Functional brain imaging of tobacco use and dependence

Abstract

While most cigarette smokers endorse a desire to quit smoking, only about 14% to 49% will achieve abstinence after 6 months or more of treatment. A greater understanding of the effects of smoking on brain function may (in conjunction with other lines of research) result in improved pharmacological (and behavioral) interventions. Many research groups have examined the effects of acute and chronic nicotine/cigarette exposure on brain activity using functional imaging; the purpose of this paper is to synthesize findings from such studies and present a coherent model of brain function in smokers. Responses to acute administration of nicotine/smoking include: a reduction in global brain activity; activation of the prefrontal cortex, thalamus, and visual system; activation of the thalamus and visual cortex during visual cognitive tasks; and increased dopamine (DA) concentration in the ventral striatum/nucleus accumbens. Responses to chronic nicotine/cigarette exposure include decreased monoamine oxidase (MAO) A and B activity in the basal ganglia and a reduction in α₄β₂ nicotinic acetylcholine receptor (nAChR) availability in the thalamus and putamen. Taken together, these findings indicate that smoking enhances neurotransmission through cortico-basal ganglia-thalamic circuits either by direct stimulation of nAChRs, indirect stimulation via DA release or MAO inhibition, or a combination of these factors. Activation of this circuitry may be responsible for the effects of smoking seen in tobacco dependent subjects, such as improvements in attentional performance, mood, anxiety, and irritability.

1. Introduction

Approximately 23% of Americans smoke cigarettes (Balluz et al., 2004). While most smokers endorse a desire to quit (Fiore et al., 2000), very few will actually quit smoking without treatment, and only about 14–49% will achieve abstinence after 6 months or more of effective treatment (Holmes et al., 2004; Hughes et al., 1999; Hurt et al., 1997; Jorenby et al., 1999; Killen et al., 1999, 2000). Because cigarette smoking carries both considerable health risks (Bartal, 2001; Mokdad et al., 2004) and high societal costs (Leistikow et al., 2000a,b), there is an urgent need for improved treatments for this condition. Functional brain imaging (in conjunction with other lines of research) holds great promise for elucidating both brain circuits and molecular targets that mediate the acute effects of cigarette smoking and chronic effects of tobacco dependence. A greater understanding of brain function associated with smoking may result in improved pharmacological (and behavioral) interventions.

Many functional brain imaging studies of tobacco use and dependence have been performed, using four primary imaging modalities: (1) functional magnetic resonance imaging (fMRI), (2) positron emission tomography (PET), (3) single photon emission computed tomography (SPECT), and (4) autoradiography. These imaging modalities have been used to determine relationships between brain function and effects of acute and chronic cigarette smoking and of smoking-related behaviors. For this review paper, the MEDLINE database was searched using keywords for the four imaging techniques mentioned above cross-referenced with the words “nicotine,” “cigarette,” and “tobacco.” Only data driven functional imaging studies are included in this review, and reference lists within papers found on MEDLINE were also examined and relevant studies included here. In order to maintain focus in this review paper, functional imaging techniques that provide measures of blood flow and metabolism (which are closely related under normal conditions (Paulson, 2002)) are combined under the general heading of brain activity (including fMRI and certain types of SPECT, PET, and autoradiography studies). Also, in order to build a cohesive model of brain activity responses to acute and chronic smoking, nicotine and cigarette studies will be reviewed together, while recognizing that cigarette smoke has many constituents other than nicotine (Baker et al., 2004; Fowles and Dybing, 2003).

The purpose of this paper is to synthesize findings from functional brain imaging studies of tobacco use and dependence, and present a coherent model of brain function in smokers. Acute brain responses to nicotine/smoking will be reviewed first, followed by chronic responses to nicotine/smoking, and concluding with a discussion of these imaging findings in the context of neuroanatomical work and the clinical effects of smoking in tobacco dependent subjects.

