulated between successful and unsuccessful quitters. suggesting the involvement of this gene cluster in ND 4. Subjective response to tobacco: Considered an in particular and in substance abuse in general. important endophenotype for ND; early subjective This 15q25.1 locus contains three genes encoding responses to nicotine exposure may be related to nAChRs subunits, which lie very close to and in future ND status. strong linkage disequilibrium (LD) with each other, so Smoking initiation is considered as a separate that most reports treat them as a cluster. To date, phenotype from ND. Genetic and environmental evidence for genetic association of these nAChR- factors that influence SI may be different from those encoding loci with ND-related phenotypes is more that influence ND, although they may also be robust and convincing than evidence for any other shared.12,81 Age of SI (and its influence on later life nAChR gene association with ND. Emphasis should smoking behavior) is another phenotype that has been be given to the fact that in the case of CHRNA5– studied from the genetic perspective (Table 2).82,83 CHRNA3–CHRNB4, GWAS findings, which are con- Development of ND requires exposure to nicotine. sidered superior to candidate-gene designs because of Therefore, the most appropriate controls for studying their ‘hypothesis-free’ approach, validate and con- factors that influence severity of ND in smokers are verge with candidate-gene studies of nAChR genes individuals who initiated smoking and had sufficient that are based on the biologically plausible role of exposure to develop ND but did not. This approach was these genes in mediating nicotine influence in the implemented by Sacconne et al.76 and Bierut et al.78 brain. This unexpected outcome of the GWAS Implementation of each phenotype category re- strongly supports the rationale of the candidate-gene quires availability of specific data, which is depen- approach in ND. dent on a priori study design. As a result, there is great variability in the ND phenotypes that are used CHRNA5–CHRNA3–CHRNB4 cluster and ND. In the and this is an important factor limiting the replic- candidate-gene study of Saccone et al.,76 seven SNPs ability of findings. A further consideration is that located within the CHRNA5–CHRNA3–CHRNB4 testing a series of different phenotypes in the same cluster were associated with ND. This large 348 study raises concern regarding spurious findings that candidate-gene study, accompanied by a GWAS are a consequence of multiple testing. The fact that performed in parallel,78 included a European the different phenotypes are mostly highly correlated ancestry (EA) sample of 1050 cases and 879 controls may somewhat mitigate this concern. Formal evalua- from the United States and Australia. In both studies, tion of the degree of correlation among different ND cases were defined as individuals with scores X4on phenotypes is greatly needed and would enhance the FTND, whereas controls were people who had efforts to compare studies that use different pheno- been exposed to smoking (must have smoked at least types. 100 cigarettes in their lifetimes) but had an FTND The instruments used to assess severity of ND differ score of 0 during the heaviest period of smoking. As from study-to-study, causing inconsistency in defin- noted above, this control selection strategy allows ing ND phenotypes. One of the most accepted genetic effects specific to ND to be examined more measurement tools is the Fagerstrom test for nicotine effectively. dependence (FTND), which is based on the Fager- Biologically, the most plausible evidence for a risk strom tolerance questionnaire (FTQ)84 with several variant was for the non-synonymous SNP, improvements. The FTND questionnaire focuses on rs16969968, in the CHRNA5 gene (P = 0.00064).76 assessing the intensity of nicotine physiological Individuals with two copies of the minor allele ‘A’ dependence (0–10 score, higher score means higher (causing substitution of the 398 amino acid aspar- ND).85,86 Heaviness of smoking index (HSI) is com- agine with aspartic acid) were nearly twice as likely to posed of two FTND items, smoking quantity (SQ) or have ND, in a recessive model of inheritance. There maximum cigarettes smoked in a 24 h period.3 Several was even stronger association of the CHRNA3 30 UTR distinct FTND/FTQ-based ND phenotypes (with SNP, rs578776 (P = 0.0003). Five other SNPs in the different FTND cut-off scores for case and control) cluster showed impressive association with ND in have been described and used for case–control this sample: rs1051730, rs684513, rs637137, association analysis (Table 2). rs3743078 and rs3813567.76

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 919 A recent paper from this group88 validated and Icelandic smokers (defined as FTNDX4, or fulfill- refined the finding in a sample different from the ment of at least three out of seven DSM-IV criteria) vs shared Saccone et al.76 candidate gene/Bierut et al.78 3506 low-quantity smokers, as a measure of ND GWAS sample and consisting of 1236 individuals resembling the Saccone et al.76 phenotype and was from 219 pedigrees of European descent . In a sample significant with P =7Â 10À15 and an odds ratio (OR) of originally collected for the study of alcoholism, the 1.4. As smoking is also a risk factor to lung cancer affected persons were defined as smoking 20 cigar- (LC) and peripheral vascular disease, association of ettes a day for at least 6 months and the unaffected the rs1051730 SNP was analyzed and found signifi- persons as smoking 10 cigarettes or fewer per day. cant for both diseases (P = 1.5 Â 10À8 and 1.4 Â 10À7, Two distinct, independent genetic associations with respectively, in combined samples). habitual smoking were found, with low correlation A relatively small case–control study (435 partici- between them. The first is marked by rs16969968 pants) of Sherva et al.89 focused on the rs16969968 (P = 0.007) and the second by rs578776 (P = 0.009). SNP. Besides replication of association of this The absolute value of D0 is high, but the correlation SNP with smoking status by itself (cases smoked between 16969968 and rs578776 is low (r2 < 0.15), more than five cigarettes a day for more than 5 years, rendering them statistically independent. As ex- whereas controls smoked between 1 and 100 cigar- pected, other SNPs that were highly correlated with ettes in their lifetime) (P = 0.01), the researchers rs16969968 (rs1051730, rs2036527, rs17486278 and explored association of this variant with early smok- rs17487223) were also associated with habitual ing experience. On the basis of the retrospective smoking (P = 0.02–0.086).88 To evaluate the effect of report of feelings on smoking the first cigarette, the the CHRNA5 non-synonymous SNP, rs16969968 rs1696968 SNP was significantly associated with (aspartic acid to asparagine in 398 position), on experiencing pleasurable response (buzz or rush) receptor function, changes in intracellular calcium among Caucasians when first smoking (P = 0.01). This after nicotine agonist exposure were measured in feeling was associated with risk of current smoking HEK293T cells expressing either a4b2a5D398 or status (P < 0.0001).89 a4b2a5N398 nAChRs. The receptor with the aspartic In agreement with Bierut et al. study,88 two acid variants (D) showed a greater maximal response different groups of SNPs were associated with ND in to nicotine than N398 receptors. However, no differ- a case–control sample of 1452 heavy smokers ence was shown in the isoform expression.88 (CPD > 30) and 1395 light smokers (CPD < 5) of EA.90 Berrettini et al.79 also obtained evidence for the Eight SNPs which are in high LD (including relevance of the CHRNA5–CHRNA3–CHRNB4 gene rs16969968 and rs1051730) were associated with cluster to ND in a GWAS (by Affymetrix 500K SNP increased smoking risk (See Table 3), while 11 SNPs chip) in two European populations combined into a (including rs578776) were protective and associated sample of B7600 individuals, but used a different with decreased risk of heavy smoking (See Table 3). phenotype, consumption of cigarettes per day (CPD). All associations withstand Bonferonni correction for For dichotomous cases–control analysis, cases were multiple testing.90 As reported previously,88 the corre- defined as individuals smoking more than 25 cigar- lation between the two separate SNP groups is low.90 ettes per day, whereas controls reported CPD less On the other hand, no association of this gene than 5. The CHRNA3 SNP rs6495308 was signifi- cluster with smoking cessation was found in a large, cantly associated with CPD consumption (pooled P- well-powered retrospective cohort study from Ger- value of 0.000069), and the CHRNA3 SNP rs951266 in many.91 Seven SNPs which are related to other ND a sub-sample. However, no SNP reached the required phenotypes were genotyped (rs684513, rs637137, GWAS level of significance, 10–7. Another CHRNA3 rs16969968, rs578776, rs1051730, rs3743078, intronic SNP, rs1317286, was strongly associated with rs3813567) in a sample of 1446 heavy smoking CPD in a third sample of B8000 persons individuals ( > 20 CPD). No significant association (P = 0.0000026) (using a panel of B6000 SNPs in with smoking cessation probability was found, B2000 genes).79 Unfortunately, there were no geno- although the study had enough power to detect even typed SNPs common to those described in the papers small effects. These results suggest that CHRNA5– of Saccone et al.,76 Bierut et al.78 and Berretini et al.79 CHRNA3–CHRNB4 variants contribute to the devel- However, SNPs in Berretini et al. study are in high LD opment of ND, but not to the likelihood of cessation and correlation to the associated SNPs of Saccone when the ND is already well established.91 et al. and Bierut et al. An important reservation is that all the GWAS and An important replication of these finding comes case–control studies reviewed above (except the from a GWAS of SQ in B11,000 Icelandic smokers.80 relatively small sample of Sherva et al.89) were This GWAS clustered the SQ phenotype, into four conducted on European origin samples. Individual levels: 0 (1–10 CPD), 1 (11–20), 2 (21–30) and 3 (31 SNP associations may be different, or even not exist at and more). Allele T of the rs1051730 SNP in CHRNA3 all, among diverse ethnic groups. For example, the was most strongly associated with SQ (P =5Â 10À16, frequency of the rs16969968 SNP is dramatically surviving a significance threshold of 2 Â 10À7). This is different in EA (37%) and African populations one of the variants reported by Saccone et al.76 (0%).88Further replication of these findings in other rs1051730 was then tested in a subset of 2394 ethnic groups is warranted.

Molecular Psychiatry 920 oeua Psychiatry Molecular

Table 3 Summary of linkage, candidate-gene association and GWAS findings in nAChr genes for various ND-related phenotypes

Gene/ SNP Position Location Smoking phenotype Study type P-value Sample size Sample origin References polymorphism (individuals)

CHRNA4 rs2236196 61448000 30 UTR Smoking quantity, heaviness of FBAS (2 candidate genes) 0.03 (dm, rm); 2037 African-American Li et al.87 smoking index, FTND 0.01(dm, rm); 0.007 (dm, rm) Physiological effect; rush or high AS (CHRNA4 gene) 0.003; 0.02; 0.03 316 Americans, Hutchison et al.112 experience, cognition mostly EA (SNP  cigarette interaction) ND and nAChRs brain multiple in variation Genetic Smoking abstinence AS (CHRNA4 gene) 0.01 353 EA Hutchison et al.112 (SNP  treatment condition  time point interaction) rs3787137 61449544 Intron 5 Smoking quantity, heaviness of FBAS (2 candidate genes) 0.05; 0.07; 0.04 2037 AA Li et al.87 smoking index, FTND (am) rs1044397 61451548 Exon 5, Nicotine addiction, FTND score, FBAS (2 candidate genes) 0.007; < 0.001; 621 Chinese male Feng et al.75 synonymous RTQ score < 0.001 rs1044396 61451578 Exon 5, Nicotine addiction, FTND score, FBAS (2 candidate genes) 0.008; 0.005; 0.006 621 Chinese male Feng et al.75 synonymous RTQ score Lerer B and Greenbaum L Smoking quantity FBAS (2 candidate genes) 0.03 (dm, rm) 2,037 EA Li et al.87 rs2273504 61458505 Intron 5 FTND, heaviness of smoking index FBAS (2 candidate genes) 0.046(dm)/0.02 2,037 EA Li et al.87 (rm); 0.06 rs6122429 61463650 50 promoter Reward (SNP  cigarette interaction) CCAS (CHRNA4 gene) 0.002 316 Americans, Hutchison et al.112 mostly EA rs3746372 61502479 Near gene Large number of cigarettes smoked AS (2 candidate genes) 0.02 241 EA schizophrenic Voineskos et al.113 daily, among schizophrenia patients patients DS20119- — Microsatellite Maximum number of cigarettes in Linkage analysis (families) Max LOD: 4.22 623 Finnish Saccone et al.76 DS20S178 markers in gene 24 h period region D20S481- — Microsatellite Smoking quantity Linkage analysis (families) Genome-wide, EA Wang et al.111 S20S480 markers in gene P = 0.032 region

CHRNB2 rs2072658 152806849 Upstream Early subjective response to tobacco: AS (2 genes) 0.002; 0.022 1068 Mixed ethnicity, Ehringer et al.115 negative physical, positive American rs12072348 152813916 Intron 5 Smoking initiation CCAS (11 genes) 0.07(al)/0.053(g) 384 Jewish females Greenbaum et al.81 rs2072660 152815345 30 UTR Early subjective response to tobacco: AS (2 genes) 0.014 1068 Mixed ethnicity, Ehringer et al.115 tobacco sensitive American Smoking initiation CCAS (11 genes) 0.02(al)/0.015(g) 384 Jewish females Greenbaum et al.81 rs2072661 152815504 30 UTR Abstinence rates (6 month follow- AS (58 genes) 0.0002; 0.0002; 412 EA Conti et al.116 up); withdrawal symptoms; time to 0.0009 relapse

CHRNB3-CHRNA6 rs10958726 42655066 50 promoter ND (FTND = 0 vs FTND > 4) CCAS (384 genes) < 0.001 1939 EA Saccone et al.76 rs13277254 42669139 CHRNB3–50 near ND (FTND = 0 vs FTND > 4) GWAS, case–control < 0.0001 1939 EA Bierut et al.78 gene rs6474413 42670221 CHRNB3–50 near ND (FTND = 0 vs FTND > 4) GWAS, case–control; GWAS 0.00009 1939 EA Saccone et al.;76 gene case–control Bierut et al.78 rs7004381 42670318 CHRNB3–50 near Number of quit attemptsa AS, partially FBAS 0.011(am)/ 1051 Ethnically diverse Hoft et al.109 gene (2 candidate genes) 0.0024(rm) rs4950 42671790 CHRNB3–50UTR Early subjective response to tobacco: AS (2 genes) 0.004; < 0.001; 1056 Ethnically diverse Zeiger et al.108 negative physical, positive, adverse 0.02 adolescents, mostly Caucasians 42671790 Number of quit attemptsa AS, partially FBAS 0.03(am)/ 1051 Ethnically diverse Hoft et al.109 (2 candidate genes) 0.021(rm) rs13280604 42678743 CHRNB3–Intron 1 Early subjective response to tobacco: AS (2 genes) 0.001; < 0.001; 1056 Ethnically diverse Zeiger et al, negative physical, positive, Adverse 0.03 adolescents, 2008108 mostly Caucasians 42678743 Endorsment of 3 or more dependence AS, partially FBAS 0.037(am); 1051 Ethnically diverse Hoft et al.109 symptoms (DSM-IV based); Number (2 candidate genes) 0,036(am)/ of quit attemptsa 0.024(rm) Table 3 Continued

Gene/ SNP Position Location Smoking phenotype Study type P-value Sample size Sample origin References polymorphism (individuals)

rs4952 42706222 CHRNB3– Exon 5 ND (FTND = 0 vs FTND = > 4) CCAS (384 genes) 0.016 1939 EA Saccone et al.76 (synonymous) rs4953 42706816 CHRNB3–Exon 5 Early subjective response to tobacco: AS (2 genes) 0.05; 0.04 1056 Ethnically diverse Zeiger et al.108 (synonymous) positive, adverse adolescents, mostly Caucasians 42706816 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.016 1939 EA Saccone et al.76 42706816 Number of quit attempts AS, partially FBAS (2 0.03(am)/ 1051 Ethnically diverse Hoft et al.109 candidate genes) 0.031(rm) rs7012713 42711460 CHRNB3–30 near Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 0.023 2847 EA Stevens et al.90 gene years) vs light (CPD < 5, at least 1 year) rs9298629 42725343 CHRNA6–30 near ND severity: low (FTQ < = 4) vs high CCAS (11 genes) 0.035 (G) 384 Jewish females Greenbaum et al.81 gene (FTQ > 6) ND rs2304297 42727356 CHRNA6–Exon 6 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) < 0.01 1939 EA Saccone et al.76 rs1072003 42739158 CHRNA6–Intron 2 ND severity: low (FTQ < = 4) vs high CCAS (11 genes) 0.053 (G) 384 Jewish females Greenbaum et al.81 (FTQ > 6) ND hCV25772398 Intron 3 Early subjective response to tobacco: AS (2 genes) < 0.0001; 0.006 1056 Ethnically diverse Zeiger et al.108 positive, adverse adolescents, mostly Caucasians rs7828365 42748471 CHRNA6–50 near Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 0.036 2847 EA Stevens et al.90 gene years) vs light (CPD < 5, at least 1 year)

CHRNA7 rs1909884 30226590 Intron 4 ND severity: low (FTQ < = 4) vs high CCAS (11 genes) 0.016 (G) 384 Jewish females Greenbaum et al.81 (FTQ > 6) ND D15S1360 — Intron 2 Smoking risk (among schizophrenia AS (1 gene) 0.015 177 Schzophrenic De-Luca et al.127 patients) patients, canada rebu n Lerer ND B and and nAChRs Greenbaum brain L multiple in variation Genetic CHRNA2 rs2280376 27373679 30 UTR Adolescent antisocial drug 50 Candidate-gene < 0.002 231 American Corley et al.123 dependence association study adolescents rs2565065 27383938 Intron 2 Smoking cessation on buproprion 11 Candidate-gene Likelihood ratio EA Heitjan et al.124 versus placebo pharmacogenetic study test, P = 0.027 D8S258 — Microsatellite Lifetime endorsment of three or more Linkage analysis Max LOD 2.7 607 Mostly EA Swan et al.120 markers in gene dependence symptoms (DSM-IV region based) Smoking intensity in schizophrenic Linkage analysis Max LOD 2.34, 98 EA Faraone et al.124 pedigrees O,OO1

CHRNA9 rs4861065 40039152 Intron 3 ND severity: low (FTQ < = 4) vs high CCAS (11 genes) 0.038(al) 384 Jewish females Greenbaum et al.81 (FTQ > 6) ND

CHRNA5-CHRNA3-CHRNB4 rs17483686 76520445 IREB2 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 5.32 Â 10À 6 2847 EA Stevens et al90 years) vs light (CPD < 5, at least 1 year) rs8034191 76593078 LOC123688 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 5.44 Â 10À 7 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs905739 76632165 Near gene cluster Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 6.98 Â 10À 7 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs2036527 76638670 Near gene cluster Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 1.14 Â 10À 7 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 oeua Psychiatry Molecular year) rs684513 76645455 CHRNA5–Intron 1 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.0087 1939 EA Saccone et al.76 FTND score AS (7 SNPs) 0.0222 1121 EA Chen et al.96 Smoking: heavy (CPD > 30 for 5 years) CCAS (7 genes) 1.13 Â 10À 7 2847 EA Stevens et al.90 vs light (CPD < 5, at least 1 year) 921 922 oeua Psychiatry Molecular

