Translational Research in Nicotine Addiction

Abstract

While commendable strides have been made in reducing smoking initiation and improving smoking cessation rates, current available smoking cessation treatment options are still only mildly efficacious and show substantial interindividual variability in their therapeutic responses. Therefore, the primary goal of preclinical research has been to further the understanding of the neural substrates and genetic influences involved in nicotine's effects and reassess potential drug targets. Pronounced advances have been made by investing in new translational approaches and placing more emphasis on bridging the gap between human and rodent models of dependence. Functional neuroimaging studies have identified key brain structures involved with nicotine-dependence phenotypes such as craving, impulsivity, withdrawal symptoms, and smoking cessation outcomes. Following up with these findings, rodent-modeling techniques have made it possible to dissect the neural circuits involved in these motivated behaviors and ascertain mechanisms underlying nicotine's interactive effects on brain structure and function. Likewise, translational studies investigating single-nucleotide polymorphisms (SNPs) within the cholinergic, dopaminergic, and opioid systems have found high levels of involvement of these neurotransmitter systems in regulating the reinforcing aspects of nicotine in both humans and mouse models. These findings and coordinated efforts between human and rodent studies pave the way for future work determining gene by drug interactions and tailoring treatment options to each individual smoker.

Worldwide, tobacco use causes more than 7 million deaths each year (Ebbert et al. 2010) and increases incidences of heart disease, stroke, and cancer in smokers (Warren et al. 2014).³ Tobacco contains the chemical nicotine, which does not directly cause tobacco-related morbidity, but is the main culprit of dependency in smokers due to its addictive properties (Stolerman and Jarvis 1995). Although the detrimental effects of smoking are generally understood and 70% of smokers indicate that they want to quit, successful attempts among smokers remain <5% (Nides 2008). Many factors contribute to the development and maintenance of nicotine addiction, and these processes have been investigated in both human populations and animal models to better understand underlying mechanisms. Nicotine, classified as a stimulant drug, produces paradoxical effects (Gilbert 1979)—acting as both a stimulant and a depressant by increasing attention, learning and memory, and information processing (Froeliger et al. 2009), while also alleviating anxiety and depression (Benowitz 1996; McClernon et al. 2006; Matta et al. 2007). Cessation of chronic nicotine produces withdrawal symptoms in both animals (Malin et al. 1992; Grabus et al. 2005) and humans (Hendricks et al. 2006; Hughes 2007), and avoidance of withdrawal symptoms is a major factor that contributes to the maintenance of smoking and relapse during quit attempts. For example, studies have also shown that the severity and duration of nicotine withdrawal symptoms strongly predict relapse (Piasecki et al. 1998, 2000).

Currently, there are three “first-line” pharmacotherapies for nicotine addiction: nicotine replacement therapy ([NRT]; transdermal patch, nasal spray, gum, inhaler, lozenge, e-cigarettes), bupropion (Wellbutrin or Zyban, a mixed norepinephrine/dopamine (DA) reuptake inhibitor), and varenicline (Chantix, a nicotinic acetylcholine receptor [nAChR] partial agonist). While these drugs do show some efficacy in the maintenance of smoking cessation by reducing cravings and withdrawal symptoms (Matta et al. 2007; Nides 2008), there is substantial interindividual variability in their therapeutic responses (Babb et al. 2017). This highlights the importance of investing in new translational approaches and placing more emphasis on bridging the gap between human and rodent models of dependence and their complementary findings. This review has been tailored to first discuss how nicotine's biological effects are examined in the realm of preclinical research—focusing on pharmacological and behavioral modeling of nicotine dependence in both humans and rodents. And second, we narrate a collection of the current translational studies showcasing parallel functional and genetic findings between human and rodent models of nicotine dependence.

NEUROBIOLOGY UNDERLYING THE ADDICTIVE NATURE OF NICOTINE

Neurochemistry of Nicotine

Nicotine, the major psychoactive and addictive component in tobacco smoke, is thought to mediate both tobacco reinforcement and dependence (Le Foll and Goldberg 2009). Nicotine from a smoked cigarette (the most efficient delivery system into the body) will reach the brain in as little as 7 seconds after inhalation (Maisto and Connors 2004), delivering, on average, 1–2 mg of nicotine per cigarette (Hoffmann and Hoffmann 1997) into the bloodstream. Once in the bloodstream, nicotine rapidly crosses the blood–brain barrier because of its highly lipophilic properties (Gahring and Rogers 2005) and can be isolated in lipid-rich, slightly basic reservoirs, such as glia (Crooks 1999). This compartmentalization of nicotine can lead to its accumulation in the brain during chronic administration (Ghosheh et al. 2001), potentially producing continued effects following termination of nicotine exposure. Once in the brain, nicotine binds to nAChRs. These receptors are pentameric ion channels, which pass Na⁺, K⁺, and Ca²⁺ ions and, thus, have the ability to alter cellular activity. Entry of these ions can directly impact cell excitability or trigger calcium-sensitive molecules, such as protein kinase C (PKC) (Soliakov and Wonnacott 2001), protein kinase A (PKA) (Dajas-Bailador et al. 2002), calmodulin-dependent protein kinase II (CAMKII) (Steiner et al. 2007), and extracellular signal-regulated kinases (ERKs) (Dajas-Bailador et al. 2002; Steiner et al. 2007). These calcium-sensitive kinases then have numerous downstream effects, including activation of transcription factors such as CREB (Chang and Berg 2001; Pandey et al. 2001; Hu et al. 2002; Brunzell et al. 2003; Walters et al. 2005) (for review of signaling effects of nicotine, see Shen and Yakel 2009).