Table 3 Continued

Gene/ SNP Position Location Smoking phenotype Study type P-value Sample size Sample origin References polymorphism (individuals)

rs871058 76645546 CHRNA5–Intron 1 Smoking cessation on buproprion Candidate-gene Likelihood ratio EA Heitjan et al.124 versus placebo pharmacogenetic study test, P = 0.019 rs17486278 76654537 CHRNA5–Intron 1 Habitual smoking (CPD > 20 vs FBAS (3 genes) 0.028 2284 EA Bierut et al.88 CPD < 10) ND severity: low (FTQ < = 4) vs high CCAS (10 genes) 0.0005 2827 EA Weiss et al.82 (FTQ > 6) ND, smoking early onset Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 7 Â 10À 8 2847 EA Stevens et al.90

years) vs light (CPD < 5, at least 1 ND and nAChRs brain multiple in variation Genetic year) rs680244 76658343 CHRNA5–Intron 1 Age at initiation of tobacco useb AS (3 genes) 0.002 2126 Americans, Schlaepfer et al.83 mostly Caucasians rs569207 76660174 CHRAN5–Intron 1 ND severity: low (FTQ < = 4) vs high CCAS (10 genes) 0.0042 2827 EA Weiss et al.82 (FTQ > 6) ND, smoking early onset Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 1.4 Â 10À 8 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) 76

rs637137 76661031 CHRNA5–Intron 2 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.0028 1939 EA Saccone et al. Lerer B and Greenbaum L Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 1.3 Â 10À 8 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs951266 76665596 CHRNA5–Intron 2 CPD (CPD < 5 vs CPD > 25) GWAS 0.0006c B15000 EA Berrettini et al.79 rs16969968 76669980 CHRNA5–Exon 6 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.0006 1939 EA Saccone et al.76 (syn) ND severity: low (FTQ < = 4) vs high CCAS (10 genes) 0.0009 2827 EA Weiss et al.82 (FTQ > 6) ND, smoking early onset Habitual smoking (CPD > 20 vs FBAS (3 genes) 0.007 2284 EA Bierut et al.88 CPD < 10) Smoking status; pleasurable buzz or CCAS (25 SNPs) 0.01; 0.01 435 Mixed ethnicity Sherva et al.89 rush during first cigarette smoking FTND score AS (7 SNPs) 0.0068; 0.0028 815, 1121 EA Chen et al.96 Smoking: Heavy (CPD > 30 for 5 CCAS (7 genes) 6.3 Â 10À 8 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs514743 76671282 CHRNA5–Intron 5 Age at initiation of tobacco useb AS (3 genes) 0.01 2126 Americans, Schlaepfer et al.83 mostly Caucasians rs578776 76675455 CHRNA3–30 UTR ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.0003 1939 EA Saccone et al.76 ND severity: low (FTQ < = 4) vs high CCAS (10 genes) 0.0048 2827 EA Weiss et al.82 (FTQ > 6) ND, smoking early onset Habitual smoking (CPD > 20 vs FBAS (3 genes) 0.009 2284 EA Bierut et al.88 CPD < 10) Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 1.37 Â 10À 6 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs1051730 76681394 CHRNA3–Exon 5 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.00099 1939 EA Saccone et al.76 SQ levels (CPD), ND (low-quantity GWAS 5 Â 10À 16 ;7Â 10À 15 13945 Iceland Thorgeirsson et smokers vs smokers) al.80 Habitual smoking (CPD > 20 vs FBAS (3 genes) 0.02 2284 EA Bierut et al.88 CPD < 10) ND severity: low (FTQ < = 4) vs high CCAS (10 genes) 0.0009 2827 EA Weiss et al.82 (FTQ > 6) ND, smoking early onset CPD; scores of FTND CCAS (3 genes) < 0.001; < 0.001 3621 Mixed ethnicity Spitz et al.103 FTND score AS (7 SNPs) 0.0237; 0.0039 815, 1121 EA Chen et al.96 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 9.3 Â 10À 8 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs3743078 76681814 CHRNA3–Intron 4 ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.011 1939 EA Saccone et al.76 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 5 Â 10À 9 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs1317286 76683184 CHRNA3–Intron 4 CPD (CPD < 5 vs CPD > 25) GWAS 0.0000026 B15000 EA Berrettini et al.79 rs11637630 76686774 CHRNA3–Intron 4 CCAS (7 genes) 5 Â 10À 9 2847 EA Stevens et al.90 Table 3 Continued

Gene/ SNP Position Location Smoking phenotype Study type P-value Sample size Sample origin References polymorphism (individuals)

Smoking: heavy (CPD > 30 for 5 years) vs light (CPD < 5, at least 1 year) rs7177514 76694461 CHRNA3–Intron 4 ND severity: low (FTQ < = 4) vs high CCAS (10 genes) 0.0032 2827 EA Weiss et al.82 (FTQ > 6) ND, smoking early onset rs6495308 76694711 CHRNA3–Intron 4 CPD (CPD < 5 vs CPD > 25) GWAS 0.000069 B15000 EA Berrettini et al.79 FTND score AS (7 SNPs) 0.0384 1121 EA Chen et al.96 rs8040868 76698236 CHRNA3–Exon 2 Age at initiation of tobacco useb AS (3 genes) 0.022 2126 Americans, Schlaepfer et al.83 mostly Caucasians rs8023462 76701789 Intergenic Age at initiation of tobacco useb AS (3 genes) < 0.001 2126 Americans, Schlaepfer et al.83 mostly Caucasians rs1948 76704454 CHRNB4–30 UTR Age at initiation of tobacco useb AS (3 genes) < 0.001 2126 Americans, Schlaepfer et al.83 mostly Caucasians rs17487223 76711042 CHRNB4–Intron 2 Habitual smoking (CPD > 20 vs FBAS (3 genes) 0.001 2284 EA Bierut et al.88 CPD < 10) Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 8.07 Â 10À 7 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs12440014 76713781 CHRNB4–Intron 2 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 3.45 Â 10À 7 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs11636605 76715933 CHRNB4–Intron 1 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 1.18 Â 10À 6 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs1316971 76717565 CHRNB4–Intron 1 Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 3.18 Â 10À 6 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs3813567 76721606 CHRNB4–50 near ND (FTND = 0 vs FTND > 4) CCAS (384 genes) 0.012 1939 EA Saccone et al.76

gene Lerer ND B and and nAChRs Greenbaum brain L multiple in variation Genetic Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 1.25 Â 10À 6 2847 EA Stevens et al.90 years) vs light (CPD < 5, at least 1 year) rs1996371 76743861 CHRNB4–50 near Smoking: heavy (CPD > 30 for 5 CCAS (7 genes) 3.33 Â 10À 4 2847 EA Stevens et al.90 gene years) vs light (CPD < 5, at least 1 year)

CHRNA10 D11S4046 — Microsatellite Habitual smoking Linkage analysis 0.0003 142 EA Gelernter et al.129 markers in gene region D11S4181 — Microsatellite FTND Linkage analysis Max LOD 2.25 505 Finnish Loukola et al.130 markers in gene region

Abbreviations: AA, African American; AS, association study; al, allele; am, additive model; CCAS, case–control association study; CPD, cigarettes per day; dm, dominant model; EA, European ancestry; FBAS, family-based association study; FTND, Fagerstrom test for nicotine dependence; FTQ, Fagerstrom tolerance questionnaire; G, genotype; GWAS, genome-wide association study; RTQ, revised tolerance questionnaire; rm, recessive model. All P-values data are uncorrected for multiple testing. a Only primary analysis results in Caucasians are presented. b Pooled results from CADD sample. c Did not pass quality control. oeua Psychiatry Molecular 923 Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 924 CHRNA5–CHRNA3–CHRNB4 involvement in age- association was shown in two samples of 815 dependent nicotine addiction. In accordance with and 1128 EA individuals, and the association the observation that onset of smoking during remained significant after correction for multiple adolescence is associated with more severe ND in testing. However, no association of any of the adult smokers,82,92,93 Weiss et al.82 studied a sample of seven SNPs genotyped in the region was found 2827 long-term smokers, using a dichotomized with cannabis abuse or dependence, in a separate definition of ND severity based on FTND scores: low sample.96 (FTND = 0–4) and high (FTND 6–10). Six SNPs were A recent study has shown association of the significantly associated with ND with P-values CHRNA5 non-synonymous SNP, rs16969968, with ranging from 0.0032–0.0005 (including rs1699968, cocaine dependence.97 Association of this SNP with rs1051730 and rs578776) only in the early onset of ND-related phenotypes was extensively described smoking group (age 16 years or younger), but not in above. Interestingly, while the minor allele (A, the late onset group (age 17 years or older). causing aspartic acid to asparagine substitution) is a In a five tagging SNP (rs680244, rs569207, risk factor for ND, it plays a protective role in cocaine rs16969968, rs578776 and rs1051730) haplotype dependence.97 The association of the same gene analysis of the CHRNA5–CHRNA3–CHRNB4 cluster, cluster with several drugs of abuse (nicotine, alcohol one haplotype (CCAGA) was detected to be a risk and cocaine dependence) may imply a common, factor, while another (CTGAG) was a protective factor general neurobiological mechanism of brain nAChR for smoking heaviness in adults.82 These predisposi- involvement in drug addiction, possibly because of tions were strongest if smoking began during adoles- cholinergic system modulation of DA release and DA- cence. These finding are in correlation to those mediated reward pathways. reported by Stevens et al.,90 who found two groups of SNPs in this cluster; one increased risk for heavy CHRNA5–CHRNA3–CHRNB4 and LC. Smoking is the smoking, while the other was protective.90 most important risk factor for LC.1,98 In the nineteenth century, LC was considered a rare disease.99 Since the CHRNA5–CHRNA3–CHRNB4 and other substance turn of the twentieth century, when cigarette smoking abuse. Comorbidity of nicotine and alcohol use became widespread in the developed world, LC rate is well documented but a common biological increased dramatically.98,99 However, the clear mechanism or genetic background has not been epidemiological link between LC and smoking was shown.9,83 Association of the CHRNA5–CHRNA3– proved conclusively only in the 1950s.98 CHRNB4 locus with alcohol dependence (defined by Two recent GWASs reported association of the DSM-IV) was shown for 16 polymorphisms (out of 42 CHRNA5–CHRAN3–CHRNB4 locus with LC but not genotyped) in two samples, spanning the CHRNA5– with ND. In a large GWAS, B317 000 SNPs (Illumina CHRNA3 region and including a SNP (rs588765) that HumanHap300 Beadchip) were analyzed in B2000 affects mRNA levels.94 These variants show low LD LC patients and 2625 controls of central European with most other ND associated SNPs reported earlier. origin and in five additional samples.100 The non- Further evidence for a possible shared genetics synonymous SNP, rs16969968, and rs8034191 (most background is supplied by Joslyn et al.95 who significant marker for LC, located in the nearby gene reported association of several SNPs in the PSMA4), which are in strong LD and highly corre- chromosomal region of CHRNA5–CHRNA3–CHRNB4 lated (D0 =1; r2 = 0.92), were found to be robustly locus (particularly rs1051730 and rs8034191) with associated with LC (rs16968869: allelic OR = 1.3, alcohol level of response, which is considered an P =1Â 10À20 overall; rs8034191: allelic OR = 1.3, intermediate phenotype for alcohol addiction.95 P =5Â 10À20 overall). The statistical risk for LC was Age at initiation of both tobacco use and alcohol observed irrespective of smoking status and was drinking was significantly associated with three other increased by these variants in current, former and SNPs within the CHRNA5ÀCHRNA3ÀCHRNB4 locus never smokers groups. However, no association (rs8023462, rs1948, rs514743) in a US sample of 1075 was detected with head and neck cancers (another individuals of diverse ethnicity. rs8023462 and smoking-related cancer). To rule out the association of rs1948 significantly predicted early age of initiation the 15q25 locus with LC through its involvement in for tobacco with a hazard ratio of 1.35 for the TT ND, the association of rs16969968 with ND was genotype of rs8023462 and a hazard ratio of 1.29 for analyzed in a subsample (for which smoking the CC genotype of rs1948.83 These findings were behavior data were available), but no association with replicated in a separate population-representative any of the individual FTQ indices or the HSI measure sample, and together they suggest that variants in was found. Other LC associated SNPs in the this gene cluster may influence and promote the early region are rs1051730, rs951266 and rs1317286, but use of tobacco and alcohol.83 they were not analyzed for ND.100 The authors Furthermore, the association of rs16969968 and of this GWAS state that the association with LC is rs1051730 with both ND (phenotype was defined as unlikely to be mediated by the association with ND, FTND score) and symptoms of alcohol use and as the increased risk was also seen among nonsmo- dependence (according to DSM-IV criteria, in another kers and was not seen in other smoking-related sample) was reported by Chen et al.96 The ND cancers.100

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 925 Further support for the importance of the increased smoking, it is also possible that genetic CHRNA5–CHRNA3–CHRNB4 locus in LC comes from variants in nAChRs directly influence mechanisms of aGWASofB1150 current and former smokers and lung carcinogenesis, without mediation by smoking B1150 controls that analyzed 315 450 tagging SNPs phenotypes. A third option is that these same loci are (Illumina HumanHap300 Beadchip) and showed that involved independently in the pathophysiology of the CHRNA3 SNP, rs1051730, and the neighboring both states.107 If so, combination of high nicotine PSM4 SNP, rs8034191, are associated with LC, with a consumption and higher LC susceptibility would be combined, adjusted OR of 1.32 (P =7Â 10À18) for synergistic. These controversial findings underscore rs1051730 and an OR of (P = 3.5 Â 10À18) for the challenge of defining and measuring behavioral rs8034191.101 This association was confirmed in two phenotypes (although some ND-related phenotypes replication sets. When the two SNPs were analyzed seem at first glance quite simple to define; for for association with two measures of ND (CPD and example, CPD). This may be a plausible explanation pack-years of exposure), there was only weak evi- of why associations with ND-related phenotypes are dence for an influence on smoking behavior, among more difficult to demonstrate than the associations former but not current smokers. with LC. These findings conflict with those of Thorgeirsson The relationship of CHRNA5–CHRNA3–CHRNB4 et al.,80 who showed that a sequence variant asso- loci with ND versus LC should be studied in the ciated with ND conferred risk for LC and peripheral future, taking into account the histological character- vascular disease through an effect on smoking istics of LC, interindividual variability of nicotine behavior. Robust evidence for association with ND metabolism and parallel exposure to other carcino- comes from the reports of Saccone et al.76, Bierut gens. Finally, as at least two independent variants et al.78 and Berretini et al.79; these reports do not within this cluster are associated with ND risk, it is address association with LC. Nevertheless, in two possible that the LC risk variants within the gene large GWASs100,101 direct and significant association cluster are independent and uncorrelated with at least with LC was shown, with weak evidence in only one one of the ND SNPs, notably the CHRNA3 30 UTR of them for association with ND. rs578776. Since the appearance of these three pivotal GWASs, intensive research efforts have been invested to eluci- CHRNB3–CHRNA6 cluster date the relationship of these loci to LC, but there are no A large body of evidence supports a contribution of clear conclusions. A systematic and comprehensive the CHRNA6–CHRNB3 gene cluster to ND (Table 3). description of all the observations is beyond the scope The strongest findings stem from the GWAS of Bierut of this review, which focuses on the role of nAChRs in et al.78 and the 348 candidate-gene study of Saccone ND. It is noted that the association of the rs8034191 and et al.,76 which are described above (Section rs1051730 SNPs with familial LC was described by Liu ‘CHRNA6–CHRNB3–CHRNB4 cluster and ND’). In et al.102 in the context of GWAS performed on a 194 the candidate-gene study, two SNPs in CHRNB3, familial LC patients and 219 control individuals sample rs6474413 and rs10958726, which are in very high LD (adjusted for age, sex and pack-years of cigarette and located in the putative 50 promoter region, exposure). However, data regarding ND in terms of accounted for the strongest signals in the analysis. FTND scores were unavailable.102 Two synonymous coding SNPs in exon 5 of CHRNB3, In a very recent follow-up publication to the GWAS rs4952 and rs4953, and rs2304297 located in 30 UTR of Amos et al.101 Spitz et al.103 combined the of CHRNA6 were also among the top signals.76 discovery and one of the replication samples and Converging evidence for association of CHRNB3 with observed that the highly correlated rs1051370 and ND came from the results of the GWAS,78 in which the rs8034191 are significant risk factors for both LC and gene appeared in the ‘top-group’. rs13277254 had a ND, contrary to the initial conclusions of Amos P-value of 6.54 Â 10À5, and was the 21st strongest et al.103 (Supplementary Table 1). Two ND-related signal in the GWAS out of 31 960 SNPs genotyped phenotypes were found to be significantly associated: with a custom array, with an OR of 1.19 among males CPD and FTND, but not other phenotypes such as age and an OR of 1.55 among females. However, this of SI.103 In contrast, when rs1051730 was tested association does not withstand correction for multiple among lifetime never smokers, no association with testing.78 The SNP (not genotyped in the parallel LC was found, supporting the etiological assumption candidate-gene study) is located 3 kb upstream of the that LC etiology is mediated by ND and these SNPs do CHRNB3 50 promoter. not contribute per se to carcinogenesis. These nAChR In the study of Stevens et al.,90 described in detail receptors are expressed in human and rodent bron- above, two SNPs were nominally significantly asso- chial epithelial cells,104 and in non-small-cell lung ciated with heavy smoking (rs7012713 in CHRNB3 cancers.105 We do not know at this time whether and rs7828365 in CHRNA6), but both did not survive polymorphisms in this cluster can directly increase correction for multiple testing. On the other hand, no the sensitivity of the target tissues to tobacco smoke association with heavy smoking in this sample was and/or nicotine toxicity.106 Besides the option that detected for any of the six ND-related CHRNB3– variants in the CHRNA5–CHRNA3–CHRNB4 loci CHRNA6 SNPs reported by Saccone et al.76 and Bierut increase LC risk solely due to involvement in et al.78

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 926 A recent study108 focused on the association of four CHRNB3 and rs7828365 in CHRNA6), but both did SNPs in the CHRNB3–CHRNA6 loci with early not survive correction for multiple testing. Nominal subjective response to tobacco. Initial subjective associations of rs9298629 (P = 0.035, genotype) and response to tobacco is important for reinforcement rs1072003 (P = 0.053, genotype) with severity of ND mechanisms and may be an endophenotype for ND, were reported by Greenbaum et al.,81 in a small sized acting as a mediator between the genotype and case–control study of 11 nAChRs gene among female dependence, differentiating individuals who develop Israeli students aged 20–30 years, divided into high ND over time from those who do not.89 The partici- (n = 127) and low ND (n = 115) groups. pants were retrospectively asked about early subjec- tive responses when they started smoking, by CHRNA4 answering 23 questions such as ‘In the period shortly Two genome-wide linkage analyses found significant after you used tobacco, did it make you feel or suggestive linkage of the 20q13.2–q13.3 region to (subjective effect)?’ to which individuals answered smoking behavior. Significant evidence for linkage of yes or no. The subjective effects were grouped into marker region DS20119-DS20S178 with maximum three factors: negative physical (lazy, drowsy and so number of cigarettes smoked in a 24-h period was on), positive (creative, sociable, energetic and so on) found in a Finnish sample of 623 individuals from and adverse (depressed, paranoid, confused and so 155 families; the peak multipoint LOD score was 4.22 on). Two independent US samples (1056 ethnically at 72.90 cM.110 The nearby CHRNA4 gene is posi- diverse adolescents and 1524 families) were stu- tioned at B98 cM on chromosome 20, although the died.108 The rs4950 and rs13280604 SNPs, located in evidence for linkage to this gene in the analysis was the CHRNB3 50 UTR and intron 1, respectively, were weak. In a genome-wide permutation linkage study of robustly associated with all three subjective response the Framingham Heart Study (FHS) population using factors to initial tobacco use, in both samples. 401 autosomal microsatellite markers, significant Significant results were also detected for the other linkage of the marker region D20S481-D20S480 with two SNPs genotyped, rs4953 (found to be associated SQ was reported (interval 78–86 cM, peak 84 cM, with ND in the Saccone et al.76 study, as mentioned genome wide P = 0.032).111 above), hCV25772398, and haplotypes composed of Several candidate-gene studies have shown evi- the four SNPs. However, no association was shown dence for possible association of the CHRNA4 gene between these SNPs and patterns of use of either with ND phenotypes. Two synonymous SNPs, tobacco or alcohol. Strong limitations of this study are rs1044396 and rs1044397 in exon 5 of the gene, were reliance on retrospective reports of young indivi- significantly associated with a protective effect duals, multiple testing and the possibility that other against ND, in a sample of 621 males from 206 substances were tried in the same period of experien- families with multiple nicotine addicted siblings cing tobacco and alcohol.108 from Anhiu province, China.75 For both SNPs (which Hoft et al.109 investigated the association of eight are in almost complete LD), significant associations SNPs in the CHRNA6–CHRNB3 cluster with smoking were found for all three addiction-related phenotypes behavior in a representative sample of US house- measured: addicted cases (FTND scores > 8) vs non- holds, which included 1051 participants, nearly half addicted controls (FTNDr2), age adjusted FTND and of them members of sibling pairs.109 ND was assessed RTQ (revised tolerance questionnaire) scores (quanti- using DSM-IV criteria and individuals who endorsed tative phenotypes). Association of the rs1044397 SNP three or more of the seven dependence symptoms with age-adjusted FTND and RTQ scores survived were considered dependent. The second phenotype Bonferonni correction. A haplotype composed of all was the number of unsuccessful quit attempts. In a six SNPs genotyped in the CHRNA4 gene (including family-based sample, the CHRNA6 rs2304297 SNP rs1044396 and rs1044397) with 22.5% frequency in and rs13280604 in CHRNB3 were associated with the population was significantly associated with three or more dependence symptoms among current protection against ND.75 Caucasian smokers. When unrelated singletons (in Li et al.87 analyzed the association of six SNPs in the whole sample or within a current frequent the CHRNA4 gene with SQ, FTND and HSI in 2037 smokers sub-population only) were examined for individuals from 602 nuclear families of EA or genetic association with ND, nominal association Africans American (AA) origin. Association of was detected in these two SNPs, as well as in rs2236196 with all three measures was reported in CHRNB3 rs7004381 and rs4950 and CHRNA6 the AA subsample, and remained significant for HSI rs892413.109 Four SNPs in the CHRNB3 gene and FTND after correction for multiple testing. In the (rs7004381, rs4950, rs4953 and rs13280604) and one EA subsample, SNPs rs1044396 and rs2273504 were CHRNA6 SNP, rs2304297, were associated with associated with SQ and FTND, respectively. Haplo- number of quit attempts, regardless of ethnicity. Only type analysis of rs2273505–rs2273504–rs2236196 the association of rs7004381 with quit attempts showed significant association withstanding multiple survives Bonferonni correction.109 testing with all the ND measures in AA females.87 In the Stevens et al.90 study, which is described in The same CHRNA4 30 UTR SNP, rs2236196, was detail above, two SNPs were nominally significantly found to be associated with the subjective physiolo- associated with heavy smoking (rs7012713 in gical effects of smoking and the experience of a rush