Nicotinic Receptors (nAChRs)

In the vertebrate central nervous system, there are 12 genes that encode 12 distinct α and β subunits, CHRNA2-10 and CHRNAB2-4. They are classified as α in the presence of a Cys-Cys pair near the start of TM1, or β when the Cys pair is absent (Le Novere and Changeux 1995). These subunits assemble together to form the pentameric ion channel in a variety of homomeric or heteromeric stoichiometries (Fowler et al. 2008). With dozens of naturally expressing nAChR subtypes (Wu and Lukas 2011), the most widely expressed subtypes in the brain are the α7 homomeric receptors and α4β2 heteromeric receptors. Variability in composition, neuroanatomical localization, and pharmacological characteristics of these receptors adds to their specificity and complexity, and potential influences on nicotine dependence. For example, the α4β2 nAChR subtype has been shown to be the predominant nAChR subtype up-regulating following chronic exposure to nicotine, as evidenced by findings in cell culture (Xiao and Kellar 2004), rodents (Schwartz and Kellar 1983; Marks et al. 1985), monkeys (Picciotto et al. 2008), and humans (Mukhin et al. 2008). The up-regulation of receptors after agonist activation is a characteristic unique to nicotine, and its full importance is not clearly understood. Studies have found that increases in nAChR density correlate with withdrawal symptoms severity (Gould et al. 2012). Additionally, positron emission tomography (PET) imaging in human smokers suggest that this up-regulation may directly contribute to smoking relapse as well. Cosgrove and colleagues (2009) demonstrated that β2-containing nAChRs remain significantly up-regulated after 1 month of abstinence, and this increase in receptor density is positively correlated to craving. Furthermore, Brody and colleagues (2006) found that as few as one to two puffs of a cigarette result in 50% occupancy of α4β2 nAChRs for >3 hours after smoking. Although correlational, this study suggests that nAChR up-regulation after chronic use could directly contribute to failed smoking cessation.

MEASURING NICOTINE DEPENDENCE IN SMOKERS

Emerging studies are characterizing how simple questionnaire-based assessments of an individual's smoking habits can provide clinical relevance and predictive power of measures such as the degree of dependency or response to smoking cessation therapies. There are several traditional methods used to measure the level of dependence and to predict the likelihood of relapse. Assessments such as the Fagerström Tolerance Questionnaire ([FTQ]; Fagerström 1978), the Heaviness Smoking Index ([HIS]; Heatherton et al. 1989), and the Fagerström Test of Nicotine Dependence ([FTND]; Heatherton et al. 1991), as well as the Diagnostic and Statistical Manual, 5th Edition (DSM-V) criteria are used by researchers to evaluate not only physical dependence but also “cognitive, behavioral, and physiological symptoms” as per the DSM-V definition. In smoking intervention trials, epidemiological studies, and genetic studies, the FTND test is the most commonly used to assess dependence characteristics such as cigarette consumption and the compulsion to use. This assessment consists of a scale of 1–10, with scores 1–2 indicating low dependence and 8+ high dependence.

Despite its popularity, studies have speculated the FTND test to have low reliability and validity, and limited ability to predict biochemical markers of dependence (Payne et al. 1994; Pomerleau et al. 1994; Etter et al. 1999), perhaps because of its “yes or no” forced-answer format. More recently developed assessments such as the Nicotine Dependence Syndrome Scale ([NDSS]; Shiffman et al. 2004) and Wisconsin Inventory of Smoking Dependence Motives ([WISDM]; Piper et al. 2004) have been designed based off a more multifactorial perspective of dependence—taking into account theories of dependence syndrome (Edwards 1986) and a multitude of motivational domains (habitual/automatic, positive affect, negative affect, etc.) (Smith et al. 2010). Collectively, these measures have been shown to predict clinically important dependence criteria such as craving, severity of withdrawal symptoms, rate of nicotine metabolism, and smoking cessation outcomes (Fagerström and Schneider 1989; Heatherton et al. 1991; Alterman et al. 1999; Breslau and Johnson 2000; Piper et al. 2004; Shiffman et al. 2004).

MODELING NICOTINE DEPENDENCE IN RODENTS

With standard assessments in place to measure nicotine dependence and nicotine exposure modalities in humans, developing valid and reproducible animal models of nicotine dependence is imperative to identify and characterize clinically relevant neuronal adaptations that occur from chronic drug use. Animal models have become useful tools in advancing our understanding of the neurobiological processes underlying initiation, maintenance, withdrawal, and relapse to smoking. Because the human condition of nicotine dependence is what drives the need for this research, it is important for translational studies to be designed with certain factors in mind. Routes of administration, length of exposure, dose, drug cues, and genetic variability, etc. can have drastically converging physiological effects between species. Matta et al. (2007) published a set of guidelines for nicotine dose and administration selection for in vivo research focusing on individual species used in modeling addiction. Their review addresses issues related to acute versus chronic exposure, nicotine metabolism, genetic background, route of administration, and behavioral responses (Matta et al. 2007), and represents a great resource for those conducting in vivo animal studies.

Routes of Nicotine Administration

To date, the majority of research on the behavioral and biological effects of nicotine in rodent models involves noncontingent exposure of nicotine, typically by either injection (subcutaneous or intraperitoneal), subcutaneous implantation of osmotic minipumps engineered to deliver a steady-state infusion of nicotine for experimentally defined lengths of administration (Malin et al. 1992), or from inhalation procedures (Alasmari et al. 2018). These types of nicotine delivery systems are instrumental in identifying the effects of acute and chronic exposure to nicotine on a wide variety of behavioral responses such as locomotor activity and anxiety-like behavior. Conversely, a contingent form of nicotine administration is self-administration, where animals are placed in an operant chamber and have the opportunity to voluntarily administer nicotine, typically delivered intravenously, upon emission of an operant response (lever-press, nose poke, etc.). This form of nicotine delivery assesses an animal's propensity to self-administer, which is advantageous when investigating a drug's reinforcing effects on behavior (Donny et al. 1995).

Rodent Behavioral Paradigms

Reward

Along with nicotine self-administration paradigms, another commonly used model of drug reward is conditioned place preference (CPP). This behavioral paradigm tests for the development of conditioned preference or aversion to an environment associated with prior drug exposure. CPP is evidenced by animals spending more time in the drug-paired environment versus vehicle-paired, showcasing a rewarding effect of the drug. Whereas aversion to a drug results in the opposite effect, animals will spend more time in the vehicle-paired environment (Prus et al. 2009). Collectively, these two models enable researchers to investigate effects of genetic or pharmacological manipulation has on nicotine reward phenotypes. Achieving nicotine-induced CPP in rodents has proven challenging compared to other drugs of abuse with some studies observing place preference (Fudala et al. 1985; Horan et al. 1997; Le Foll and Goldberg 2005), while others observe no drug effect or aversion to the drug (Clarke and Fibiger 1987; Jorenby et al. 1990; Parker 1992). It remains unclear why there are inconsistencies with this behavioral model, but some speculate it could be due to a very narrow dose–response curve with the difference between the rewarding and aversive doses of nicotine being very small (Benowitz 1990).

Nicotine's rewarding effects have also been associated with deficits in impulsive control. Impulsivity is generally characterized as a predisposition to make risky decisions without adequate forethought. Deficits in impulse control increases vulnerability to dependence in humans (Verdejo-García et al. 2008) and rodents (Dalley et al. 2007; Belin et al. 2008). Human impulsivity is evaluated by a variety of behavioral tasks that measure different types of impulsivity (e.g., motor, nonplanning, and attentional [Patton et al. 1995; Moeller et al. 2001]). These can be further simplified as either impulsive action, the ability to withhold a motor response, or impulsive choice, where motivational forces affect impulsive responding. In rodents, the stop-signal task (SST), the go/no-go task, and the five-choice serial reaction time task (Robbins 2002; Bari et al. 2008; Dalley et al. 2008) have high face validity in measuring motor and attentional aspects of impulsivity. The overall experimental premise of these tasks is to measure response inhibition or impulsive action—the ability to withhold a pre-potent response (Evenden 1999; D'Amour-Horvat and Leyton 2014).