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 927 or high, in a sample of 316 participants of diverse with SI (P = 0.02 for allele; P = 0.015 for genotype). ethnicity tested for subjective responses to smoking in Multi-SNP haplotypes, including rs2072660, were the laboratory.112 After 8 h of smoking abstinence, protective against SI (P = 0.036). None of the findings each participant smoked three high-nicotine cigar- withstood correction for multiple testing.81 ettes, and was then assessed for four-dimensions of Association of the rs2072658 SNP, located 42 bp behavioral sensitivity to nicotine (endophenotype of upstream to the CHRNB2 transcription initiation site ND): (1) physiological, (for example, nausea and heart with initial subjective response to both tobacco and pounding), (2) rush or high, (3) cognitive and (4) alcohol, was reported by Ehringer et al.115 in a sample reward. A subjective response of reward was asso- of 1068 ethnically diverse young adults (mean age ciated with the 50 promoter SNP rs6122429. Subse- B18 years). The research assessment tools, methods quently, the same research group reported that and limitations are similar to those described for this rs2236196 was associated with smoking cessation type of study in relation to CHRNB3–CHRNA6108 in outcomes, in the context of a clinical trial of Section ‘CHRNA5–CHRNA3–CHRNB4 cluster’. transdermal nicotine patch versus nicotine nasal rs2072658 was associated with two tobacco pheno- spray for smoking abstinence (there were no signifi- types: tobacco negative physical and tobacco positive. cant difference in rs2236196 genotype between the Interestingly, the same SNP was also associated to two two treatment groups). Smoking abstinence was alcohol phenotypes: alcohol negative physical and assessed by self-report and exhaled carbon monoxide alcohol adverse. rs2072660, located in CHRNB2 exon reading. Among the 353 participants in this trial, 6, was associated with another tobacco phenotype, rs2236196 was associated with smoking abstinence at sensitive, which relates to feeling ‘nauseous’ or the end of treatment and 6 months later (P = 0.01).112 ‘dizzy’. However, no association of either SNP was To investigate the biological relevance of these two found with pattern of use.115 SNPs, luciferase activity of a construct containing the The rs2072661 SNP in the 30 UTR region of two different alleles was assayed as a marker for gene CHRNB2 was associated with success of pharma- expression.112 The construct containing the T allele of cotherapy for ND. Conti et al.116 performed a study of the 30 UTR rs2236196 SNP showed greater luciferase 58 candidate genes (genotyping 1295 SNPs) in a activity than the construct containing the C allele. In sample of 222 individuals receiving bupropion and postmortem brain tissue, a4b2-receptor binding in the 195 receiving placebo (all of EA). Independent of NAc was greater among carriers of the TC genotype treatment, rs2072661 was associated with abstinence than carriers of the TT genotype.112 Regarding the rate at the end of treatment (adjusted P = 0.01), and rs6122429 promoter SNP, luciferase activity in the after a 6-month follow-up period (adjusted P = 0.0002, construct containing the G-allele was fivefold greater significant with adjustment for the number of genes than the A-allele construct, probably due to a tested). The highly correlated 30 UTR rs2072660 SNP difference in transcription factor binding to this had a similar impact on these phenotypes. Further- promoter region. This unique study design of several more, rs2072661 had a role in the time to relapse and levels of experimental approaches (cell culture, withdrawal symptoms at target quit date.116 receptor binding in postmortem brain tissue, sub- However, several independent studies did not jective effects in a human laboratory model and demonstrate significant association of CHRNB2 with treatment outcome data) provides important informa- smoking-related behaviors or ND phenotypes (for tion regarding how genetic variation may affect ND both individual SNPs and haplotypes). In the Feng phenotypes at the biological level.112 et al.75 study (research details described in Section Association study of the CHRNA4 gene and heavy ‘CHRNA4’) rs2072658 and rs2072660 discussed smoking among schizophrenia patients showed sig- above, as well as other SNPs including rs2072661, nificant association between rs3746372 and large did not show significant association with ND. number of cigarettes smoked daily, as well as Negative results for the same SNPs were reported in interaction between this SNP and rs3787116.113 two independent US samples—the study by Silver- Schizophrenia patients are frequent (up to 80%, a man et al.,117 in which an association with SI or ND much higher rate than in the general population) and (similar phenotypes to Greenbaum et al.81) was heavy smokers, and the association between schizo- studied, and that by Li et al.,87 described in Section phrenia and smoking is of special interest. Tobacco ‘CHRNA4’. No significant associations of individual use is thought to alleviate neurocognitive deficits(s) of markers or haplotypes with smoking risk and beha- this disorder and to be an attempt of self-medicate.114 vior were detected by Leuders et al.,118 and by Voineskos et al.,113 in a sample of schizophrenic CHRNB2 smokers. In the study of Stevens et al.,90 no associa- Greenbaum et al.81 genotyped 39 SNPs located in 11 tions with heavy smoking were found for 10 SNPs nAChRs genes in a sample of Jewish, Israeli women, analyzed within the CHRNA4 and CHRNB2 genes. all undergraduate students aged 20–30 years. The Recently, highly significant gene–gene interactions sample consisted of 142 non-initiators of smoking and for the gene pairs CHRNA4–CHRNB2, CHRNB2– 242 initiators (SI) of whom 127 were defined as high NTRK2, CHRNA4–BDNF and CHRNA4–NTRK2 were ND (FTQX6) and 115 were low ND (FTQ < 4) . The reported with regard to ND in a case–control rs2072660 SNP in CHRNB2 was nominally associated sample.119 The SNPs included in the CHRNB2–

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 928 CHRNA4 gene pair interaction model were sample of Greenbaum et al.,81 the CHRNA7 intron 4 rs2072660, rs2072661 (CHRNB2) rs2273504, SNP, rs1909884, showed nominal genotypic rs2229959 and rs2236196 (CHRNA4).119 The impor- association with severity of ND (P = 0.016) (low tance of these findings lies in the notion that although versus high ND). A logistic regression model a contribution of CHRNB2 to ND may not be observed predicting high ND and taking into account the when this gene is analyzed separately, it may still be effect of background variables showed a significant involved in ND, due to dependency of a CHRNB2 contribution of two CHRNA7 SNPs, rs883473 and effect on specific variants in other genes (CHRNA4 or rs1909884, the CHRNA5 SNP rs680244, and the NTRK2). Conventional single-gene association analy- interaction between the CHRNA7 SNP rs2337980 sis may not be useful in the case of CHRNB2 and its and neuroticism. The model predicted 40% of the role in ND genetic would need to be analyzed jointly variance. No association of this gene with SI was with other genes.119 observed.81 In a report from a Swiss sample, no association of two dinucleotide polymorphisms, Other nAChR loci D15S1360 and D76630, with smoking status was Compared with the data regarding the four loci found.125 presented above, less evidence supports involvement CHRNA7, which is a notable candidate gene for of four other nAChR loci (CHRNA2, CHRNA7, schizophrenia,126 has been studied in the context of CHRNA9 and CHRNA10) in genetic susceptibility schizophrenia-related smoking (See Section to ND. ‘CHRNA4’). Association of the 113-bp allele of the D15S1360 marker in CHRNA7 (intron 2) with smok- CHRNA2. A few genome-wide linkage analysis ing risk was reported in a sample of 177 schizo- studies showed some degree of significant or phrenia patients from Canada.127 However, no suggestive evidence for the 8q21 region with peak interaction between smoking status and the promoter scores close to CHRNA2. In the SMOFAM sample of À86C/T SNP genotype was found by Zammit et al.128 607 individuals from 158 families, in which 738 among UK schizophrenia patients. autosomal, microsatellite markers were examined,120 the D8S258 marker was linked to tobacco CHRNA9 and 10. Suggestive linkage for ND at the dependence, using a phenotype based on DSM-IV 11p15 locus containing CHRNA10 was reported in (fulfillment of three or more out of seven ND criteria two independent genome-wide linkage studies.129,130 during lifetime) with maximal LOD score of 2.7. This Studying 142 individuals from 12 families (European linkage supports data from the Collaborative Study on American) of the Family Study of Panic disorder the Genetic of Alcoholism (COGA) that used two (FSPD) sample, Gelernter et al.129 genotyped 416 phenotypes (ever smoked and average number of microsatellite markers and detected linkage of packs per year) with both showing modest linkage to D11S4046 with habitual smoking (P = 0.0003). 8p22–23.121 CHRNA2 has also been implicated in a Affected persons were defined as those who had smoking intensity phenotype among schizophrenic smoked more than one pack per day for at least a year pedigrees of European American and AA origin in a or at least half pack per day for at least 10 years. In the linkage study using the permutation method. The Finnish twin families sample (FTF), encompassing highest peak LOD score of 2.34 in the CHRNA2 505 individuals from 153 families, the linkage containing 8p21–22 region (P-value 0.001) was shown findings of Gelernter et al.129 were supported for ND for the marker D8S258 in the EA sample (P-value of (FTND X4 scores). D11S4181 had a modest maximal 0.003 in the combined sample, with LOD score LOD score of 2.25.130 However, the linkage signal was of 1.91).122 in the telomeric region, raising questions concerning The CHRNA2 30 UTR SNP, rs2280376, was nomin- the reliability of the finding.130 ally associated with antisocial drug dependence The study of Greenbaum et al.81 is, to the best of our during adolescence, in a sample of 231 young knowledge, the only candidate-gene study that ana- Caucasian American males in treatment for drug lyzed the association of CHRNA9 and CHRNA10 with dependence and delinquency versus matched com- ND. CHRNA9 was nominally associated with ND munity controls.123 Analyzing 1500 SNPs distributed severity (intron 3 SNP rs4861065, P = 0.038, genotype) across 50 candidate genes, rs2280376 emerged with a but no positive results were shown for CHRNA10.As P-value of 0.002 after permutation. Another intronic mentioned in the Introduction section, these two CHRNA2 SNP, rs2565065, was nominally associated genes are expressed in the inner ear and not in the with smoking cessation, in a clinical trial of bupro- brain, making them less plausible for genetic studies. pion versus placebo for smoking cessation in an EA sample, using Bayesian analysis of a panel of 59 SNPs 124 Genetic manipulation of nAChR genes in mouse in 11 candidate genes. models CHRNA7. Although a large body of neurobiological Genetically modified animals are highly valuable evidence supports an important role for the CHRNA7 for identifying the physiological and behavioral role gene in ND, its genetic contribution to ND has not of individual neuronal nAChRs subtypes, particularly been extensively studied. In the Jewish, Israeli female in the absence of pharmacological agents with

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 929 specificity for many of the nAChR subtypes.131 Knock b2-subunit plays a major role in nicotine discrimina- in (KI) and knock out (KO) mutations in nAChR genes tive stimulus and taste aversion effects and these provide a useful research tool to evaluate the properties may affect psychological aspects of ND.139 particular contribution of specific receptor subunits Sensitivity to nicotine’s stimulus properties is sub- to ND susceptibility, including modulation of dopa- stantially impaired in b2-KO mice (compared with minergic transmission by the cholinergic sys- WT), in two experimental models: drug discrimina- tem.131,132 Recently, several review papers have tion and conditioned taste aversion. The effect is addressed physiological and behavioral phenotypes specific to nicotine, as no difference between WT and relevant to ND of nAChR gene manipulation in KO is detected with regard to morphine-involved mice.77,131,133 In this short section and the companion discrimination tests.139 Mice lacking the b2-subunit table (Table 4), we will briefly highlight some of the do not exhibit nicotine CPP (conditioned place findings most relevant to the current discussion but preference), a measure of nicotine reward.140 The not describe them in detail. effect of nicotine on acoustic startle response sensi- tivity was assessed in WT compared with b2-KO mice CHRNB2 and a significant reduction in startle amplitude after The involvement of the CHRNB2 subunit in the nicotine administration was detected in the KO reinforcing effect of nicotine was shown in two mice.141 elegant studies based on the notion that the reinfor- cing properties of nicotine are mediated by increasing CHRNB4 DA release and transmission in the mesolimbic DA The appearance of withdrawal symptoms after cessa- system. The elimination of the gene in KO tion of smoking is an important clinical effect of mice causes a significant decrease in DA release in nicotine. In the study of Salas et al.,72 KO mice the ventral striatum after nicotine stimulation as had decreased signs of nicotine withdrawal compared compared with wild-type (WT) animals.134 Consis- with WT. The mice were exposed to nicotine for 13 tently, intravenous self-administration of nicotine is days (through minipump delivery), and then with- attenuated in these mutant mice. This attenuation is drawal symptoms were induced by intraperitoneal specific to nicotine and there is no effect on the injection of the nicotinic receptor antagonist meca- reinforcing of cocaine in the mutant mice.134 Subse- mylamine. Somatic withdrawal symptoms (chewing, quent work by Maskos et al.135 confirmed the scratching, shaking and so on) were scored, and found importance of CHRNB2 for nicotine reinforcement. to be substantially decreased in the KO mice. The researchers used a lentivirus vector to re-express Hyperalgesia, a non-somatic nicotine withdrawal the chrnb2 gene in the VTA of b2-KO mice (the virus sign, was induced in WT, but not b4À/À mice.72 was stereotactically injected to this brain region). Re- By contrast, no differences in withdrawal symp- expression of the chrnb2 subunits caused recovery of toms were seen in chrnb2 KO mice compared with nicotine-elicited DA release (recorded using in vivo WT. Lack of physical nicotine withdrawal symptoms intracerebral microdialysis in awake, freely moving in chrnb2 KO mice was further shown by Besson mice) and nicotine self-administration, as well as et al.142 and Jackson et al.,143 providing robust restoration of cognitive function, measured by slow evidence for this finding. These results suggest that exploratory behavior in an open field.135 nAChRs subtypes involved in positive-reinforcing Interestingly, b2*-receptors in the VTA and the mechanisms of nicotine (such as b2-subunits) might substantia nigra are involved in different DA-depen- be different from those subtypes that contribute to its dent locomotion types (only exploratory behavior for negative-reinforcing effects (such as b4-subunits).72 VTA b2*, compared with a variety of locomotion b4-containing receptors are found in the autonomic phenotypes for nigral b2*), as observed when b2-gene nervous system, a fact that may explain their is rescued in chrnb2 KO mice.136 On the electro- involvement in autonomic-mediated somatic signs physiological level, spontaneous firing patterns of of nicotine withdrawal.131 VTA dopaminergic neurons are different among b2À/ À mice compared with WT, and re-expression of this CHRNA6 and CHRNB3 subunit restores the original firing patterns.137 The role of a6- and b3-subunits in the nicotinic The importance of the b2-subunit in stimulation of modulation of DA neurons and their effects on ND are DA by nicotine and its relevance to ND is demon- not completely understood. Both subunits colocalize strated by abolition of nicotine-mediated DA re- in striatal dopaminergic neurons.64 As mentioned sponses in b2-KO mice (such as DA-dependent before, the b3-subunit has an accessory role and locomotor activation, after repeated nicotine treat- cannot form an acetylcholine binding site. It was ment).138 This suggests that the b2-subunit plays a noticed that b3-subunit deletion greatly reduces a6* fundamental role in the sustained activation of the nAChR expression in the DA cell body (VTA) and DA system during chronic administration of nico- terminal region (striatum),32 while mRNA levels for tine.138 a6 or other subunits were not changed in these Other behavioral studies in chrnb2 KO mice areas in b3-KO mice.68 These findings strikingly provide further evidence for the importance of support the role of the b3-subunit in a6b3* receptor this subunit in nicotine-behavior phenotypes. The formation and targeting to DA nerve terminals.32

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 930 Table 4 nAChRs: genetically manipulated mouse models and reported phenotypes (main relative findings)

Gene Manipulation Phenotype References

CHRNB2 KO Decreased DA release in the striatum, after nicotine Picciotto et al.134 stimulation. Attenuation of nicotine Intravenous self-administration. KO Abolishment of locomotor response after nicotine King et al.138 administration KO Deficient in sensitivity to nicotine discrimination and Shoaib et al.139 conditioned taste aversion. KO Change of spontaneous firing pattern of dopaminergic VTA Mameli-Engvall et al.137 neurons KO Deletion of nicotine CPP (conditioned place preference) Walter et al.140 KO Significant decrease in acoustic startle response after Owens et al.141 nicotine administration KO No change in withdrawal symptoms compared with WT Besson et al.142 Reexpression of Recovery of nicotine-elicited dopamine release and nicotine Maskos et al.135 the gene in KO self-administration mice VTA Restoration of cognitive function Reexpression of Different dopamine dependent locomotion types in receptor Avale et al.136 the gene in KO restoration at VTA vs SN mice

CHRNA4 KI: single-point Receptor hypersensitive to nicotine Tapper et al.146 mutation (Leu90 Enhanced reinforcing and reward behavior to Ala) Altered nicotine tolerance and sensitization KO Striatal dopamine levels do not increase in response to Marubio et al.147 nicotine KO Morphological changes in nigrostriatal dopaminergic Parish et al.148 neurons, impaired dopamine transporter levels KO No systemic nicotine self-administration Pons et al.145 Mice strain Altered nicotine consumption and preference Butt et al.150 A529T polymorphism

CHRNB4 KO Decreased signs of nicotine withdrawal compared with WT Salas et al.72 KO Significantly decrease the sensitivity to nicotine-induced Salas et al.157; Kedmi seizures and hypolocomotion et al.229

CHRNB3 KO Alternation DA relapse in the striatum following nicotine Cui et al, 200368 stimulation KO Altered locomotor activity and prepulse inhibition (PPI) of Cui et al.68 acoustic startle response