Withdrawal

Termination of repeated or chronic administration of a drug results in physiological states consistent with drug withdrawal (see review by Malin 2001). The symptoms of nicotine withdrawal in mice reported consist of somatic signs, such as teeth-chattering, excessive grooming, tremors, arching of the back, etc. (Malin et al. 1992), as well as affective changes and deficits in cognition (see review in De Biasi and Salas 2008). When conducting rodent studies modeling affective withdrawal symptoms, there are a variety of behavioral tests that measure anxiety-like and depressive-like symptoms in rodents and are sensitive to nicotine withdrawal-induced behaviors. For example, the open field exploration test consists of placing the animal, singly, in a brightly lit, unprotected open area. By doing so, it elicits anxiety-like behavior from social isolation of being separated from cage mates, and the stress of being in a novel test environment (Crawley 1985). A second paradigm, the elevated plus-maze/elevated zero-maze is also a well-established paradigm for assessing anxiety-like behavior in mice (Walf and Frye 2007). Mice are placed on a maze apparatus ∼1 m from the floor and given the choice of spending time in open, unprotected maze arms or in the enclosed, protected arms of a maze. In both of these paradigms, mice tend to avoid open, brightly lit areas, preferring darker and more enclosed areas, but administration of anxiolytic drugs, such as benzodiazepines, result in increased time exploring in the open area or arms of the apparatus, indicative of an anxiolytic response (review in Pellow and File 1986). Last, the novelty-induced hypophagia test is designed to measure the latency of a mouse to feed in a novel test environment. Mice naturally avoid exploring novel environments yet are motivated to approach and consume highly palatable food. This inhibition in feeding behavior, often referred to as hyponeophagia, is a reliable indicator of anxiety-like behavior in mouse and rats (Deacon 2011) and is found to be sensitive to nicotine (Hussmann et al. 2014) and withdrawal (Turner et al. 2013, 2014; Yohn et al. 2014; Fisher et al. 2017).

Nicotine withdrawal is associated with deficits in neurocognitive function including working memory, sustained attention, and response inhibition, all of which have been successfully modeled in rodents and developed to model the cognitive deficits observed in humans. Associative learning is often measured in contextual fear conditioning where an animal associates a certain context with a foot shock. Studies have successfully shown that nicotine enhances contextual fear conditioning (Gould and Higgins 2003) or the recall of this associative memory, while withdrawal from nicotine impairs recall (Davis et al. 2005; Davis and Gould 2009; Portugal and Gould 2009; Raybuck and Gould 2009). Animal models of memory typically employ maze tasks, such as radial-arm maze and Morris water maze, which require rodents to use visuospatial cues to learn the location of food. Withdrawal in rats results in deficits in performance of these tasks (Levin et al. 1990, 2006), while nicotine has been shown to reverse memory impairments (Levin et al. 1993). Altogether, these tests are beneficial in evaluating the effect of pharmacological or genetic manipulation on impaired cognitive functions seen in nicotine dependence.

TRANSLATIONAL STUDIES INVESTIGATING FUNCTIONAL CORRELATES OF NICOTINE DEPENDENCE

Over the last half-century, scientists have queried the physiological processes underlying the transition from casual to habitual and motivated drug use. Olds and Milner were among the first to demonstrate that certain brain regions act as “pleasure” centers (Olds and Milner 1954; Heath 1963, 1972), leading to the identification of brain regions and their neuronal pathways that make up what is now referred to as the mesocortic and mesolimbic pathways. The mesocortic or “reward” pathway consists of dopaminergic cell bodies originating in the ventral tegmental area (VTA) that project to and terminate in the nucleus accumbens (NAc), a region recognized for its role in translating motivation into action (Mogenson et al. 1980). Natural rewards, such as food or sex, as well as all examined drugs of abuse (alcohol, amphetamine, nicotine, opiates, cocaine, etc.) lead to activation of this reward pathway via an extracellular increase of DA in the NAc (Carlezon and Wise 1996; Wise 1996; Di Chiara 1998; Fredholm and Svenningsson 2003). Nicotine-induced release of DA into the synapse, and subsequent activation of dopaminergic receptors on neighboring cells, is believed to modulate neural activity, ultimately influencing synaptic transmission and strengthening neural networks associated with motivated and goal-directed behaviors (for review, see Salamone et al. 2016).

Neuroimaging tools such as structural and functional magnetic resonance imaging (sMRI, fMRI) and PET have been essential in forward translating preclinical research and elucidating acute pharmacological, neurocognitive, and long-term neurobiological effects of smoking on the brain (Kober and Deleone 2011), as discussed below. The NAc possesses reciprocal connections with several other limbic and cortical regions. This complex interplay between regions is believed to carry information about executive and motor plans, behavioral flexibility, learning and memory, and emotional processing (Goto and Grace 2008). The limbic system is a group of structurally and functionally related areas of the brain that underlie not just reward-related events and motivated behaviors, but also learning and memory and emotional processing. These mesolimbic regions consist of the hypothalamus, amygdala, hippocampus, and several other intimately connected regions, such as the prefrontal cortex (PFC), VTA, and basal ganglia (Goto and Grace 2008).

Negative Affect and Cognition

Supporting data in human (Picciotto et al. 2002; Pomerleau et al. 2005) and animal models (Davis and Gould 2009; Kenney et al. 2012; Turner et al. 2013, 2014; Wilkinson et al. 2013; Kutlu et al. 2016; Fisher et al. 2017) link hippocampal function with negative cognitive and affective nicotine withdrawal symptoms. The hippocampus is no longer thought of as a homogenous structure, but instead an eloquently designed region divided by distinct functional and anatomical differences along its anterior (ventral) to posterior (dorsal) axis. Interestingly, both human (McClernon et al. 2016) and rodent studies (Kenney et al. 2012; Turner et al. 2013; Gould et al. 2014; Fisher et al. 2017; Xia et al. 2017) have demonstrated that these substructures differentially facilitate nicotine dependence. Dorsal hippocampal activity is found to mediate the context-dependent rewarding aspects of nicotine in rodents (Xia et al. 2017), as well as the cognitive impairments seen during withdrawal (Wilkinson et al. 2013; Fisher et al. 2017). Whereas, Fisher et al. (2017) demonstrated the ventral hippocampus has little effect on cognitive performance but has a profound effect on anxiety-like responses during nicotine withdrawal (Fisher et al. 2017). In smokers, structural and functional imaging studies have mirrored these findings with hippocampal activity being associated with negative affect, reactivity to smoking images and stimuli, and quitting success (Franklin et al. 2007; McClernon et al. 2008; Froeliger et al. 2010; Janes et al. 2010). For example, McClernon et al. (2016), found that photographed locations personally associated with smoking increased blood oxygen levels in the posterior (dorsal) hippocampus, which was associated with higher craving levels and greater ad lib smoking.

Furthermore, it has been demonstrated that the hippocampus is also critical for context-induced reinstatement of drug-seeking behavior in mice via projections to both the PFC and NAc (Crombag et al. 2008). Excitatory inputs from the hippocampus to the NAc increase action-potential neuronal firing by DA-mediated receptor activation, leading to drug-seeking behavior. Blockade of glutamatergic signaling within this circuit prevents this response (Crombag et al. 2008). In smokers, chronic nicotine induced dysregulation of this corticostriatal circuitry has been shown to impact relapse rates. Resting-state functional connectivity (rsFC) fMRI, a method used for assessing functionally connected neural networks (Shmuel and Leopold 2008), has proven to be effective in characterizing the pathophysiology of substance use disorders (Lu and Stein 2013). rsFC fMRI studies have indicated that the severity of nicotine dependence is inversely associated with the strength of activity within the corticostriatal circuitry (Hong et al. 2009), with stronger corticostriatal connectivity associated with reducing smoking prior to making a quit attempt (Froeliger et al. 2017a), maintaining smoking abstinence, less craving, and greater positive affect (Froeliger et al. 2015). These findings highlight the complex interplay between limbic and cortical regions of the dependent brain in many aspects of addiction—from drug-seeking behavior to the manifestation of negative withdrawal symptoms, and ultimately to relapse.

Impulsivity

In addition to limbic-mediated affect and learning/memory mechanisms, nicotine dependence is evidenced by significant disruptions to multiple forms of executive function, including impulse control (Hester et al. 2010). Impulsivity is characterized by the tendency to engage in risky or maladaptive behaviors without proper forethought regarding the negative consequences that can occur from such behaviors (Evenden 1999; de Wit 2009; Jupp et al. 2013) and is a hallmark characteristic of many neuropsychological disorders, including addiction. Rodent models of impulsivity share strong parallels with tasks used to assess this trait in humans, and reveal similar circuitry is involved in their performance. For example, rodent models of nicotine dependence have demonstrated that prenatal (Schneider et al. 2011, 2012; Zhang et al. 2018), adolescent (Trauth et al. 2000; Counotte et al. 2009, 2011), and adult (Alasmari et al. 2018; Zhou et al. 2018; Cole et al. 2019) exposure to nicotine increases impulsivity and nicotine consumption (Schneider et al. 2012), through stimulation of α4β2 nAChRs in the infralimbic cortex (Tsutsui-Kimura et al. 2010; Ohmura et al. 2017). In addition to increased consumption of nicotine, studies have found high-impulsive rodents are more willing to work for nicotine in self-administration paradigms than their low-impulsive counterparts (Diergaarde et al. 2008). Diergaarde et al. (2008) found this phenotype associated with reduced DA release in corticostriatal areas such as the NAc and medial PFC. This is in line with numerous other reports showcasing that DA augmenting drugs (e.g., amphetamines, cocaine, DA reuptake inhibitors, etc.) increase premature responding (D'Amour-Horvat and Leyton 2014).