CHRNA3 Heterozygous a3 Partially resistant to nicotine-induced seizures Salas et al.157 þ /À Normal sensitivity to the hypolocomotor effects of nicotine

CHRNA6 KO No systemic nicotine self-administration Pons et al.145 KI: single-point Increased DA neuron firing in VTA; hyperactive behavior Drenan et al.132 mutation (Leu90 to Ser)

CHRNA7 KO Decreased somatic withdrawal symptoms but normal Salas et al.155 nicotine tolerance KO Involvement in physical nicotine withdrawal signs (somatic Jackson et al.143 signs and hyperalgesia)

CHRNA5 KO Less sensitive to nicotine-induced seizure and locomotor Salas et al.158 activity KO Reduction in some physical aspects of nicotine withdrawal Jackson et al.143 (but not affective withdrawal signs)

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 931 Nicotine-related behavior has not been assayed in Leu90Ala mice developed a significant place prefer- KO mice but analysis in these mutant mice ence for nicotine. Knock-in mice were much more showed that a sub-population of nAChRs, which both likely to develop tolerance compared with WT, contains the b3-subunit and is sensitive to inhibition assayed by nicotine-induced hypothermia with by the snail toxin a-conotoxin MII (a-CtxMII), reg- chronic nicotine infusion.146 ulates striatal DA neurotransmission following Similar to b2-KO mice, a4-KO exhibit striatal DA nicotine stimulation.68 a-CtxMII is a toxin which levels that do not increase in response to nicotine, inhibits nicotine-induced DA release in rat striatal supporting the notion that a4b2* nAChRs are neces- synaptosomes.144 Deletion of the b3-subunit altered sary for DA release.147 Morphological changes in locomotor activity and prepulse inhibition of acoustic nigrostriatal dopaminergic neurons in drug-naive startle response, both partially regulated by DA adult a4(À/À)-mice have been described; KO mice transmission, supporting a role for b3-containing had significantly larger terminal arbors than WT.148 In nAChRs in modulating these behaviors.68 addition, KO mice had impaired DA transporter Champtiaux et al.62 did not find morphological levels and impairment in locomotor activity in alternation in DA pathways in KO mice. Using response to DA agonist and antagonist.148 autoradiography, they found a complete disappear- The mouse strain dependent A529 polymorphism ance of a-CtxMII binding in a6À/À mice, claiming in the a4-subunit modulates nicotine intake among that the a6-subunit is an essential component of the different strains. This functional polymorphism (ala- native-binding site of this toxin.62 nine to threonine residue exchange at position 529, A recent study has shown the central role of both located in the second intracellular loop), which was VTA a6 and VTA a4-subunit in acute nicotine shown to influence nicotine-induced 86Rb þ efflux, reinforcement.145 chrna6 KO and KO nico- plays a pivotal role in nicotine but not ethanol, tine-naive mice did not display self-administration of consumption and preference.149,150 This effect most nicotine intravenously (at pharmacological doses), likely occurs through changes in the sensitivity of a4- whereas WT mice did. When the chrna6 and chrna4 containing nAChR to nicotine, possibly due to change genes were re-expressed (using lentiviral vector) in of the ratio between high and low affinity a4b2*- the a6À/À VTA and a4À/À VTA (respectively), the receptors.151 reinforcement property of nicotine was restored.145 This important study confirms that CHRNA6 and CHRNA4 (as well as CHRNB2) expression in the VTA is both necessary and sufficient for self-administra- CHRNA7 tion of nicotine.145 Despite the widespread expression of a7-subunit Drenan et al.132 used genetically manipulated a6- containing receptors in the brain, studies in KO mice nicotine-hypersensitive mice (substitution of Leu have not identified a role for this receptor in nicotine with Ser on the 90 residue in the M2 domain) to show reward and conditioning.77,131 a7-KO studies have the role of this subunit in modulating DA transmis- failed to support a role for this subunit in sensitivity sion. The hypersensitive a6-mice showed increased to the behavioral effects of nicotine. a7À/À nicotine- DA neuron firing in the VTA compared with WT. naive mice do not display different IV self-adminis- Moreover, administration of a selective-a6 agonist tration of nicotine (model of acute reinforcement) stimulated greater striatal DA release in the mutated than WT,145 and a7-KO mice did not show weakened mice. On the behavioral level, KI animals were responses to nicotine in a drug discrimination hyperactive both in the home cage and when experiment nor difference from WT in sensitivity to introduced to a novel environment, consistent with the locomotor depressant effects of nicotine.152,153 The increased DA release.132 Taken together, these obser- development of nicotine tolerance did not differ vation point to specific involvement of the a6* between a7-KO and WT mice.154 Somatic withdrawal receptor in regulation of DA release. signs (after intraperitoneal injection of nAChR an- tagonist to induce nicotinic withdrawal) are reduced CHRNA4 in KO mice relative to WT littermates.155 In addition to the role of the CHRNA4 subunit in the However, other reports suggest that a7-subunit in- nicotine self-administration model of acute reinforce- volvement in withdrawal is restricted to somatic signs ment described above,145 involvement of the a4- and does not affect other withdrawal symptoms such nAChR subunit in nicotine-induced reward, toler- as affective ones,77,143 whereas others do not support ance, and sensitization was demonstrated in engi- the role of this subunit in somatic withdrawal signs neered mutant.146 Single-point mutation in the pore- either.156 This discrepancy may result from differ- forming M2 domain of the a4-subunit-encoding gene ences in treatment regimens and protocols.155 Some (Leu90 to Ala90) was introduced by the researchers. As caution is required in interpreting these results. The a result, the a4-subunit containing receptors of this relative absence of nicotine-behavior phenotypes in knock-in (KI) mouse became hypersensitive to nico- a7-KO mice does not necessarily mean lack of a tine.146 At nicotine doses 50-fold less than that biological role for the a7-subunit in these phenotypes; required in WT mice to elicit reinforcing (measured such a role may be compensated by other nAChR by CPP test) and reward behavior, the mutated subunits, such as a4 and b2.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 932 Other receptors section, we point out possible mechanism by which A relatively small number of studies have been neuronal nAChR variants may influence the pharma- reported in the literature regarding nicotine-related cological profile of the subunits and of the receptor behaviors of other nAChR genetically manipulated subtypes containing them. mice. a3À/À mice suffer from high perinatal mortality which impairs research on nicotine effects in geneti- cally manipulated mice totally lacking this subunit. General overview Compared with WT, heterozygous a3 þ /À mice were Receptor desensitization. A well-documented partially resistant to nicotine-induced seizures but characteristic of nAChRs is that they are reversibly manifested normal sensitivity to the hypolocomotor desensitized when chronically exposed to an agonist. effects of nicotine (possibly due to partial expression 157 Desensitization is a temporary decrease or loss of of this subunit in the heterozygous mice). a5À/À the biological reaction of nAChRs after continuous mice lacking , the gene cluster neighbor of or repetitive exposure to nicotinic agonist , reach adulthood and behave normally under stimulation.21,34 In vitro electrophysiological measure- basal conditions, but display lower sensitivity to 158 ments have shown that prolonged acetylcholine or short-term effects of nicotine than WT mice. nicotine application (in a time scale of minutes) Resistance of a5À/À mice to nicotine was observed produce a progressive decline of the currents carried in a range of nicotine doses (both low and high doses), by nAChRs.160–162 Several models of desensitization and affected seizure sensitivity as well as several have been proposed, suggesting that this process other behavioral effects, especially those concerned represents an allosteric protein behavior.46 According with locomotor activity. Involvement of the a5- to the classical Katz and Thesleff163 model, nAChRs subunit in some physical aspects of nicotine with- exist in three states: resting, activated and drawal (but not affective nicotine withdrawal signs) 143 desensitized. Agonist binds to the resting receptor has been reported. Treatment of nicotine-depen- (usually with low-receptor affinity), and activates it. dent a5-KO mice with a nicotinic antagonist (meca- Then the receptor is desensitized (with high agonist mylamine) caused reduction in somatic signs affinity), and subsequently recovers. Recovery from compared with WT, whereas anxiety-related behavior desensitization is agonist independent.164 Different was still present. a5-subunit containing receptors are nAChRs have different probabilities of desensiti- found in the PNS (including ganglionic a5b4*- zation in response to varying concentrations of receptors), and may play a role in mediating these 143 nicotine. This fact may have significance for ND due somatic withdrawal signs. to the changing concentrations of nicotine resulting from cigarette smoking throughout the day.16 Molecular function of neuronal nAChRs in ND: pharmacological and biochemical aspects nAChR upregulation. After chronic nicotine A wide range of neuropsychopharmacological data exposure, high-affinity agonist binding to nicotinic support the notion that ND involves more than one receptors is increased (particularly the a4b2*- specific nicotinic receptor subtype (for recent com- subtype) in both animal models and humans. This prehensive reviews see: Rose16; Picciotto et al.).159 process is termed nicotine-induced upregulation22,165 This supports our hypothesis that different variants in and may be involved in the pathophysiology of ND. multiple neuronal nAChR subtypes may mediate This striking effect of chronic nicotine exposure was genetic vulnerability to ND. The addiction process is shown in postmortem radioligand-binding studies complicated, and depends on many factors including performed in the brains of smokers and nonsmokers. pharmacological characteristics of the receptor such An increase in [3H]-nicotine- and [3H]-Ach-binding as differential affinity for nicotine, subunit composi- sites was observed in the brains of smokers, and was tion and stoichiometry.16 shown to be dependent on nicotine concentration.166 Three fundamental features of the cholinergic Moreover, numerous studies have shown that long- receptor system are discussed here briefly: differential term exposure to nicotine (or other nAChR ligands) receptor affinity for nicotine, receptor activation- causes an increase in the total amount of brain desensitization cycle and nicotinic receptor upregula- labeling by [3H]-nicotine.167 Actually, nAChR tion after chronic nicotine exposure or withdrawal. upregulation has been reported in several different These pharmacological traits have a crucial influence in vivo and in vitro systems,21 after various nicotine- on ND-related behavior. As receptor desensitization exposure regimens and in many brain areas, such as (decreased response to agonist) and upregulation after the striatum, cortex, midbrain and hippocampus.168 chronic nicotine exposure (with potential for in- The mechanism of nAChR upregulation, probably creased response to agonist) act in opposite direc- involves increase in receptor number, rather than tions, these mechanisms seem compensatory to each augmentation of receptor affinity for nicotine.21,169 other.159 So far only very limited research has been Several mechanisms has been proposed to carried out regarding the possible influence of explain it, such as accelerated receptor transition identified genetic variants in nAChR-encoding genes from the endoplasmic reticulum to the surface on these pharmacological traits. In the following membrane.170

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 933 nAChR activation-desensitization cycle and ND tors on glutamatergic afferents remain active and In considering the possible contribution of receptor enhance glutamate excitation of the DA neurons. activation and desensitization to the rewarding effect It is neurobiologically reasonable that genetic of nicotine in the brain, the attention of most variants in a4b2* and a7* nAChRs may directly and researchers has traditionally focused on a4b2*-recep- indirectly influence the affinity of the receptor tors, which comprise the majority of nicotinic containing the subunit for ligand activation-desensi- receptor sites in the brain, and bind nicotine with tization cycle and/or other pharmacological traits of high affinity.18 However, the nicotine concentrations interest. Hutchison et al.112 assessed the influence of present throughout the day in the brains of smokers the CHRNA4 30 UTR SNP, rs22236196, on a4b2 causes saturation and desensitization of these recep- binding in human postmortem brain tissue. Indivi- tors at an early stage, and they are not available for duals carrying the TC genotype for this SNP showed nicotine action for most of the day.16 Using the highly greater binding in the NAc than carriers of the TT selective 2-FA radioligand enabling visualization of genotype.112 Bierut et al.88 studied the influence of the brain a4b2*-nAChRs with positron emission tomo- amino acid change caused by CHRNA5 rs16969968 on graphy in humans, Brody et al.47 showed that daily the maximal response to nicotinic agonist. They smokers have nearly complete saturation of brain found that a4b2a5 receptors with the aspartic acid a4b2*-receptors throughout the day. Moreover, one to variants (D398) had more than two times greater two puffs of a cigarette resulted in 50% occupancy of maximal response to nicotine than a4b2a5 receptors a4b2*-nAChRs for 3.1 h after smoking. Smoking a full with asparagine.88The implication is that the de- cigarette (or more than one) resulted in more than creased nAChR function is associated with the 88% brain receptor occupancy, Thus, according to increased risk for ND. However, there were no this neuroimaging study, 96–98% of a4b2*-nAChRs differences in expression pattern between the two 88 172 in the brains of nicotine-dependent smokers are isoforms. Mao et al. estimated that (a4b2)2a5 occupied during the day. These findings show that receptors constitute 11–37% of the total a4b2* in nicotine-dependent smokers, an almost complete nAChR population, depending on the brain region. shift of a4b2*-receptors to the high-affinity desensi- Accordingly, the potential influence of the tization state takes place during the day.47 According a4b2a5N398/a4b2a5D398 variant on total brain to Benowitz et al.,8 the nicotine concentration in the a4b2* receptor response to nicotine is probably blood of chronic smokers during the day is in the substantial. range of 60–300 nM. As even low levels of nicotine Other genetic variants in the CHRNA5–CHRNA3– exposure results in substantial occupancy and desen- CHRNB4 cluster associated with ND phenotypes may sitization of brain a4b2*-nAChRs, it is plausible that also have functional role due to influence on receptor high-affinity nicotine binding to a4b2*-receptors is response to nicotine, ligand affinity and activation- not in itself a sufficient mechanism to explain desensitization cycle. As discussed in the Section continued smoking in nicotine-addicted individuals ‘Pharmacological properties of neuronal nAChRs and throughout the day.16,171 their involvement in human disorders’, b4-expression On the other hand, nAChRs with low affinity for in the VTA is low, but b4-containing receptors are nicotine are not susceptible to rapid saturation, and found in the habenula, locus coeruleus, IPN and can continue functioning and responding throughout cerebellum.58,71,173 a3b4* have lower-binding affinity the day. Thus, selective desensitization of some than do a4b2* receptors in transfected cells,174 they nAChR subtypes, while maintaining the activity and do not become fully occupied and saturated at the functionality of others may play a significant role in nicotine concentration range detected in smokers’ ND. One important receptor subtype to consider in plasma during the day,16 and are more resistant to this context is a7, also widely expressed in the brain. rapid desensitization than are a4b2* receptors.175 a7*-containing receptors in the VTA/SNc are much Taking receptor affinity, saturation and differential less susceptible than a4b2* to desensitization at the desensitization considerations into account, Rose16 daily blood and brain nicotine concentrations of suggested that b4-subunit containing receptors con- smokers.171 Up to 500 nM concentration of nicotine tribute to the reinforcing effect of nicotine. Further (beyond the nicotine concentration range of chronic research on the pharmacological importance of ND- smokers, as discussed above) causes very little associated variants (or variants tagged by them) in the desensitization of VTA/SNc a7*-nAChR currents.171 CHRNA5–CHRNA3–CHRNB4 cluster is clearly war- As mentioned in Section ‘nAChR subunits’, DA ranted. release in the NAc is enhanced by glutamatergic excitatory drive from the laterodorsal tegmentum and Multiple nAChR upregulation and ND pedunculopontine tegmentum. The glutamatergic afferents reaching VTA DA neurons express presy- a4b2* receptor upregulation: possible contribution to naptic a7* nicotinic receptors, which have a rela- ND and withdrawal. Buisson and Bertrand169 tively low-nicotine affinity, and are not highly reported that human a4b2 nAChRs expressed in desensitized by the nicotine concentrations achieved HEK-293 cells are upregulated after chronic nicotine by smokers.9 So, while nicotine concentrations are exposure, with an increase of up to 70% of the high elevated and a4b2* receptor desensitized, a7* recep- affinity receptor fraction. Several in vivo imaging