Complementary to these findings, human studies have shown that individuals who are regular smokers also tend to score higher on measures of impulsivity than those who have never smoked (Mitchell 1999) and may be more sensitive to the acute rewarding effects of nicotine (Perkins et al. 2008). Measures of inhibitory control among individuals with substance use disorders are used as transdiagnostic indicators of relapse vulnerability across many substances of abuse (Moeller et al. 2016). Smokers, when compared to nonsmokers, exhibit significant performance deficits on inhibitory control tasks (Luijten et al. 2011; Nestor et al. 2011; Dinur-Klein et al. 2014); and abstaining from smoking has been found to further disrupt inhibitory control performance (Powell et al. 2004; Kozink et al. 2010a). Moreover, performance on these tasks has been shown to be predictive of smoking relapse (Kozink et al. 2010a; Powell et al. 2010; Froeliger et al. 2017b) and the capacity to resist ad lib smoking in a laboratory setting (Mueller et al. 2009; Froeliger et al. 2017b).

In humans, impulsive control behaviors are proposed to be carried out via glutamatergic-mediated excitation from the PFC to subthalamic nucleus (STN) and then to pallidum, in turn exerting GABAergic-mediated inhibition from the pallidum to the thalamus (Jahanshahi et al. 2015). Evidence of these interconnections have been provided through the use of tractography (Aron et al. 2007), a 3D-modeling technique used to visualize nerve tracts. Functional evidence for the involvement of this circuitry in inhibitory control performance is collected through combined techniques such as task-based fMRI and electrocorticography (Swann et al. 2012), inhibitory control task-based effective connectivity (Rae et al. 2015), and lesion studies (Aron et al. 2003). Neuroimaging studies have found that, compared to nonsmokers, smokers exhibit less gray-matter volume within the corticothalamic pathway, including the inferior frontal gyrus (IFG) (Brody et al. 2004; Gallinat et al. 2006; Fritz et al. 2014) and thalamus (Liao et al. 2012; Franklin et al. 2014), and that smoking abstinence increases IFG blood oxygen level–dependent (BOLD) response during inhibitory control (Kozink et al. 2010a) and other neurocognitive tasks (160; Froeliger et al. 2012; Kozink et al. 2010b). Additionally, relative hyperactivity in the IFG and weaker inhibitory control task-based functional connectivity between the IFG and thalamus (corticothalamic circuit) among smokers at baseline (i.e., pre-quit) has been demonstrated to be associated with worse smoking outcomes following a quit attempt, emphasizing the potential use of fMRIs in predicting the likelihood of relapse due to deficits in impulse control (Froeliger et al. 2017b). This collection of data from both rodent and human studies examining nicotine's effects on impulsivity suggest that interventions (medications, behavioral therapies, etc.) aimed to improve inhibitory control might aid in increasing successful smoking cessation rates.

TRANSLATIONAL STUDIES INVESTIGATING GENETIC CORRELATES OF NICOTINE DEPENDENCE

With the use of genome-wide association studies (GWAS) and candidate gene studies, scientists now have access to hundreds of potential therapeutically relevant gene targets. The awareness of a link between genetic and biological factors and smoking cessation therapy effectiveness drives the use of pharmacogenomic research as a platform for identifying new drug targets. Research shows that the genetic variants collected from pharmacogenomic screens may predict therapeutic response, thus improving treatment outcomes. For example, gender differences play an important role in smoking cessation medication effectiveness—with women less likely to quit than men (Wetter et al. 1999). It is believed that identification and mechanistic understanding of the genetic variations underlying nicotine- dependence phenotypes will prove valuable in developing new smoking cessation therapies. In this next section, we will review how twin studies, linkage studies, candidate gene studies and GWAS in human cohorts have taken great strides in identifying contributors of genetic variability associated with nicotine addiction. We will also discuss etiologically relevant animal models of addiction and highlight how they have helped expand our knowledge of gene function and how genetic variation relates to smoking phenotypes.

nAChR Polymorphisms in Smokers

Whether by using a candidate gene approach or a GWAS approach, a number of studies have implicated single-nucleotide polymorphisms (SNPs) occurring in nicotinic subunit genes in the etiology of smoking. The most widely evaluated example of this is the CHRNA5-CHRNA3-CHRNB4 gene cluster, which has been examined for associations with nicotine dependence phenotypes, withdrawal symptoms, and smoking cessation. A number of recent GWAS and pathway-based studies have identified SNPs in this gene cluster associated with heaviness of smoking and/or nicotine dependence (Bierut et al. 2007; Saccone et al. 2007; Berrettini et al. 2008; Thorgeirsson and Stefansson 2008; Caporaso et al. 2009; Thorgeirsson et al. 2010). SNPs within the CHRNA5 gene in particular have been hypothesized to mediate the rewarding effects of nicotine. For example, Berrettini et al. (2008) found SNPs within the CHRNA5 gene to be associated with an increase in reported cigarettes smoked per day (CPD). Other aspects of nicotine dependence, such as nicotine tolerance, smoking initiation, craving, withdrawal severity, and inability to stop smoking, have also been associated with CHRNA5/A3/B4 SNPs (Baker et al. 2009), collectively linking these variants to smoking behavior and a higher risk of developing addiction to nicotine.