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 934 studies of a4b2* receptor upregulation have been in vivo. In humans, evaluation of the relevance of performed in humans, using highly selective SNPS to upregulation can be studied in postmortem radioligands combined with positron emission brain tissue from smokers but also in living volun- tomography scans. Wullner et al.176 used 2-FA- teers using positron emission tomography and SPECT positron emission tomography to evaluate a4b2 radioligands. nAChRs in smokers and nonsmokers. The total brain distribution volume of the radioligand in vivo was Upregulation of non-a4b2*-subunit containing significantly increased in the brains of smokers, receptors. Recent studies show that not only a4b2- especially in the cerebellum and brainstem. Staley containing receptors, but also other neuronal nAChR et al.177 imaged b2* nAChRs in human smokers using are upregulated after nicotine exposure. The subunit [123I]5-IA SPECT, and compared receptor availability patterns of upregulation differ from one another in during early abstinence with corresponding values in terms of time course and required nicotine nonsmokers. As expected, the densities of brain b2* concentration.165 In vitro studies in cells transfected nAChR were higher in the striatum, cerebellum and with homomeric a7 and various heteromeric a3b2, cerebral cortex of smokers during early abstinence, a6b2 and a3b4 receptors show upregulation induced compared with nonsmokers. Moreover, there was a by nicotine, but data from in vivo studies is more correlation between somatosensory cortex b2* nAChR variable and complex.26 Upregulation of a7-subunit availability and the urge to smoke.177 Upregulation containing nAChRs after chronic nicotine treatment following smoking cessation is time dependent, and 3 has been reported in several brain regions of the weeks later nAChRs were downregulated to the levels mouse, including striatum, midbrain and cortex.180,181 of nonsmokers.178 Upregulation of human a7-subunits was shown in The possible contribution of nAChR upregulation human embryonic kidney (HEK) cells after chronic to the pathophysiology of ND is not fully understood, exposure to nicotine and other receptor ligands.182 but is believed to be mediated by midbrain DA Data on a6* receptors upregulation is less clear. On release. Vezina et al.168 have proposed that nicotine the one hand, some in vivo reports support exposure produces temporary nAChR upregulation, upregulation of a6-containing receptors in rats leading to augmentation of midbrain dopaminergic during long-term self-administration of nicotine.183 neuron excitation and the induction of long-term Although substantially higher nicotine potentiation. concentrations are required to upregulate a6b2 and Another hypothesis regarding the role of upregula- a3b2 receptors compared with the a4b2* receptor, tion in ND is that upregulation is a compensatory this high level of nicotine may be reached during the response of the cholinergic system following desensi- fast fluctuation of nicotine in smokers’ serum, tization of neuronal nAChRs after chronic nicotine immediately after smoking a cigarette. This implies exposure.159,161 The increase in receptor number that different phases of nicotine intake during the day enables the cholinergic system to counteract the involve different receptor subtypes upregulation.165 chronic deficits caused by the initial rapid desensi- On the other hand, other papers describe no change or tization due to chronic agonist exposure.179 The even downregulation of this subunit. McCallum increase in nAChR number may be the underlying et al.184 did not observe upregulation of striatal a3*/ mechanism of nicotine tolerance and ND.21 a6* receptor subunits, possible due to differences in As explained in the Section ‘nAChR upregulation’, nicotine exposure regimens and concentrations. Perry the upregulation process seems to result from an et al.185 reported downregulation of a6* receptors increase in a4b2* receptor numbers expressed on the number and function in the striatum. membrane, although the exact mechanism is not yet The a6- and b3-subunit upregulation is of special clear.159 Genetic variants within these two subunits interest because these subunits are expressed in midbrain may influence upregulation processes; of special DA neurons, and modulate DA release.186,187 It is possible importance are variants in the UTRs of the genes that some regulatory variants within the CHRNA6– (such as CHRNB2 30 UTR rs2072660 and rs2072661or CHRNB3 gene cluster are associated with upregulation CHRNA4 30 UTR rs2236196) and promoters, which and others with downregulation of the subunits. have a regulatory role in gene expression. For As mentioned in the Section ‘nAChR activation- example, as described in Section ‘CHRNA4’, the desensitization cycle and ND’, (a4b2)2a5 receptors CHRNA4 50 promoter rs6122429 and CHRNA4 30 constitute 11–37% of the total a4b2* nAChRs.172 UTR rs2236196 affect luciferase construct activity, a However, while a4b2* nAChR upregulate after marker of gene expression.112 In the construct con- chronic nicotine administration, it seems that the taining the G-allele, expression was fivefold greater particular (a4b2)2a5 subtype does not. (a4b2)2a5 than the A-allele construct.112 Different baseline resistance to nicotine-induced upregulation probably levels of gene and protein expression, as a result of has implications for the mechanism of ND, due to the genetic variants in regulatory regions, may have a suggested contribution of a4b2* nAChR upregulation profound impact on the upregulation potential of to ND.172 As mentioned, a5 is an accessory subunit, receptors after nicotine exposure. To our knowledge, and some of the variants described in the CHRNA5– no correlation has been sought between specific SNPs CHRNA3–CHRNB4 cluster may be involved in the and upregulation patterns, either in cell cultures or upregulation resistance mechanisms.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 935 Finally, important reservations should be noted ing behavior. In a study of serotonergic candidate regarding non-a4b2 subtype upregulation. Consis- genes in our sample of young Israeli women,204 SNPs tently, higher nicotine levels are required for upregu- in five serotonin receptor genes (HTR1A, HTR1B, lation of these receptors compared with the a4b2* HTR2A, HTR2C and HTR6) and the 5-HT transporter- subtype. These higher concentrations differ from the linked polymorphic region (5-HTTLPR) were geno- physiological accumulation of nicotine levels in typed. In a logistic regression model for SI, the smokers’ serum and brain, which rises slowly and is personality trait, novelty seeking was a significant relatively sustained during the day, after hours of risk factor. A five SNP CACCC haplotype in HTR6 was continued smoking.21,165 Further study is warranted to a strong protective factor against SI (OR = 0.26; determine the functional role of upregulation in the P = 0.007). The interaction of HTR6-C276T genotype addiction process in humans, and the possible and lifetime traumatic experience contributed influence of genetic variants within nAChRs genes strongly to the risk of SI (OR = 13.88, P = 0.0001). on it. Similar results were obtained for ND in a study of dopaminergic candidate genes in the same sample of 205 nAChR genes, personality traits, neurocognition young Israeli women. Novelty seeking was a risk “ and ND factor (OR = 1.07, P = 0.03); the DAT1_E15 274— DAT1_VNTR G-9 haplotype (OR = 0.37, P = 0.001) Although genetic factors play a significant role in ND- and the interaction between trauma and a related behavior, other factors such as demographic DAT1_E15“274—DAT1_VNTR C-9 haplotype and family background, psychopathology, life experi- (OR = 0.15, P = 0.01) were protective.205 ence, personality traits and cognitive function are A potential interaction between nAChR-encoding pivotally involved in this complex trait whether genes and cognitive function in susceptibility to ND is directly or interactively.81,188,189 Before dependence of great interest because of earlier reports on the effect on nicotine is established, factors such as these (as of nicotine on cognitive processes and the association well as genetic factors) influence the likelihood of of cigarette smoking with specific cognitive attri- initiating smoking (SI) which is a prerequisite for ND butes.206,207 It has been suggested that smoking may be to develop. A further complexity that must be taken used as a form of self-medication by some individuals into account is that personality, psychopathology and because of beneficial acute effects on cognitive cognitive function are themselves influenced by functioning, mainly in the attention domain.208,209 A genetic factors.190,191 It is thus clear that complex large body of evidence supports the involvement of models are required to effectively study the role of neuronal nAChRs in cognitive functions. For exam- genes and environment in ND-related behaviors. A ple, selective-a7 agonists show efficacy in learning comprehensive evaluation of this subject is beyond and memory tasks in animals40 and a role of the a4b2- the scope of this review. However, we will address subunit in cognition is supported by KO and general concepts that are important and findings that antagonist studies.20 Thus, genetic variations in illustrate the concepts, as summarized in Supple- nAChRs that influence neurocognitive functions mentary Table 2. may affect smoking behavior through this mediation, It is well established that the tobacco use is as discussed above for personality traits. associated with personality traits such as neuroti- cism, impulsivity, extraversion, novelty seeking, Attention, impulsivity, nAChRs and susceptibility harm avoidance and antisociality.192–195 Given the to ND well-established role of genetics in personality there Attention deficit/hyperactivity disorder (ADHD) is is growing interest in identifying specific genetic characterized by attentional, hyperactive and impul- variants that may be implicated in personality sivity symptom clusters that manifest as cognitive traits.190,196 An early, much studied and highly deficits on tests of attention and executive functio- debated finding is the association of novelty seeking, n.20.Epidemiological reports show a high prevalence with a functional variant in the DA D4 receptor of smoking among individuals diagnosed with ADHD gene.197,198 Similar findings have implicated genetic compared with nonsmokers. According to Lambert variants in personality traits such as harm avoid- et al.,210 46% of adolescents with ADHD were daily ance199 and neuroticism.200 Thus, although genetic smokers by the age of 17 years compared with 24% of variants may influence ND directly, they may also do age-matched controls. The finding continues into so through their influence on personality traits that adulthood, the figures being 35 vs 16%, respec- are related to cigarette smoking and use of other tively.210 substances.191,201 For example, interaction between Accumulating evidence suggests that cholinergic functional variants in the promoter of the serotonin system dysregulation contributes to the pathophysiol- receptor gene (5-HTTLPR-L/5-HTTLPR-S) and neuro- ogy of ADHD, based on animal, pharmacological and ticism was found to influence smoking behavior.202,203 imaging studies.20,211,212 There are several reports Identifying the genetic contribution of nAChR- linking genetic findings in nAChR-encoding genes encoding genes to personality traits could shed to ADHD. Notwithstanding differences in phenotype further light on the complex interaction between definition, association of several SNPs in CHRNA4 genes, environment and personality factors in smok- has been reported: rs7267923 by Brookes et al.;213

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 936 rs2273505, rs2273506 and rs6090384 (survives multi- cism was not significant in itself. It is of interest that ple testing correction) by Todd et al.214 with replica- Greenbaum et al.81 found a strong contribution of tion of rs2273505 by Lee et al.;211 intron 1 background and psychological factors to SI with a dinucleotide repeat215 and rs3787141.211 Some of relatively small contribution of variations in nAChRs these findings were not replicated by other groups genes. On the other hand, the contribution of back- (for example, Bobb et al.216). Although inconsistent, ground and psychological characteristics to severity the findings can be interpreted as supporting involve- of ND was smaller but that of nAChR genetic variants ment of the CHRNA4 gene in ADHD. Impulsivity, a larger.81 central feature of ADHD, is a recognized personality Further support for the influence of nAChRs on trait associated with smoking behavior.193,194 Thus, neuroticism comes from genetically manipulated genetic variants in CHRNA4 may contribute to ND- mice. The behavior of b4À/À mice differs from WT related behavior both directly and as risk factors for (b4 þ / þ ) littermates, when examined on behavioral ADHD. Interestingly, the synonymous CHRNA4 SNP test related to anxiety.70 b4-KO mice showed in- RS1044396 was found to be associated with inter- creased exploratory behavior and increased climbing individual differences in visuospatial attention.217 activity,70 interpreted as decreased anxiety-like beha- Furthermore, although not modulating working mem- vior. Taken together, these findings point to the ory by itself, this variant was involved in the possibility that CHRNB4 involvement in ND is at interaction of attention and working memory.218 least partially mediated by the effect of this nAChR Recent work by our group suggests that young subunit on anxiety. In another relevant study, point women who smoke have a neurocognitive profile that mutation of a single amino acid in the a4-subunit M2 is characterized by small but significant impairments transmembrane region (leucine-to-serine) generated of sustained attention and response inhibition.209 We nicotine hypersensitive a4-receptors.222 These mice suggest that the improvement by nicotine of cognitive displayed increased level of anxiety (mostly in the function in these domains may predispose young elevated plus maze), as well as other characteristic women who initiate cigarette smoking to maintain the such as poor motor learning and excessive locomotion habit.209 Subsequently, Rigbi et al.189 showed that the activity, that was eliminated by low doses of nicotine performance on the relevant test measures (for injection.222 example, response inhibition) is modified by specific The role of the a4-subunit in anxiety in a stressful variants in nAChR genes, mostly in CHRNA4, setting was assessed by Ross et al.223 a4-KO mice CHRNA7 and CHRNB3 but also in other nAChR show behavioral features consistent with heightened genes.189 basal levels of anxiety and increased exploratory activity in stressful settings,223 consistent with the nAChR genes, neuroticism and ND reports of Labarca et al.222 Finally, as with deletion of A connection between neuroticism and smoking is the a4 and b4-subunit, chrnb3 KO mice showed well established, both in adults and adoles- altered anxiety-like behavior, when compared with cents.188,194,219 Neuroticism is associated with mani- WT, assessed by a series of behavioral tests.224 festations of anxiety and depression.220 Goodwin and Thus, in addition to their direct influence on ND- Hamilton201 suggested a connection between cigarette related phenotypes, neuronal nAChRs genes may smoking, panic attacks and neuroticism. In the mediate predisposition to neuroticism, a personality genetic context, as discussed above, the serotonin trait that is a susceptibility factor for smoking. transporter polymorphism has been reported to interact with neuroticism, to influence several ND- Conclusions related phenotypes.202,203 Nicotinic cholinergic receptors modulate the re- In this review, we have surveyed several lines of lease of neurotransmitters that have critical roles in evidence that support the involvement of genetic the regulation of anxiety (DA and serotonin).18 In variation in different neuronal nAChRs subunits in humans, according to smokers’ accounts, smoking ND. Findings from three genetic research approaches may have anxiolytic effects, which are likely attribu- (linkage analysis, candidate-gene association studies table to the nicotine contained in cigarettes.221 Taking and GWAS) analyzing various ND phenotypes with into account the psychological characteristics and different assessment tools support the involvement of genetic variants, controlling for the background several nAChR genes in ND. By far, the strongest and variables and using a best-fitting logistic regression most compelling evidence is for association of model of SI, Greenbaum et al.81 found that a 5 SNP CHRNA5–CHRNA3–CHRNB4 gene cluster variants haplotypes in CHRNB2, neuroticism and novelty with several ND-related phenotypes, as studied seeking were significantly associated with SI extensively in the last 2 years. Mouse models with (P = 5.9 Â 10À14, Nagelkerke r2 = 0.30). The model to genetic manipulation of nAChR genes display beha- predict high ND as compared with low ND vioral phenotypes related to ND, show the role of the (P = 2.24 Â 10À7, Nagelkerke r2 = 0.40), included two cholinergic system in regulating dopaminergic trans- SNPs in CHRNA7 (rs1909884 and rs883473), one in mission and provide a major tool to evaluate the CHRNA5 (rs680244) and the interaction of CHRNA7 contribution of each subunit to ND susceptibility. SNP rs2337980 with neuroticism, although neuroti- Pharmacological and biochemical properties of

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 937 receptor subunits, such as affinity for nicotine, (although these surprising findings may be explained desensitization and receptor upregulation, differ by compensation of other subunits). However, rela- among subtypes and functionally relevant genetic tively modest data support the contribution of these variation in these receptors is likely to play a role in three genes to ND relevant personality traits, such as differential susceptibility to ND. Various nicotinic impulsivity (CHRNA4,) neuroticism (CHRNB2, receptors are expressed on mesolimbic DA neurons CHRNA7) and to neurocognitive function (CHRNA4, and are believed to modulate the reinforcing proper- CHRNA7). ties of nicotine. Finally, neuronal nAChRs variants are associated with personality traits (such as neuroti- Possible missing links cism and impulsivity) and neurocognitive functions, which may contribute to ND. Taken together, these We suggest that the a3, a5-6 and b3-4 nAChR subunit- data suggest that genetic susceptibility to ND is linked encoding genes play a much more pivotal role in ND to several nAChR subtype genes and variants in each than hitherto appreciated and that functional, most subunit gene may have independent as well as likely regulatory variants within them may contribute interactive contributions to ND at the molecular level. differentially to ND-related behavior (see Table 5). On the basis of previous neurobiological and Although neurobiological knowledge of the physio- pharmacological knowledge, CHRNA4 and CHRNB2 logical role of the receptors containing these subunits are highly plausible candidate genes for ND. a4b2* in ND is relatively limited compared with available receptors are widely expressed in the brain and bind data regarding a4b2* and a7* receptors, recent reports nicotine with high affinity.22 a4b2* nAChRs are support their critical role as modulators and acces- upregulated in the brains of smokers.166 A partial sory subunits in nAChRs. Moreover, impressive, agonist of this receptor, varenicline, is a newly robust genetic associations with ND phenotypes have approved smoking cessation drug. However, contin- been reported. As neurobiological research advances, ued nicotine consumption during the day by heavy the complexity of nAChR composition is increasingly smokers is difficult to explain, due to the rapid recognized. In fact, the number of neurobiologically saturation and desensitization of the receptor.16 relevant receptor subtypes (with distinct biophysical Studies in genetically manipulated mice clearly point and pharmacological properties) is larger than that to an essential role of both genes in ND (especially estimated previously.18 For example, current knowl- positive reinforcement), although reports are some- edge indicates that DA nerve terminals in the striatum what more impressive for CHRNB2 than CHRNA4. express as many as five different nAChR subtypes: However, genetic finding do not correlate with the a4b2, a4a5b2, a4a6b2b3, a6b2b3 and a6b2.226 This central pharmacological role of these two subunits. functional diversity of nAChR receptors in terms of Modest support for the involvement of the CHRNA4 composition and role is highly influenced by the a3, gene in ND is provided by relatively small candidate- a5–6 and b3–4 receptor subunits, not paid much of gene association studies and linkage studies, but not attention until recently (Table 6). by large scale GWASs (See Table 3). For the CHRNB2 As presented in Table 3, there is considerable gene, most candidate-gene studies fail to show support for the association of the CHRNA5– association with ND-related phenotypes. This sur- CHRNA3–CHRNB4 gene cluster with ND, based on prising contrast between solid, well-established neu- GWAS and case–control study results, mainly from robiological knowledge of ND mechanisms, receptor EA samples, encompassing thousands of participants pharmacological traits (high nicotine binding and and using several measures of ND. Particularly upregulation) and hypothesis supporting experiments convincing associations have been reported for the in mouse models on the one hand, and modest non-synonymous, CHRNA5 exon 6 SNP, rs16969968, genetic findings as well as rapid receptor saturation and the CHRNA3 exon 5 SNP, rs1051730. Another and desensitization on the other, supports the notion variant, not correlated to the above two SNPs, the of a potential contribution of genetic variation in CHRNA3 30 UTR SNP, rs578776, is independently additional neuronal nAChR subtypes to ND suscept- associated with ND. The fact that ‘hypothesis-free’ ibility. findings of GWAS and the biologically plausible, The findings regarding CHRNA7 are somewhat candidate-gene approach converge in this case is similar. As for CHRNA4 and CHRNB2, neurobiologi- striking. cal and pharmacological data support a major in- The accessory a5-subunit modulates biophysical volvement in ND. This receptor is expressed in most and pharmacological properties of nAChRs CNS regions, including VTA and striatum. It binds (like calcium permeability or desensitization nicotine with low affinity, is not desensitized in the rate).27,172,227,228 Indeed, variation in rs16969968 in- brains of smokers during the day (in contrast to fluences the maximal response of a4b2a5 receptors to a4b2*),171 and upregulates after chronic nicotine nicotine.88 Although a5-KO mice are viable, so far administration.225 However, both genetic findings only limited research on ND-related features has been and experiments in mouse models do not support a reported, including that mutant mice are less sensi- substantial link to ND. Furthermore, a7-KO mice do tive to nicotine-induced seizure and locomotor not display ND-related behavior, with the exception activity158,229 than WT, and show a reduction in of some aspects of nicotine withdrawal symptoms physical withdrawal symptoms143 (See Table 4).

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 938 Table 5 Brain nAChRs: key pharmacological data that may be applicable in terms of possible influence of specific genetic variants on ND

a4b2-containing nAChRs *High nicotine affinity, rapid saturation.47 *Desensitization at nicotine concentration daily range, found in chronic smokers.16 *Nicotine receptor upregulation, after chronic nicotine admission and in the early abstinence.169,176 * CHRNA4 30 UTR SNP rs22236196 influences a4b2 binding in human postmortem brain tissue.112

a7-containing nAChRs *Lower nicotine affinity than a4b2.171 *Significantly less susceptible to desensitization at nicotine concentration daily range, found in chronic smokers.9 *Nicotine receptor upregulation, after chronic nicotine admission.180,181

a5-containing receptor *a4b2a5N398/a4b2a5D398 Variants (refers to rs16969968) alters response to nicotine agonist.88 *Assembly of a5toa4b2-containing receptor, prevents receptor upregulation.172

a6-containing receptor a6* receptors response to chronic nicotine administration is not clear: some papers reports Upregulation,165 whereas other papers report no change184 or even downregulation.185

b4-containing nAChRs *Less prone to receptor occupancy and desensitization, compared with a4b2* receptors.16

Table 6 Summary of possible contribution of nACnR genes to ND

Genetic Mouse model Pharmacological traits Personality traits/neurocognition

CHRNA2 þþ þ þ CHRNA3 þþþ þ þ þ CHRNA4 þþ þþþ þþþ þþ CHRNA5 þþþ þ þþ þ CHRNA6 þþ þþ þþ þ CHRNA7 þ þ þþþ þþ CHRNB2 þ þþþ þþþ þþ CHRNB3 þþ þþ þ þ CHRNB4 þþ þþ þþ þþ

‘ þ ’ —no or little evidence support possible involvement in contribution to ND-related phenotypes. ‘ þþ’ —medium support for possible involvement in contribution to ND-related phenotypes. ‘ þþþ’ —convincing evidence for possible involvement in contribution to ND-related phenotypes.