nAChR Studies in Mouse Models

Transgenic mice with subunit deletion, mutation, or overexpression have been useful in defining the contribution of nAChR subtypes to specific functions (Champtiaux and Changeux 2004). Studies from nicotinic receptor knockout (KO) mice have helped to elucidate the relative contributions of specific subunits to discrete behaviors pertinent to nicotine dependence and withdrawal (for review, see Fowler et al. 2008). For example, studies evaluating the function of the individual subunits encoded from the CHRNA5/A3/B4 gene cluster, have found that Chrnb4 and Chrna5 KO mice show similar phenotypes, such as reduced signs of withdrawal symptoms (Salas et al. 2004; Jackson et al. 2008), decreased somatic signs, resistance to nicotine-induced seizures, and alterations in locomotor activity (Kedmi et al. 2004). Interestingly, Frahm et al. (2011) found that overexpression of the Chrnb4 subunit results in strong aversion to nicotine. This effect was reversed by viral-mediated, site-specific expression of a Chrna5 variant (D398N) associated with high risk of nicotine dependence in humans (rs1696998) (Bierut et al. 2008; Saccone et al. 2009; Bierut 2010). Furthermore, a recently generated transgenic mouse model (tgCHRNA5/A3/B4), overexpressing the human CHRNA5/A3/B4 cluster, is reported to have increased sensitivity and preference to nicotine (Gallego et al. 2012). Together, data collected from these mouse models suggests a role of the CHRNA4/A3/B4 gene cluster in the rewarding and aversive properties of nicotine.

DRD2 Polymorphisms in Smokers

In addition to the cholinergic system, many studies have also shown that genetic variation within the dopaminergic system is associated with nicotine phenotypes and smoking cessation outcomes (Herman et al. 2014). DA acts through five receptor subtypes (D1–D5), with these subtypes further classified under two broad receptor families: D1-like (D1 and D5 receptors) and D2-like (D2–D4 receptors). These two families have opposing signal transduction functions, but are believed to work concordantly, modulating dopaminergic signaling (Hasbi et al. 2011). Genome-wide linkage analyses have shown that the DRD2 region specifically of chromosome 11 (11q23), is linked with increased risk for cigarette smoking (Gelernter et al. 2004). Located ∼10 kb downstream from the DRD2 gene is the widely published DRD2/ANKK1 Taq1A polymorphism (rs1800497), linked to nicotine dependence and smoking cessation outcomes (Huang et al. 2009, 2015; Stapleton et al. 2011; David et al. 2013; Mayer et al. 2015; Hirasawa-Fujita et al. 2017). The minor allele [rs1800497(T)] is associated with a reduced number of DA receptors in the brain (Pohjalainen et al. 1998) and increased risk of smoking (Noble et al. 1994; Comings et al. 1996). Interestingly, a neuroimaging study found that individuals carrying this minor allele learned to avoid actions with negative consequences less efficiently (Klein et al. 2007), which can be postulated to influence addictive behaviors. Additionally, variants within the DA transporter, SLC6A3 (rs28363170) have been linked to smoking behaviors and are hypothesized to influence DA transmission (Lerman et al. 1999; Sabol et al. 1999). These polymorphisms collectively are found to underlie individual differences in nicotine-dependence phenotypes such as smoking risk (Lerman et al. 1999), cigarette craving (Erblich et al. 2004, 2005), smoking reward and reinforcement (Perkins et al. 2008), and likelihood of relapse (Sabol et al. 1999).

DRD2 Studies in Mouse Models

Many studies have shown the behavioral impacts of dopaminergic function in rodent models of nicotine dependence as well. In rodents, voluntary self-administration of nicotine leads to increased D2 receptor levels (Novak et al. 2010) in striatal regions of the brain. This observation was also seen in rats after prolonged withdrawal from nicotine, suggesting that elevated D2 levels could be participating in the hypersensitivity following nicotine exposure (Novak et al. 2010). Blocking DA release has been shown to alter the rewarding effects of nicotine as measured by self-administration (Corrigall 1999) and CPP studies (Sun et al. 2014). In rats, site-specific blockade of DA transmission in the VTA specifically reverses the conditioning properties of nicotine from aversive to rewarding (Sun et al. 2014), while systemic and site-specific antagonism of D2 receptors attenuates cue-induced reinstatement of nicotine-seeking behaviors (Liu et al. 2010). In addition to mediating the drug-seeking effects of nicotine, D2 antagonists have also been observed to inhibit nicotine's improvement of memory retrieval in models of stress (Keshavarzian et al. 2018), showcasing D2 receptor function in mediating not just the reinforcing aspects of nicotine dependence but also memory and stress responsivity.

OPRM1 Polymorphisms in Smokers

A significant neurotransmitter system also relevant to nicotine-induced reward is the endogenous opioid system. In smokers, nicotine leads to increase release of β-endorphins, an endogenous µ-opioid receptor (MOR) ligand (Pomerleau et al. 1983). Administration of naloxone, a MOR antagonist, is shown to reduce nicotine reward (Rukstalis et al. 2005). Furthermore, a variant within the coding region of exon 1 of the opioid receptor µ1 gene (OPRM1) (rs1799971) has been identified encoding a nonsynonymous substitution of asparagine (Asn) to aspartic acid (Asp) [Asn40Asp, A > G (A118G)] in the extracellular amino terminus of MOR, resulting in loss of a glycosylation site (Bond et al. 1998; Beyer et al. 2004).

Nicotine-dependence studies investigating rs1799971 genotypes found that female carriers of A/G, G/G alleles are associated with reduced reinforcing value of nicotine, while in males there was no association (Ray et al. 2006). Additionally, carriers of A/G, G/G alleles in a separate study were found to have better smoking cessation outcomes when using transdermal nicotine patches, despite gender (Lerman et al. 2004). Brain imaging studies have reported that carriers of the G allele have larger magnitudes of DA release in response to nicotine smoking than those in the right caudate and right ventral pallidum (Domino et al. 2012), while A allele carriers exhibit higher levels of MOR-binding potential or receptor availability (Ray et al. 2011) and significant increases in cerebral spinal fluid in regions associated previously with cigarette cravings (Wang et al. 2008).

OPRM1 Studies in Rodent Models

Rodent studies have also found the rewarding properties of nicotine to be mediated in part by the MORs (Berrendero et al. 2010; Charbogne et al. 2014). Binding of β-endorphin to MORs on GABAergic interneurons within the NAc decreases inhibitory activity, resulting in disinhibition of dopaminergic neurons and subsequent elevations in DA release (Johnson and North 1992). Both MOR antagonist and MOR KO studies display attenuation of the reinforcing effects of nicotine (Berrendero et al. 2002; Walters et al. 2005; Ismayilova and Shoaib 2010; Liu and Jernigan 2011).

Great technological advancement has been made in engineering humanized mouse models to study the function of polymorphisms within the OPRM1 gene. Mague et al. (2009) generated a mutant mouse line that possessed the mouse equivalent (A112G, N38D) of the human SNP (rs1799971). Studies using this mouse model were designed to evaluate the mechanism underlying the changes associated with the human OPRM1 A118G SNP. Their findings demonstrated that mice harboring the A112G SNP display several phenotypic similarities to humans, including reduced messenger RNA (mRNA) expression of MOR and morphine-mediated antinociception (Mague et al. 2009). Further biochemical experiments demonstrated that this SNP results in reduced N-glycosylation, stability (Huang et al. 2012), and expression of MOR protein (Wang et al. 2012), as well as altered hippocampal function (Mague et al. 2015). Additionally, Ramchandani et al. (2011) generated a murine model of OPRM1 A118G SNP (rs1799971) by replacing the mouse exon 1 with the human exon 1 carrying the A118 or G118 allele through site-directed mutagenesis. This mouse model has been used extensively in modeling addiction of multiple drugs of abuse, including nicotine (Bernardi et al. 2016), alcohol (Bond et al. 1998; Bilbao et al. 2015; Henderson-Redmond et al. 2018), cocaine (Freet et al. 2015), and opioids (Robinson et al. 2015; Freet et al. 2018). These humanized mouse models offer a broad utility in the evaluation and prediction of impacts that genetic variation can have on addiction phenotypes.

CONCLUSIONS

Smoking is a massive public health problem with a great need for research to improve treatment outcomes. While commendable strides have been made in reducing smoking initiation and improving smoking cessation rates, current available smoking cessation treatment options are still mildly efficacious. In response to the need of better treatment options, the primary goal of ongoing research has been to further the understanding of the neural substrates and genetic influences involved in nicotine's effects and reassess potential drug targets. Functional neuroimaging studies have identified key brain structures involved with nicotine-dependence phenotypes such as craving, impulsivity, withdrawal symptoms, and smoking cessation outcomes. Following up with these findings, rodent-modeling techniques have made it possible to dissect the neural circuits involved in these motivated behaviors and ascertain mechanisms underlying nicotine's interactive effects on brain structure and function. Furthermore, translational studies investigating SNPs within the cholinergic, dopaminergic, and opioid systems have found high levels of involvement of these neurotransmitter systems in regulating the reinforcing aspects of nicotine. These findings and coordinated efforts between human and rodent studies pave the way for future work, determining gene-by-drug interactions and tailoring treatment options to each individual smoker.