The nearby a3-gene, located within a very high LD The CHRNA6–CHRNB3 gene cluster, is, in our view block, is also expressed in VTA and other ND relevant an important candidate for ND susceptibility. a6- and brain regions, and this receptor subtype is upregu- b3-subunits colocalize in DA neurons, and both are lated after nicotine exposure.165 a3b4* receptors have expressed in the VTA. b3 (such as a5) is an auxiliary lower binding affinity than a4b2* receptors in subunit, assembles with other neuronal nAChR a- and transfected cells.174 Unfortunately, homozygous a3- b-subunits, and is important for the correct assembly, KO mice are not viable, complicating the assessment stability and transport of a6* nAChRs in dopaminer- of the involvement of this receptor subtype in ND. gic neurons.26,32 Knockout mice for CHRNB4, the third member of Another nAChRs sub-population, a6b2* receptors, this gene cluster, mainly show alternations in with- are located in the striatum, and mediate DA release.64 drawal symptoms, and anxiety-like behavior,72 corre- Interestingly, both a6- and b3-subunits are present in lating with the high expression of this receptor in the a4a6b2b3 receptors on dopaminergic neurons in the PNS. This may suggest CHRNB4 involvement in striatum and NAc. According to some papers, this neuroticism, a personality trait related to ND. The receptor is characterized by possessing the highest b4-subunit does not become saturated at the levels of affinity for nicotine among all nicotinic receptor nicotine found in smokers blood16 and is less subtypes reported until now, and may play an susceptible than a4b2* receptors to the process of important role in regulating DA-related behaviors, rapid desensitization.175 rendering it relevant to the reinforcing effects of

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 939 nicotine.30,132,226 Genetic support for the involvement upregulation. As mentioned in Section ‘a4b2* recep- of the CHRNA6–CHRNB3 cluster in ND has been tor upregulation: possible contribution to ND and reported in one GWAS and a large (B300) candidate- withdrawal’, combining neuroimaging and genetic gene association study on the same sample,76,78 and approaches for evaluating these aspects should be by several other candidate-gene studies analyzing ND fruitful. and other smoking-related phenotypes.90,108,109 Data Three independent GWASs have reported associa- from KO mice are supplementary: chrna6 KO nico- tion of the CHRNA5–CHRNA3–CHRNB4 cluster with tine-naive mice do not display self-administration of LC.80,100,101 Some of these studies suggest that this nicotine (in contrast to WT), but when the chrna6 gene cluster does not confer susceptibility to LC by an gene is re-expressed in the a6À/À VTA, the reinforce- effect on susceptibility to ND but directly,100,101 ment property of nicotine is restored.145 KI mice although another does support this mechanism for model of hypersensitive a6-receptor shows increased LC and peripheral vascular disease.80 Functional DA neurons firing in VTA, an hyperactive behavior, studies of the significantly associated SNPs in non- compared with WT.132 b3-KO mice show altered DA neuronal tissues (lungs and peripheral arteries) are a release in the striatum after nicotine stimulation, high priority. altered locomotor activity and prepulse inhibition of The contribution of particular SNPs to some aspects acoustic startle response.68 of ND can be evaluated in humans by assessing behavioral sensitivity to nicotine, as has been carried Pivotal issues for future research out by a few groups.108,112,115 At the level of animal The data presented here support our hypothesis that models, much behavioral research has been carried multiple neuronal nAChR subunits contribute to out on KO mice, mostly CHRNA4, CHRNA7, CHRNB2 genetic susceptibility to ND. However, many ques- and CHRNB4. Further studies of CHRNA2-3, tions are still open, and much research is required to CHRNA5-6 and CHRNB3 genetically modified mouse at least partially answer some of them. Future models are required. KI mouse models of some research should encompass all nAChRs and assess interesting SNPs (such as rs16969968) should be the independent and interactive role of each subunit created and analyzed. in ND-related phenotypes. Ultimately, the full ‘cata- In conclusion, ND is a complex trait that involves log’ of ND genes may contain all the neuronal not only genetic factors, but also demographic and nAChRs-encoding genes, each responsible for a familial factors, life experience, personality traits and particular intermediate phenotype (reinforcement, neurocognitive factors. The design of future studies withdrawal, tolerance, sensitization and so on). and sample recruitment protocols should take this Gene–gene interactions may also be important, and into consideration, and assess at these factors. Models it is possible that the genetic contribution of some predicting ND-related behavior should include these nAChRs to ND may mainly be as ‘modifier genes’, variables in addition to genetic ones. Moreover, the mediating the genetic contribution of other nAChR direct contribution of variants within nAChR genes to genes, as well as non-nAChR genes. To achieve these personality traits (such as neuroticism and impulsiv- goals, uniform definitions of ND-related phenotypes ity) and neurocognitive function associated with ND are required. As shown in Table 2, the variable should be studied separately. The answers to these definitions currently used by different research and many other open questions will be achieved by groups complicate replication studies and lead to using a multidisciplinary approach that combines inconsistent results. genetics, pharmacology, biochemistry, psychology Some nAChR-encoding genes, especially the and genetically manipulated animal models. The CHRNA5–CHRNA3–CHRNB4 gene cluster, were importance of such efforts for smoking prevention found to be associated not only with ND but also and for the development of new treatments for with addiction to additional substances of abuse such smoking cessation is obvious. Nevertheless, it should as alcohol and cocaine.94,97 These findings suggest the be noted that the research in the field of ND should existence of a general pathophysiological role for not be limited to nAChR genes; other promising nAChRs in addiction behavior. Other subunit genes candidate genes exist and others will no doubt should be similarly assessed. emerge. Further research is also required to understand the functional role of variants associated with ND. Two Conflict of interest pivotal studies have been reported so far revealing an effect of the CHRNA5 SNP, rs16969968, on maximal The authors declare no conflict of interest. receptor response to nicotine,88 and the CHRNA4 30 UTR SNP, rs2236196, on expression of the gene and References on a4b2 binding in postmortem brain tissue.112 Further studies at the neurobiological level on other 1 Annual smoking-attributable mortality, years of potential life lost, SNPs that have been associated with ND is essential, and productivity losses—United States, 1997–2001. MMWR Morb not only in terms of gene and protein expression but Mortal Wkly Rep 2005; 54: 625–628. 2 Benowitz NL. Neurobiology of nicotine addiction: implications also in the context of influence on receptor ligand- for smoking cessation treatment. Am J Med 2008; 121(4 Suppl 1): binding affinity, activation-desensitization cycle and S3–10.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 940 3 Li MD. Identifying susceptibility loci for nicotine dependence: 28 Arneric SP, Holladay M, Williams M. Neuronal nicotinic 2008 update based on recent genome-wide linkage analyses. Hum receptors: a perspective on two decades of drug discovery Genet 2008; 123: 119–131. research. Biochem Pharmacol 2007; 74: 1092–1101. 4 State-specific prevalence of current cigarette smoking among 29 Briggs CA, Gubbins EJ, Marks MJ, Putman CB, Thimmapaya R, adults and secondhand smoke rules and policies in homes and Meyer MD et al. Untranslated region-dependent exclusive workplaces—United States, 2005. MMWR Morb Mortal Wkly Rep expression of high-sensitivity subforms of alpha4beta2 and 2006; 55: 1148–1151. alpha3beta2 nicotinic acetylcholine receptors. Mol Pharmacol 5 Benowitz NL. Clinical pharmacology of nicotine: implications for 2006; 70: 227–240. understanding, preventing, and treating tobacco addiction. Clin 30 Drenan RM, Nashmi R, Imoukhuede P, Just H, McKinney S, Pharmacol Ther 2008; 83: 531–541. Lester HA. Subcellular trafficking, pentameric assembly, and 6 Carmines EL. Evaluation of the potential effects of ingredients subunit stoichiometry of neuronal nicotinic acetylcholine recep- added to cigarettes. Part 1: cigarette design, testing approach, and tors containing fluorescently labeled alpha6 and beta3 subunits. review of results. Food Chem Toxicol 2002; 40: 77–91. Mol Pharmacol 2008; 73: 27–41. 7 Baker RR, Pereira da Silva JR, Smith G. The effect of tobacco 31 Le Novere N, Corringer PJ, Changeux JP. The diversity of ingredients on smoke chemistry. Part I: Flavourings and addi- subunit composition in nAChRs: evolutionary origins, physiolo- tives. Food Chem Toxicol 2004; 42(Suppl): S3–37. gic and pharmacologic consequences. J Neurobiol 2002; 53: 8 Benowitz NL. Nicotine addiction. Prim Care 1999; 26: 611–631. 447–456. 9 Dani JA, Harris RA. Nicotine addiction and comorbidity with 32 Gotti C, Moretti M, Clementi F, Riganti L, McIntosh JM, Collins alcohol abuse and mental illness. Nat Neurosci 2005; 8: 1465–1470. AC et al. Expression of nigrostriatal alpha 6-containing nicotinic 10 Hukkanen J, Jacob III P, Benowitz NL. Metabolism and disposi- acetylcholine receptors is selectively reduced, but not elimi- tion kinetics of nicotine. Pharmacol Rev 2005; 57: 79–115. nated, by beta 3 subunit gene deletion. Mol Pharmacol 2005; 67: 11 Batra V, Patkar AA, Berrettini WH, Weinstein SP, Leone FT. The 2007–2015. genetic determinants of smoking. Chest 2003; 123: 1730–1739. 33 Laviolette SR, van der Kooy D. The neurobiology of nicotine 12 Sullivan PF, Kendler KS. The genetic epidemiology of smoking. addiction: bridging the gap from molecules to behaviour. Nat Rev Nicotine Tob Res 1999; 1(Suppl 2): S51–S57;, discussion S69–70. Neurosci 2004; 5: 55–65. 13 Li MD, Cheng R, Ma JZ, Swan GE. A meta-analysis of estimated 34 Buisson B, Bertrand D. Nicotine addiction: the possible role genetic and environmental effects on smoking behavior in male of functional upregulation. Trends Pharmacol Sci 2002; 23: and female adult twins. Addiction 2003; 98: 23–31. 130–136. 14 Lessov-Schlaggar CN, Pergadia ML, Khroyan TV, Swan GE. 35 Helbig I, Mefford HC, Sharp AJ, Guipponi M, Fichera M, Franke Genetics of nicotine dependence and pharmacotherapy. Biochem A et al. 15q13.3 microdeletions increase risk of idiopathic Pharmacol 2008; 75: 178–195. generalized epilepsy. Nat Genet 2009; 41: 160–162. 15 Uhl GR, Liu QR, Drgon T, Johnson C, Walther D, Rose JE. 36 Martin-Ruiz CM, Haroutunian VH, Long P, Young AH, Davis KL, Molecular genetics of nicotine dependence and abstinence: Perry EK et al. Dementia rating and nicotinic receptor expression whole genome association using 520,000 SNPs. BMC Genet in the prefrontal cortex in schizophrenia. Biol Psychiatry 2003; 2007; 8:10. 54: 1222–1233. 16 Rose JE. Multiple brain pathways and receptors underlying 37 Court J, Martin-Ruiz C, Piggott M, Spurden D, Griffiths M, Perry tobacco addiction. Biochem Pharmacol 2007; 74: 1263–1270. E. Nicotinic receptor abnormalities in Alzheimer’s disease. Biol 17 Jorenby DE, Hays JT, Rigotti NA, Azoulay S, Watsky EJ, Williams Psychiatry 2001; 49: 175–184. KE et al. Efficacy of varenicline, an alpha4beta2 nicotinic 38 Lee M, Martin-Ruiz C, Graham A, Court J, Jaros E, Perry R et al. acetylcholine receptor partial agonist, vs placebo or sustained- Nicotinic receptor abnormalities in the cerebellar cortex in release bupropion for smoking cessation: a randomized con- autism. Brain 2002; 125(Part 7): 1483–1495. trolled trial. JAMA 2006; 296: 56–63. 39 Gotti C, Moretti M, Bohr I, Ziabreva I, Vailati S, Longhi R et al. 18 Gotti C, Zoli M, Clementi F. Brain nicotinic acetylcholine Selective nicotinic acetylcholine receptor subunit deficits iden- receptors: native subtypes and their relevance. Trends Pharmacol tified in Alzheimer’s disease, Parkinson’s disease and dementia Sci 2006; 27: 482–491. with Lewy bodies by immunoprecipitation. Neurobiol Dis 2006; 19 Kalamida D, Poulas K, Avramopoulou V, Fostieri E, Lagoumintzis 23: 481–489. G, Lazaridis K et al. Muscle and neuronal nicotinic acetylcholine 40 Olincy A, Stevens KE. Treating schizophrenia symptoms with an receptors. Structure, function and pathogenicity. FEBS J 2007; alpha7 nicotinic agonist, from mice to men. Biochem Pharmacol 274: 3799–3845. 2007; 74: 1192–1201. 20 Wilens TE, Decker MW. Neuronal nicotinic receptor agonists for 41 Nestler EJ. Is there a common molecular pathway for addiction? the treatment of attention-deficit/hyperactivity disorder: focus on Nat Neurosci 2005; 8: 1445–1449. cognition. Biochem Pharmacol 2007; 74: 1212–1223. 42 Mansvelder HD, McGehee DS. Cellular and synaptic mechanisms 21 Ochoa EL, Lasalde-Dominicci J. Cognitive deficits in schizo- of nicotine addiction. J Neurobiol 2002; 53: 606–617. phrenia: focus on neuronal nicotinic acetylcholine receptors and 43 Klink R, de Kerchove d’Exaerde A, Zoli M, Changeux JP. smoking. Cell Mol Neurobiol 2007; 27: 609–639. Molecular and physiological diversity of nicotinic acetylcholine 22 Hogg RC, Raggenbass M, Bertrand D. Nicotinic acetylcholine receptors in the midbrain dopaminergic nuclei. J Neurosci 2001; receptors: from structure to brain function. Rev Physiol Biochem 21: 1452–1463. Pharmacol 2003; 147: 1–46. 44 Azam L, Winzer-Serhan UH, Chen Y, Leslie FM. Expression of 23 Solda G, Boi S, Duga S, Fornasari D, Benfante R, Malcovati M neuronal nicotinic acetylcholine receptor subunit mRNAs et al. In vivo RNA–RNA duplexes from human alpha3 and within midbrain dopamine neurons. J Comp Neurol 2002; 444: alpha5 nicotinic receptor subunit mRNAs. Gene 2005; 345: 260–274. 155–164. 45 Everitt BJ, Robbins TW. Neural systems of reinforcement for drug 24 Gault J, Robinson M, Berger R, Drebing C, Logel J, Hopkins J et al. addiction: from actions to habits to compulsion. Nat Neurosci Genomic organization and partial duplication of the human 2005; 8: 1481–1489. alpha7 neuronal nicotinic acetylcholine receptor gene 46 Wang H, Sun X. Desensitized nicotinic receptors in brain. Brain (CHRNA7). Genomics 1998; 52: 173–185. Res Brain Res Rev 2005; 48: 420–437. 25 Gotti C, Clementi F. Neuronal nicotinic receptors: from structure 47 Brody AL, Mandelkern MA, London ED, Olmstead RE, Farahi J, to pathology. Prog Neurobiol 2004; 74: 363–396. Scheibal D et al. Cigarette smoking saturates brain alpha 4 beta 2 26 Gaimarri A, Moretti M, Riganti L, Zanardi A, Clementi F, Gotti C. nicotinic acetylcholine receptors. Arch Gen Psychiatry 2006; 63: Regulation of neuronal nicotinic receptor traffic and expression. 907–915. Brain Res Rev 2007; 55: 134–143. 48 Quik M, Polonskaya Y, Gillespie A, Jakowec M, Lloyd GK, 27 Kuryatov A, Onksen J, Lindstrom J. Roles of accessory subunits Langston JW. Localization of nicotinic receptor subunit mRNAs in alpha4beta2(*) nicotinic receptors. Mol Pharmacol 2008; 74: in monkey brain by in situ hybridization. J Comp Neurol 2000; 132–143. 425: 58–69.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 941 49 Conejero-Goldberg C, Davies P, Ulloa L. Alpha7 nicotinic conotoxin MII-binding nicotinic acetylcholine receptors that acetylcholine receptor: a link between inflammation and neuro- modulate dopamine release and related behaviors. J Neurosci degeneration. Neurosci Biobehav Rev 2008; 32: 693–706. 2003; 23: 11045–11053. 50 Palma E, Maggi L, Barabino B, Eusebi F, Ballivet M. Nicotinic 69 Gotti C, Moretti M, Zanardi A, Gaimarri A, Champtiaux N, acetylcholine receptors assembled from the alpha7 and beta3 Changeux JP et al. Heterogeneity and selective targeting of subunits. J Biol Chem 1999; 274: 18335–18340. neuronal nicotinic acetylcholine receptor (nAChR) subtypes 51 Charpantier E, Barneoud P, Moser P, Besnard F, Sgard F. Nicotinic expressed on retinal afferents of the superior colliculus and acetylcholine subunit mRNA expression in dopaminergic neu- lateral geniculate nucleus: identification of a new native nAChR rons of the rat substantia nigra and ventral tegmental area. subtype alpha3beta2(alpha5 or beta3) enriched in retinocollicular Neuroreport 1998; 9: 3097–3101. afferents. Mol Pharmacol 2005; 68: 1162–1171. 52 Elliott KJ, Jones JM, Sacaan AI, Lloyd GK, Corey-Naeve J. 6- 70 Salas R, Pieri F, Fung B, Dani JA, De Biasi M. Altered anxiety- hydroxydopamine lesion of rat nigrostriatal dopaminergic neu- related responses in mutant mice lacking the beta4 subunit of the rons differentially affects nicotinic acetylcholine receptor sub- nicotinic receptor. J Neurosci 2003; Jul: 16, ; 23: 6255-6263. unit mRNA expression. J Mol Neurosci 1998; 10: 251–260. 71 Quick MW, Ceballos RM, Kasten M, McIntosh JM, Lester RA. 53 Ishii K, Wong JK, Sumikawa K. Comparison of alpha2 nicotinic Alpha3beta4 subunit-containing nicotinic receptors dominate acetylcholine receptor subunit mRNA expression in the function in rat medial habenula neurons. Neuropharmacology central nervous system of rats and mice. J Comp Neurol 2005; 1999; 38: 769–783. 493: 241–260. 72 Salas R, Pieri F, De Biasi M. Decreased signs of nicotine 54 Quik M, Vailati S, Bordia T, Kulak JM, Fan H, McIntosh JM et al. withdrawal in mice null for the beta4 nicotinic acetylcholine Subunit composition of nicotinic receptors in monkey striatum: receptor subunit. J Neurosci 2004; 24: 10035–10039. effect of treatments with 1-methyl-4-phenyl-1,2,3,6-tetrahydro- 73 Gomez-Casati ME, Fuchs PA, Elgoyhen AB, Katz E. Biophysical pyridine or L-DOPA. Mol Pharmacol 2005; 67: 32–41. and pharmacological characterization of nicotinic cholinergic 55 Xu W, Gelber S, Orr-Urtreger A, Armstrong D, Lewis RA, Ou CN receptors in rat cochlear inner hair cells. J Physiol 2005; 566 et al. Megacystis, mydriasis, and defect in mice (Part 1): 103–118. lacking the alpha3 neuronal nicotinic acetylcholine receptor.. 74 Lips KS, Pfeil U, Kummer W. Coexpression of alpha 9 and alpha Proc Natl Acad Sci USA 1999; 96: 5746–5751. 10 nicotinic acetylcholine receptors in rat dorsal root ganglion 56 Perry DC, Xiao Y, Nguyen HN, Musachio JL, Davila-Garcia MI, neurons. Neuroscience 2002; 115: 1–5. Kellar KJ. Measuring nicotinic receptors with characteristics of 75 Feng Y, Niu T, Xing H, Xu X, Chen C, Peng S et al. A common alpha4beta2, alpha3beta2 and alpha3beta4 subtypes in rat tissues haplotype of the nicotine acetylcholine receptor alpha 4 subunit by autoradiography. J Neurochem 2002; 82: 468–481. gene is associated with vulnerability to nicotine addiction in 57 Gahring LC, Persiyanov K, Dunn D, Weiss R, Meyer EL, Rogers men. Am J Hum Genet 2004; 75: 112–121. SW. Mouse strain-specific nicotinic acetylcholine receptor 76 Saccone SF, Hinrichs AL, Saccone NL, Chase GA, Konvicka K, expression by inhibitory interneurons and astrocytes in the Madden PA et al. Cholinergic nicotinic receptor genes implicated dorsal hippocampus. J Comp Neurol 2004; 468: 334–346. in a nicotine dependence association study targeting 348 candidate 58 Turner JR, Kellar KJ. Nicotinic cholinergic receptors in the rat genes with 3713 SNPs. Hum Mol Genet 2007; 16:36–49. cerebellum: multiple heteromeric subtypes. J Neurosci 2005; 25: 77 Portugal GS, Gould TJ. Genetic variability in nicotinic 9258–9265. acetylcholine receptors and nicotine addiction: converging 59 Flora A, Schulz R, Benfante R, Battaglioli E, Terzano S, Clementi evidence from human and animal research. Behav Brain Res F et al. Neuronal and extraneuronal expression and regulation of 2008; 193: 1–16. the human alpha5 nicotinic receptor subunit gene. J Neurochem 78 Bierut LJ, Madden PA, Breslau N, Johnson EO, Hatsukami D, 2000; 75: 18–27. Pomerleau OF et al. Novel genes identified in a high-density 60 Zoli M, Moretti M, Zanardi A, McIntosh JM, Clementi F, Gotti C. genome wide association study for nicotine dependence. Hum Identification of the nicotinic receptor subtypes expressed on Mol Genet 2007; 16: 24–35. dopaminergic terminals in the rat striatum. J Neurosci 2002; 22: 79 Berrettini W, Yuan X, Tozzi F, Song K, Francks C, Chilcoat H et al. 8785–8789. Alpha-5/alpha-3 nicotinic receptor subunit alleles increase risk 61 Salminen O, Murphy KL, McIntosh JM, Drago J, Marks MJ, for heavy smoking. Mol Psychiatry 2008; 13: 368–373. Collins AC et al. Subunit composition and pharmacology of two 80 Thorgeirsson TE, Geller F, Sulem P, Rafnar T, Wiste A, classes of striatal presynaptic nicotinic acetylcholine receptors Magnusson KP et al. A variant associated with nicotine mediating dopamine release in mice. Mol Pharmacol 2004; 65: dependence, lung cancer and peripheral arterial disease. Nature 1526–1535. 2008; 452: 638–642. 62 Champtiaux N, Han ZY, Bessis A, Rossi FM, Zoli M, Marubio L 81 Greenbaum L, Kanyas K, Karni O, Merbl Y, Olender T, Horowitz A et al. Distribution and pharmacology of alpha 6-containing et al. Why do young women smoke? I. Direct and interactive nicotinic acetylcholine receptors analyzed with mutant mice. effects of environment, psychological characteristics and J Neurosci 2002; 22: 1208–1217. nicotinic cholinergic receptor genes. Mol Psychiatry 2006; 11: 63 Moretti M, Vailati S, Zoli M, Lippi G, Riganti L, Longhi R et al. 312–322, 223. Nicotinic acetylcholine receptor subtypes expression during rat 82 Weiss RB, Baker TB, Cannon DS, von Niederhausern A, Dunn retina development and their regulation by visual experience. DM, Matsunami N et al. A candidate gene approach identifies the Mol Pharmacol 2004; 66: 85–96. CHRNA5-A3-B4 region as a risk factor for age-dependent nicotine 64 Champtiaux N, Gotti C, Cordero-Erausquin M, David DJ, addiction. PLoS Genet 2008; Jul: 4, : e1000125. Przybylski C, Lena C et al. Subunit composition of functional 83 Schlaepfer IR, Hoft NR, Collins AC, Corley RP, Hewitt JK, nicotinic receptors in dopaminergic neurons investigated with Hopfer CJ et al. The CHRNA5/A3/B4 gene cluster variability as knock-out mice. J Neurosci 2003; 23: 7820–7829. an important determinant of early alcohol and tobacco initiation 65 Gotti C, Moretti M, Gaimarri A, Zanardi A, Clementi F, Zoli M. in young adults. Biol Psychiatry 2008; 63: 1039–1046. Heterogeneity and complexity of native brain nicotinic receptors. 84 Fagerstrom KO. Measuring degree of physical dependence to Biochem Pharmacol 2007; 74: 1102–1111. tobacco smoking with reference to individualization of treatment. 66 Salminen O, Drapeau JA, McIntosh JM, Collins AC, Marks MJ, Addict Behav 1978; 3: 235–241. Grady SR. Pharmacology of alpha-conotoxin MII-sensitive sub- 85 Heatherton TF, Kozlowski LT, Frecker RC, Fagerstrom KO. types of nicotinic acetylcholine receptors isolated by breeding of The Fagerstrom Test for Nicotine Dependence: a revision of null mutant mice. Mol Pharmacol 2007; 71: 1563–1571. the Fagerstrom Tolerance Questionnaire. Br J Addict 1991; 86: 67 Le Novere N, Zoli M, Changeux JP. Neuronal nicotinic receptor 1119–1127. alpha 6 subunit mRNA is selectively concentrated in catecholami- 86 Rustin TA. Assessing nicotine dependence. Am Fam Physician nergic nuclei of the rat brain. Eur J Neurosci 1996; 8: 2428–2439. 2000; 62: 579–584, 591–572. 68 Cui C, Booker TK, Allen RS, Grady SR, Whiteaker P, Marks MJ 87 Li MD, Beuten J, Ma JZ, Payne TJ, Lou XY, Garcia V et al. Ethnic- et al. The beta3 nicotinic receptor subunit: a component of alpha- and gender-specific association of the nicotinic acetylcholine

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 942 receptor alpha4 subunit gene (CHRNA4) with nicotine depen- 107 Wacholder S, Chatterjee N, Caporaso N. Intermediacy and gene- dence. Hum Mol Genet 2005; 14: 1211–1219. environment interaction: the example of CHRNA5-A3 region, 88 Bierut LJ, Stitzel JA, Wang JC, Hinrichs AL, Grucza RA, Xuei X smoking, nicotine dependence, and lung cancer. J Natl Cancer et al. Variants in nicotinic receptors and risk for nicotine Inst 2008; 100: 1488–1491. dependence. Am J Psychiatry 2008; 165: 1163–1171. 108 Zeiger JS, Haberstick BC, Schlaepfer I, Collins AC, Corley RP, 89 Sherva R, Wilhelmsen K, Pomerleau CS, Chasse SA, Rice JP, Crowley TJ et al. The neuronal nicotinic receptor subunit genes Snedecor SM et al. Association of a single nucleotide poly- (CHRNA6 and CHRNB3) are associated with subjective responses morphism in neuronal acetylcholine receptor subunit alpha 5 to tobacco. Hum Mol Genet 2008; 17: 724–734. (CHRNA5) with smoking status and with ‘pleasurable buzz’ 109 Hoft NR, Corley RP, McQueen MB, Schlaepfer IR, Huizinga D, during early experimentation with smoking. Addiction 2008; 103: Ehringer MA. Genetic association of the CHRNA6 and CHRNB3 1544–1552. genes with tobacco dependence in a nationally representative 90 Stevens VL, Bierut LJ, Talbot JT, Wang JC, Sun J, Hinrichs AL sample. Neuropsychopharmacology 2009; 34: 698–706. et al. Nicotinic receptor gene variants influence susceptibility to 110 Saccone SF, Pergadia ML, Loukola A, Broms U, Montgomery GW, heavy smoking. Cancer Epidemiol Biomarkers Prev 2008; 17: Wang JC et al. Genetic linkage to chromosome 22q12 for a heavy- 3517–3525. smoking quantitative trait in two independent samples. Am J 91 Breitling LP, Dahmen N, Mittelstrass K, Illig T, Rujescu D, Raum E Hum Genet 2007; 80: 856–866. et al. Smoking cessation and variations in nicotinic acetylcholine 111 Wang D, Ma JZ, Li MD. Mapping and verification of susceptibility receptor subunits alpha-5, alpha-3, and beta-4 genes. Biol loci for smoking quantity using permutation linkage analysis. Psychiatry 2009; 65: 691–695. Pharmacogenomics J 2005; 5: 166–172. 92 Chassin L, Presson CC, Rose JS, Sherman SJ. The natural history 112 Hutchison KE, Allen DL, Filbey FM, Jepson C, Lerman C, of cigarette smoking from adolescence to adulthood: demo- Benowitz NL et al. CHRNA4 and tobacco dependence: from gene graphic predictors of continuity and change. Health Psychol regulation to treatment outcome. Arch Gen Psychiatry 2007; 64: 1996; 15: 478–484. 1078–1086. 93 Lando HA, Thai DT, Murray DM, Robinson LA, Jeffery RW, 113 Voineskos S, De Luca V, Mensah A, Vincent JB, Potapova N, Sherwood NE et al. Age of initiation, smoking patterns, and risk Kennedy JL. Association of alpha4beta2 nicotinic receptor and in a population of working adults. Prev Med 1999; 29(6 Part 1): heavy smoking in schizophrenia. J Psychiatry Neurosci 2007; 32: 590–598. 412–416. 94 Wang JC, Grucza R, Cruchaga C, Hinrichs AL, Bertelsen S, Budde 114 Adler LE, Olincy A, Waldo M, Harris JG, Griffith J, Stevens K JP et al. Genetic variation in the CHRNA5 gene affects mRNA et al. Schizophrenia, sensory gating, and nicotinic receptors. levels and is associated with risk for alcohol dependence. Mol Schizophr Bull 1998; 24: 189–202. Psychiatry 2008; 14: 501–510. 115 Ehringer MA, Clegg HV, Collins AC, Corley RP, Crowley T, Hewitt 95 Joslyn G, Brush G, Robertson M, Smith TL, Kalmijn J, Schuckit M JK et al. Association of the neuronal nicotinic receptor beta2 et al. Chromosome 15q25.1 genetic markers associated with level subunit gene (CHRNB2) with subjective responses to alcohol and of response to alcohol in humans. Proc Natl Acad Sci USA 2008; nicotine. Am J Med Genet B Neuropsychiatr Genet 2007; 144B: 105: 20368–20373. 596–604. 96 Chen X, Chen J, Williamson VS, An SS, Hettema JM, Aggen SH et 116 Conti DV, Lee W, Li D, Liu J, Van Den Berg D, Thomas PD et al. al. Variants in nicotinic acetylcholine receptors alpha5 and Nicotinic acetylcholine receptor beta2 subunit gene implicated in alpha3 increase risks to nicotine dependence. Am J Med Genet a systems-based candidate gene study of smoking cessation. Hum B Neuropsychiatr Genet, e-pub ahead of print; 8 January 2009. Mol Genet 2008; 17: 2834–2848. 97 Grucza RA, Wang JC, Stitzel JA, Hinrichs AL, Saccone SF, 117 Silverman MA, Neale MC, Sullivan PF, Harris-Kerr C, Wormley Saccone NL et al. A risk allele for nicotine dependence in B, Sadek H et al. Haplotypes of four novel single nucleotide CHRNA5 is a protective allele for cocaine dependence. Biol polymorphisms in the nicotinic acetylcholine receptor beta2- Psychiatry 2008; 64: 922–929. subunit (CHRNB2) gene show no association with smoking 98 Toh CK. The changing epidemiology of lung cancer. Methods Mol initiation or nicotine dependence. Am J Med Genet 2000; 96: Biol 2009; 472: 397–411. 646–653. 99 Witschi H. A short history of lung cancer. Toxicol Sci 2001; Nov: 118 Lueders KK, Hu S, McHugh L, Myakishev MV, Sirota LA, 64, : 4-6. Hamer DH. Genetic and functional analysis of single 100 Hung RJ, McKay JD, Gaborieau V, Boffetta P, Hashibe M, Zaridze nucleotide polymorphisms in the beta2-neuronal nicotinic D et al. A susceptibility locus for lung cancer maps to nicotinic acetylcholine receptor gene (CHRNB2). Nicotine Tob Res 2002; acetylcholine receptor subunit genes on 15q25. Nature 2008; 452: 4: 115–125. 633–637. 119 Li MD, Lou XY, Chen G, Ma JZ, Elston RC. Gene-gene interactions 101 Amos CI, Wu X, Broderick P, Gorlov IP, Gu J, Eisen T et al. among CHRNA4, CHRNB2, BDNF, and NTRK2 in nicotine Genome-wide association scan of tag SNPs identifies a dependence. Biol Psychiatry 2008; 64: 951–957. susceptibility locus for lung cancer at 15q25.1. Nat Genet 2008; 120 Swan GE, Hops H, Wilhelmsen KC, Lessov-Schlaggar CN, Cheng 40: 616–622. LS, Hudmon KS et al. A genome-wide screen for nicotine 102 Liu P, Vikis HG, Wang D, Lu Y, Wang Y, Schwartz AG et al. dependence susceptibility loci. Am J Med Genet B Neuropsy- Familial aggregation of common sequence variants on 15q24-25.1 chiatr Genet 2006; 141B: 354–360. in lung cancer. J Natl Cancer Inst 2008; 100: 1326–1330. 121 Bergen AW, Korczak JF, Weissbecker KA, Goldstein AM. A 103 Spitz MR, Amos CI, Dong Q, Lin J, Wu X. The CHRNA5-A3 region genome-wide search for loci contributing to smoking and on chromosome 15q24-25.1 is a risk factor both for nicotine alcoholism. Genet Epidemiol 1999; 17(Suppl 1): S55–S60. dependence and for lung cancer. J Natl Cancer Inst 2008; 100: 122 Faraone SV, Su J, Taylor L, Wilcox M, Van Eerdewegh P, Tsuang 1552–1556. MT. A novel permutation testing method implicates sixteen 104 Maus AD, Pereira EF, Karachunski PI, Horton RM, Navaneetham nicotinic acetylcholine receptor genes as risk factors for smoking D, Macklin K et al. Human and rodent bronchial epithelial cells in schizophrenia families. Hum Hered 2004; 57: 59–68. express functional nicotinic acetylcholine receptors. Mol Phar- 123 Corley RP, Zeiger JS, Crowley T, Ehringer MA, Hewitt JK, Hopfer macol 1998; 54: 779–788. CJ et al. Association of candidate genes with antisocial 105 Lam DC, Girard L, Ramirez R, Chau WS, Suen WS, Sheridan S drug dependence in adolescents. Drug Alcohol Depend 2008; et al. Expression of nicotinic acetylcholine receptor subunit 96: 90–98. genes in non-small-cell lung cancer reveals differences between 124 Heitjan DF, Guo M, Ray R, Wileyto EP, Epstein LH, Lerman C. smokers and nonsmokers. Cancer Res 2007; 67: 4638–4647. Identification of pharmacogenetic markers in smoking cessation 106 Volkow N, Rutter J, Pollock JD, Shurtleff D, Baler R. One SNP therapy. Am J Med Genet B Neuropsychiatr Genet 2008; 147B: linked to two diseases-addiction and cancer: a double whammy? 712–719. Nicotine addiction and lung cancer susceptibility. Mol Psychiatry 125 Stassen HH, Bridler R, Hell D, Weisbrod M, Scharfetter C. 2008; 13: 990–992. Ethnicity-independent genetic basis of functional psychoses: a

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 943 genotype-to-phenotype approach. Am J Med Genet B Neuropsy- 144 Kulak JM, Nguyen TA, Olivera BM, McIntosh JM. Alpha- chiatr Genet 2004; 124B: 101–112. conotoxin MII blocks nicotine-stimulated dopamine release in 126 Sanders AR, Duan J, Levinson DF, Shi J, He D, Hou C et al. No rat striatal synaptosomes. J Neurosci 1997; 17: 5263–5270. significant association of 14 candidate genes with schizophrenia 145 Pons S, Fattore L, Cossu G, Tolu S, Porcu E, McIntosh JM et al. in a large European ancestry sample: implications for psychiatric Crucial role of alpha4 and alpha6 nicotinic acetylcholine genetics. Am J Psychiatry 2008; 165: 497–506. receptor subunits from ventral tegmental area in systemic 127 De Luca V, Wang H, Squassina A, Wong GW, Yeomans J, Kennedy nicotine self-administration. J Neurosci 2008; 28: 12318–12327. JL. Linkage of M5 muscarinic and alpha7-nicotinic receptor 146 Tapper AR, McKinney SL, Nashmi R, Schwarz J, Deshpande P, genes on 15q13 to schizophrenia. Neuropsychobiology 2004; 50: Labarca C et al. Nicotine activation of alpha4* receptors: 124–127. sufficient for reward, tolerance, and sensitization. Science 128 Zammit S, Spurlock G, Williams H, Norton N, Williams N, 2004; 306: 1029–1032. O’Donovan MC et al. Genotype effects of CHRNA7, CNR1 and 147 Marubio LM, Gardier AM, Durier S, David D, Klink R, Arroyo- COMT in schizophrenia: interactions with tobacco and cannabis Jimenez MM et al. Effects of nicotine in the dopaminergic system use. Br J Psychiatry 2007; 191: 402–407. of mice lacking the alpha4 subunit of neuronal nicotinic 129 Gelernter J, Liu X, Hesselbrock V, Page GP, Goddard A, Zhang H. acetylcholine receptors. Eur J Neurosci 2003; 17: 1329–1337. Results of a genomewide linkage scan: support for chromosomes 148 Parish CL, Nunan J, Finkelstein DI, McNamara FN, Wong JY, 9 and 11 loci increasing risk for cigarette smoking. Am J Med Waddington JL et al. Mice lacking the alpha4 nicotinic receptor Genet B Neuropsychiatr Genet 2004; 128B: 94–101. subunit fail to modulate dopaminergic neuronal arbors and 130 Loukola A, Broms U, Maunu H, Widen E, Heikkila K, Siivola M possess impaired dopamine transporter function. Mol Pharmacol et al. Linkage of nicotine dependence and smoking behavior on 2005; 68: 1376–1386. 10q, 7q and 11p in twins with homogeneous genetic background. 149 Dobelis P, Marks MJ, Whiteaker P, Balogh SA, Collins AC, Stitzel Pharmacogenomics J 2008; 8: 209–219. JA. A polymorphism in the mouse neuronal alpha4 nicotinic 131 Mineur YS, Picciotto MR. Genetics of nicotinic acetylcholine receptor subunit results in an alteration in receptor function. Mol receptors: Relevance to nicotine addiction. Biochem Pharmacol Pharmacol 2002; 62: 334–342. 2008; 75: 323–333. 150 Butt CM, King NM, Hutton SR, Collins AC, Stitzel JA. Modula- 132 Drenan RM, Grady SR, Whiteaker P, McClure-Begley T, McKin- tion of nicotine but not ethanol preference by the mouse Chrna4 ney S, Miwa JM et al. In vivo activation of midbrain dopamine A529T polymorphism. Behav Neurosci 2005; 119: 26–37. neurons via sensitized, high-affinity alpha 6 nicotinic acetylcho- 151 Kim H, Flanagin BA, Qin C, Macdonald RL, Stitzel JA. The line receptors. Neuron 2008; 60: 123–136. mouse Chrna4 A529T polymorphism alters the ratio of high to 133 Fowler CD, Arends MA, Kenny PJ. Subtypes of nicotinic low affinity alpha 4 beta 2 nAChRs. Neuropharmacology 2003; acetylcholine receptors in nicotine reward, dependence, and 45: 345–354. withdrawal: evidence from genetically modified mice. Behav 152 Stolerman IP, Chamberlain S, Bizarro L, Fernandes C, Schalkwyk Pharmacol 2008; 19: 461–484. L. The role of nicotinic receptor alpha 7 subunits in nicotine 134 Picciotto MR, Zoli M, Rimondini R, Lena C, Marubio LM, Pich discrimination. Neuropharmacology 2004; 46: 363–371. EM et al. Acetylcholine receptors containing the beta2 subunit 153 Tritto T, McCallum SE, Waddle SA, Hutton SR, Paylor R, Collins are involved in the reinforcing properties of nicotine. Nature AC et al. Null mutant analysis of responses to nicotine: deletion 1998; 391: 173–177. of beta2 nicotinic acetylcholine receptor subunit but not alpha7 135 Maskos U, Molles BE, Pons S, Besson M, Guiard BP, Guilloux JP subunit reduces sensitivity to nicotine-induced locomotor de- et al. Nicotine reinforcement and cognition restored by targeted pression and hypothermia. Nicotine Tob Res 2004; 6: 145–158. expression of nicotinic receptors. Nature 2005; 436: 103–107. 154 Naylor C, Quarta D, Fernandes C, Stolerman IP. Tolerance to 136 Avale ME, Faure P, Pons S, Robledo P, Deltheil T, David DJ et al. nicotine in mice lacking alpha7 nicotinic receptors. Psychophar- Interplay of beta2* nicotinic receptors and dopamine pathways macology (Berl) 2005; 180: 558–563. in the control of spontaneous locomotion. Proc Natl Acad Sci 155 Salas R, Main A, Gangitano D, De Biasi M. Decreased withdrawal USA 2008; 105: 15991–15996. symptoms but normal tolerance to nicotine in mice null for the 137 Mameli-Engvall M, Evrard A, Pons S, Maskos U, Svensson TH, alpha7 nicotinic acetylcholine receptor subunit. Neuropharma- Changeux JP et al. Hierarchical control of dopamine cology 2007; 53: 863–869. neuron-firing patterns by nicotinic receptors. Neuron 2006; 50: 156 Grabus SD, Martin BR, Imad Damaj M. Nicotine physical 911–921. dependence in the mouse: involvement of the alpha7 nicotinic 138 King SL, Caldarone BJ, Picciotto MR. Beta2-subunit-containing receptor subtype. Eur J Pharmacol 2005; 515: 90–93. nicotinic acetylcholine receptors are critical for dopamine- 157 Salas R, Cook KD, Bassetto L, De Biasi M. The alpha3 and beta4 dependent locomotor activation following repeated nicotine nicotinic acetylcholine receptor subunits are necessary for administration. Neuropharmacology 2004; 47(Suppl 1): 132–139. nicotine-induced seizures and hypolocomotion in mice. Neuro- 139 Shoaib M, Gommans J, Morley A, Stolerman IP, Grailhe R, pharmacology 2004; 47: 401–407. Changeux JP. The role of nicotinic receptor beta-2 subunits in 158 Salas R, Orr-Urtreger A, Broide RS, Beaudet A, Paylor R, De Biasi nicotine discrimination and conditioned taste aversion. Neuro- M. The nicotinic acetylcholine receptor subunit alpha 5 mediates pharmacology 2002; 42: 530–539. short-term effects of nicotine in vivo. Mol Pharmacol 2003; 63: 140 Walters CL, Brown S, Changeux JP, Martin B, Damaj MI. The 1059–1066. beta2 but not alpha7 subunit of the nicotinic acetylcholine 159 Picciotto MR, Addy NA, Mineur YS, Brunzell DH. It is not receptor is required for nicotine-conditioned place preference in ‘either/or’: activation and desensitization of nicotinic acetylcho- mice. Psychopharmacology (Berl) 2006; 184: 339–344. line receptors both contribute to behaviors related to nicotine 141 Owens JC, Balogh SA, McClure-Begley TD, Butt CM, Labarca C, addiction and mood. Prog Neurobiol 2008; 84: 329–342. Lester HA et al. Alpha 4 beta 2* nicotinic acetylcholine 160 Fenster CP, Rains MF, Noerager B, Quick MW, Lester RA. receptors modulate the effects of ethanol and nicotine on Influence of subunit composition on desensitization of neuronal the acoustic startle response. Alcohol Clin Exp Res 2003; 27: acetylcholine receptors at low concentrations of nicotine. 1867–1875. J Neurosci 1997; 17: 5747–5759. 142 Besson M, David V, Suarez S, Cormier A, Cazala P, Changeux JP 161 Fenster CP, Hicks JH, Beckman ML, Covernton PJ, Quick MW, et al. Genetic dissociation of two behaviors associated with Lester RA. Desensitization of nicotinic receptors in the central nicotine addiction: beta-2 containing nicotinic receptors are nervous system. Ann NY Acad Sci 1999; 868: 620–623. involved in nicotine reinforcement but not in withdrawal 162 Corringer PJ, Bertrand S, Bohler S, Edelstein SJ, Changeux JP, syndrome. Psychopharmacology (Berl) 2006; 187: 189–199. Bertrand D. Critical elements determining diversity in agonist 143 Jackson KJ, Martin BR, Changeux JP, Damaj MI. Differential role binding and desensitization of neuronal nicotinic acetylcholine of nicotinic acetylcholine receptor subunits in physical and receptors. J Neurosci 1998; 18: 648–657. affective nicotine withdrawal signs. J Pharmacol Exp Ther 2008; 163 Katz B, Thesleff S. A study of the desensitization produced by 325: 302–312. acetylcholine at the motor end-plate. J Physiol 1957; 138: 63–80.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 944 164 Giniatullin R, Nistri A, Yakel JL. Desensitization of nicotinic ACh portionate increase of the alpha6 subunit. Mol Pharmacol 2004; receptors: shaping cholinergic signaling. Trends Neurosci 2005; 65: 611–622. 28: 371–378. 184 McCallum SE, Parameswaran N, Bordia T, Fan H, McIntosh JM, 165 Walsh H, Govind AP, Mastro R, Hoda JC, Bertrand D, Vallejo Y et Quik M. Differential regulation of mesolimbic alpha 3/alpha 6 al. Up-regulation of nicotinic receptors by nicotine varies with beta 2 and alpha 4 beta 2 nicotinic acetylcholine receptor sites receptor subtype. J Biol Chem 2008; 283: 6022–6032. and function after long-term oral nicotine to monkeys. 166 Breese CR, Marks MJ, Logel J, Adams CE, Sullivan B, Collins AC J Pharmacol Exp Ther 2006; 318: 381–388. et al. Effect of smoking history on [3H]nicotine binding 185 Perry DC, Mao D, Gold AB, McIntosh JM, Pezzullo JC, Kellar KJ. in human postmortem brain. J Pharmacol Exp Ther 1997; 282: Chronic nicotine differentially regulates alpha6- and beta3- 7–13. containing nicotinic cholinergic receptors in rat brain. J Pharma- 167 Perry DC, Davila-Garcia MI, Stockmeier CA, Kellar KJ. Increased col Exp Ther 2007; 322: 306–315. nicotinic receptors in brains from smokers: membrane binding 186 Exley R, Clements MA, Hartung H, McIntosh JM, Cragg SJ. and autoradiography studies. J Pharmacol Exp Ther 1999; 289: Alpha6-containing nicotinic acetylcholine receptors dominate 1545–1552. the nicotine control of dopamine neurotransmission in nucleus 168 Vezina P, McGehee DS, Green WN. Exposure to nicotine and accumbens. Neuropsychopharmacology 2008; 33: 2158–2166. sensitization of nicotine-induced behaviors. Prog Neuropsycho- 187 Meyer EL, Yoshikami D, McIntosh JM. The neuronal nicotinic pharmacol Biol Psychiatry 2007; 31: 1625–1638. acetylcholine receptors alpha 4* and alpha 6* differentially 169 Buisson B, Bertrand D. Chronic exposure to nicotine upregulates modulate dopamine release in mouse striatal slices. J Neurochem the human (alpha)4((beta)2 nicotinic acetylcholine receptor 2008; 105: 1761–1769. function. J Neurosci 2001; 21: 1819–1829. 188 Kendler KS, Neale MC, Sullivan P, Corey LA, Gardner CO, 170 Darsow T, Booker TK, Pina-Crespo JC, Heinemann SF. Exocytic Prescott CA. A population-based twin study in women of trafficking is required for nicotine-induced up-regulation of smoking initiation and nicotine dependence. Psychol Med alpha 4 beta 2 nicotinic acetylcholine receptors. J Biol Chem 1999; 29: 299–308. 2005; 280: 18311–18320. 189 Rigbi A, Kanyas K, Yakir A, Greenbaum L, Pollak Y, Ben-Asher E 171 Wooltorton JR, Pidoplichko VI, Broide RS, Dani JA. Differential et al. Why do young women smoke? V. Role of direct and desensitization and distribution of nicotinic acetylcholine re- interactive effects of nicotinic cholinergic receptor gene variation ceptor subtypes in midbrain dopamine areas. J Neurosci 2003; 23: on neurocognitive function. Genes Brain Behav 2008; 7: 164–172. 3176–3185. 190 Ebstein RP. The molecular genetic architecture of human 172 Mao D, Perry DC, Yasuda RP, Wolfe BB, Kellar KJ. The personality: beyond self-report questionnaires. Mol Psychiatry alpha4beta2alpha5 nicotinic cholinergic receptor in rat brain is 2006; 11: 427–445. resistant to up-regulation by nicotine in vivo. J Neurochem 2008; 191 Kreek MJ, Nielsen DA, Butelman ER, LaForge KS. Genetic 104: 446–456. influences on impulsivity, risk taking, stress responsivity and 173 Vincler MA, Eisenach JC. Immunocytochemical localization of vulnerability to drug abuse and addiction. Nat Neurosci 2005; 8: the alpha3, alpha4, alpha5, alpha7, beta2, beta3 and beta4 1450–1457. nicotinic acetylcholine receptor subunits in the locus coeruleus 192 Pomerleau CS, Pomerleau OF, Flessland KA, Basson SM. of the rat. Brain Res 2003; 974: 25–36. Relationship of Tridimensional Personality Questionnaire scores 174 Xiao Y, Kellar KJ. The comparative pharmacology and up- and smoking variables in female and male smokers. J Subst regulation of rat neuronal nicotinic receptor subtype binding Abuse 1992; 4: 143–154. sites stably expressed in transfected mammalian cells. J Pharma- 193 Lipkus IM, Barefoot JC, Williams RB, Siegler IC. Personality col Exp Ther 2004; 310: 98–107. measures as predictors of smoking initiation and cessation in the 175 Olale F, Gerzanich V, Kuryatov A, Wang F, Lindstrom J. Chronic UNC Alumni Heart Study. Health Psychol 1994; 13: 149–155. nicotine exposure differentially affects the function of human 194 Patton D, Barnes GE, Murray RP. A personality typology of alpha3, alpha4, and alpha7 neuronal nicotinic receptor subtypes. smokers. Addict Behav 1997; 22: 269–273. J Pharmacol Exp Ther 1997; 283: 675–683. 195 Etter JF, Pelissolo A, Pomerleau C, De Saint-Hilaire Z. Associa- 176 Wullner U, Gundisch D, Herzog H, Minnerop M, Joe A, Warnecke tions between smoking and heritable temperament traits. Nico- M et al. Smoking upregulates alpha4beta2* nicotinic acetylcho- tine Tob Res 2003; 5: 401–409. line receptors in the human brain. Neurosci Lett 2008; 430: 196 Reif A, Lesch KP. Toward a molecular architecture of personality. 34–37. Behav Brain Res 2003; 139: 1–20. 177 Staley JK, Krishnan-Sarin S, Cosgrove KP, Krantzler E, Frohlich 197 Ebstein RP, Novick O, Umansky R, Priel B, Osher Y, Blaine D E, Perry E et al. Human tobacco smokers in early abstinence have et al. Dopamine D4 receptor (D4DR) exon III polymorphism higher levels of beta2* nicotinic acetylcholine receptors than associated with the human personality trait of Novelty Seeking. nonsmokers. J Neurosci 2006; 26: 8707–8714. Nat Genet 1996; 12: 78–80. 178 Mamede M, Ishizu K, Ueda M, Mukai T, Iida Y, Kawashima H 198 Benjamin J, Li L, Patterson C, Greenberg BD, Murphy DL, Hamer et al. Temporal change in human nicotinic acetylcholine receptor DH. Population and familial association between the D4 after smoking cessation: 5IA SPECT study. J Nucl Med 2007; 48: dopamine receptor gene and measures of Novelty Seeking. Nat 1829–1835. Genet 1996; 12: 81–84. 179 Schwartz RD, Kellar KJ. In vivo regulation of [3H]acetylcholine 199 Osher Y, Hamer D, Benjamin J. Association and linkage of recognition sites in brain by nicotinic cholinergic drugs. anxiety-related traits with a functional polymorphism of the J Neurochem 1985; 45: 427–433. serotonin transporter gene regulatory region in Israeli sibling 180 Pakkanen JS, Jokitalo E, Tuominen RK. Up-regulation of beta2 pairs. Mol Psychiatry 2000; 5: 216–219. and alpha7 subunit containing nicotinic acetylcholine receptors 200 van den Oord EJ, Kuo PH, Hartmann AM, Webb BT, Moller HJ, in mouse striatum at cellular level. Eur J Neurosci 2005; 21: Hettema JM et al. Genomewide association analysis followed by a 2681–2691. replication study implicates a novel candidate gene for neuroti- 181 Nuutinen S, Ahtee L, Tuominen RK. Time and brain region cism. Arch Gen Psychiatry 2008; 65: 1062–1071. specific up-regulation of low affinity neuronal nicotinic receptors 201 Goodwin R, Hamilton SP. Cigarette smoking and panic: the role during chronic nicotine administration in mice. Eur J Pharmacol of neuroticism. Am J Psychiatry 2002; 159: 1208–1213. 2005; 515: 83–89. 202 Lerman C, Caporaso NE, Audrain J, Main D, Boyd NR, Shields 182 Molinari EJ, Delbono O, Messi ML, Renganathan M, Arneric SP, PG. Interacting effects of the serotonin transporter gene and Sullivan JP et al. Up-regulation of human alpha7 nicotinic neuroticism in smoking practices and nicotine dependence. Mol receptors by chronic treatment with activator and antagonist Psychiatry 2000; 5: 189–192. ligands. Eur J Pharmacol 1998; 347: 131–139. 203 Hu S, Brody CL, Fisher C, Gunzerath L, Nelson ML, Sabol SZ 183 Parker SL, Fu Y, McAllen K, Luo J, McIntosh JM, Lindstrom JM et al. Interaction between the serotonin transporter gene and et al. Up-regulation of brain nicotinic acetylcholine receptors in neuroticism in cigarette smoking behavior. Mol Psychiatry 2000; the rat during long-term self-administration of nicotine: dispro- 5: 181–188.

Molecular Psychiatry Genetic variation in multiple brain nAChRs and ND L Greenbaum and B Lerer 945 204 Lerer E, Kanyas K, Karni O, Ebstein RP, Lerer B. Why do young differences in visuospatial attention. J Cogn Neurosci 2005; 17: women smoke? II. Role of traumatic life experience, psychologi- 1611–1620. cal characteristics and serotonergic genes. Mol Psychiatry 2006; 218 Greenwood PM, Sundararajan R, Lin MK, Kumar R, Fryxell KJ, 11: 771–781. Parasuraman R. Both a Nicotinic SNP and a Noradrenergic SNP 205 Segman RH, Kanyas K, Karni O, Lerer E, Goltser-Dubner T, Modulate Working Memory Performance when Attention is Pavlov V et al. Why do young women smoke? IV. Role of genetic Manipulated. J Cogn Neurosci 2008; e-pub ahead of print 18 variation in the dopamine transporter and lifetime traumatic November 2008. experience. Am J Med Genet B Neuropsychiatr Genet 2007; 144B: 219 Byrne DG, Byrne AE, Reinhart MI. Personality, stress and the 533–540. decision to commence cigarette smoking in adolescence. 206 Rezvani AH, Levin ED. Cognitive effects of nicotine. Biol J Psychosom Res 1995; 39: 53–62. Psychiatry 2001; 49: 258–267. 220 Middeldorp CM, Cath DC, Van Dyck R, Boomsma DI. The co- 207 Newhouse PA, Potter A, Singh A. Effects of nicotinic stimulation morbidity of anxiety and depression in the perspective of genetic on cognitive performance. Curr Opin Pharmacol 2004; 4: 36–46. epidemiology. A review of twin and family studies. Psychol Med 208 Dinn WM, Aycicegi A, Harris CL. Cigarette smoking in a student 2005; 35: 611–624. sample: neurocognitive and clinical correlates. Addict Behav 221 Kassel JD, Unrod M. Smoking, anxiety, and attention: support for 2004; 29: 107–126. the role of nicotine in attentionally mediated anxiolysis. 209 Yakir A, Rigbi A, Kanyas K, Pollak Y, Kahana G, Karni O et al. J Abnorm Psychol 2000; 109: 161–166. Why do young women smoke? III. Attention and impulsivity as 222 Labarca C, Schwarz J, Deshpande P, Schwarz S, Nowak MW, neurocognitive predisposing factors. Eur Neuropsychopharmacol Fonck C et al. Point mutant mice with hypersensitive alpha 4 2007; 17: 339–351. nicotinic receptors show dopaminergic deficits and increased 210 Lambert NM, Hartsough CS. Prospective study of tobacco anxiety. Proc Natl Acad Sci USA 2001; 98: 2786–2791. smoking and substance dependencies among samples of 223 Ross SA, Wong JY, Clifford JJ, Kinsella A, Massalas JS, Horne MK ADHD and non-ADHD participants. J Learn Disabil 1998; 31: et al. Phenotypic characterization of an alpha 4 neuronal 533–544. nicotinic acetylcholine receptor subunit knock-out mouse. 211 Lee J, Laurin N, Crosbie J, Ickowicz A, Pathare T, Malone M et al. J Neurosci 2000; 20: 6431–6441. Association study of the nicotinic acetylcholine receptor alpha4 224 Booker TK, Butt CM, Wehner JM, Heinemann SF, Collins AC. subunit gene, CHRNA4, in attention-deficit hyperactivity dis- Decreased anxiety-like behavior in beta3 nicotinic receptor order. Genes Brain Behav 2008; 7: 53–60. subunit knockout mice. Pharmacol Biochem Behav 2007; 87: 212 Potter AS, Newhouse PA, Bucci DJ. Central nicotinic cholinergic 146–157. systems: a role in the cognitive dysfunction in attention-deficit/ 225 Kawai H, Berg DK. Nicotinic acetylcholine receptors containing hyperactivity disorder? Behav Brain Res 2006; 175: 201–211. alpha 7 subunits on rat cortical neurons do not undergo long- 213 Brookes K, Xu X, Chen W, Zhou K, Neale B, Lowe N et al. The lasting inactivation even when up-regulated by chronic nicotine analysis of 51 genes in DSM-IV combined type attention deficit exposure. J Neurochem 2001; 78: 1367–1378. hyperactivity disorder: association signals in DRD4, DAT1 and 16 226 Grady SR, Salminen O, Laverty DC, Whiteaker P, McIntosh JM, other genes. Mol Psychiatry 2006; 11: 934–953. Collins AC et al. The subtypes of nicotinic acetylcholine 214 Todd RD, Lobos EA, Sun LW, Neuman RJ. Mutational analysis of receptors on dopaminergic terminals of mouse striatum. Biochem the nicotinic acetylcholine receptor alpha 4 subunit gene in Pharmacol 2007; 74: 1235–1246. attention deficit/hyperactivity disorder: evidence for association 227 Ramirez-Latorre J, Yu CR, Qu X, Perin F, Karlin A, Role L. of an intronic polymorphism with attention problems. Mol Functional contributions of alpha5 subunit to neuronal acet- Psychiatry 2003; 8: 103–108. ylcholine receptor channels. Nature 1996; 380: 347–351. 215 Comings DE, Gade-Andavolu R, Gonzalez N, Wu S, Muhleman D, 228 Gerzanich V, Wang F, Kuryatov A, Lindstrom J. alpha 5 Subunit Blake H et al. Multivariate analysis of associations of 42 genes in alters desensitization, pharmacology, Ca þþ permeability and ADHD, ODD and conduct disorder. Clin Genet 2000; 58: 31–40. Ca þþ modulation of human neuronal alpha 3 nicotinic 216 Bobb AJ, Addington AM, Sidransky E, Gornick MC, Lerch JP, receptors. J Pharmacol Exp Ther 1998; 286: 311–320. Greenstein DK et al. Support for association between ADHD and 229 Kedmi M, Beaudet AL, Orr-Urtreger A. Mice lacking two candidate genes: NET1 and DRD1. Am J Med Genet B neuronal nicotinic acetylcholine receptor beta4-subunit Neuropsychiatr Genet 2005; 134B: 67–72. and mice lacking both alpha5- and beta4-subunits are highly 217 Greenwood PM, Fossella JA, Parasuraman R. Specificity of the resistant to nicotine-induced seizures. Physiol Genomics 2004; effect of a nicotinic receptor polymorphism on individual 17: 221–229.

Supplementary Information accompanies the paper on the Molecular Psychiatry website (http:// www.nature.com/mp)

Molecular Psychiatry