2. Brain function responses to acute nicotine administration and cigarette smoking

2.1. Brain activity responses to nicotine/cigarette administration

Many functional brain imaging studies have been performed examining the effects of administration of nicotine or cigarette smoking compared with a placebo or control state (Table 1). Though a wide range of brain regions have been reported to have altered activity in response to nicotine or cigarette smoking, several global and regional findings have been replicated, leading to general conclusions about the acute effects of nicotine or smoking on brain activity.

Table 1.

Functional brain imaging studies of nicotine or cigarette administration

Authors	Subjects	Method	Intervention	Results
London et al (London et al., 1988b; London et al., 1988a)	Rats	2-deoxy-d-[1-14C]glucose autoradiography	SC nic (0.1 to 1.75 mg/kg)	↑nicotine rich regions, including thal, cereb, visual system, others
Rourke et al (Rourke et al., 1997)	8 smokers; 8 former smokers; 17 non-smokers	iodine-123 iodoamphetamine (IMP) SPECT	Smokers smoked the morning of the scan; other groups did not	↓cortical uptake of IMP (a measure of blood flow) in current smokers compared to other groups
Stein et al (Stein et al., 1998)	16 smokers	fMRI	IV nic (0.75- 2.25 mg/70 kg wt) vs. placebo	↑R NAc and bilateral amyg, cingulate, frontal lobes, thal, others
Marenco et al (Marenco et al., 2000)	Rats- chronically nic exposed vs. nic naive	2-deoxy-d-[1-14C] glucose autoradiography	SC nic (0.4 mg/kg) vs. saline	↑thal, superior colliculus in chronically exposed; ↑thal, superior colliculus, medial habenula and dorsal lateral geniculate in nic naive
Domino et al (Domino et al., 2000a)	18 smokers	15O-PET	Nic nasal spray vs. pepper spray	↑thal, pons, visual cortex, cereb
Domino et al (Domino et al., 2000b)	11 smokers	FDG-PET	Nic nasal spray vs. pepper spray	Small ↓global; ↑ L IFG, L PC, R thal, visual cortex; ↓-normalized L ins and R inf occ ctx
Zubieta et al (Zubieta et al., 2001)	18 smokers	15O-PET	Nic nasal spray vs. pepper spray	↑ anterior thal; ↓-L ant temp and R amyg
Rose et al (Rose et al., 2003)	34 smokers	15O-PET	Cigarette vs. no nic control conditions	↑ L frontal factor (incl prefrontal and ACC), ↓-L amyg rCBF
Yamamoto et al (Yamamoto et al., 2003)	10 smokers	99mTc-ECD SPECT	Cigarette vs. abstinence	↓global blood flow
Stapleton et al (Stapleton et al., 2003a)	4 smokers; 2 non-smokers	2 FDG-PETs (fully quantified)	IV nic (1.5 mg) versus placebo	↓global and most regions studied

All regional changes represent normalized activity, unless otherwise stated. SC = subcutaneous; nic = nicotine; thal = thalamus; cereb = cerebellum; SPECT = single photon emission computed tomography; fMRI = functional magnetic resonance imaging; IV = intravenous; R = right; L = left; NAc = nucleus accumbens; amyg = amygdala; FDG = ¹⁸F-fluorodeoxyglucose; PET = positron emission tomography; IFG = inferior frontal gyrus; PC = posterior cingulate; ins = insula; inf occ ctx = inferior occipital cortex; ant = anterior; temp = temporal lobe; ACC = anterior cingulate cortex.

One common finding is that administration of nicotine (Domino et al., 2000b; Stapleton et al., 2003b) or cigarette smoking (Yamamoto et al., 2003) during scanning results in decreased global brain activity. Similarly, smokers who smoke ad lib prior to SPECT scanning (including the morning of the scan) have decreased global brain activity compared to former smokers and non-smokers (Rourke et al., 1997). These findings are generally supported by studies using transcranial Doppler ultrasound or the Xe 133 inhalation method to measure responses to smoking, with some (Cruickshank et al., 1989; Kubota et al., 1983, 1987; Rogers et al., 1983), but not all (Kodaira et al., 1993; Terborg et al., 2002), studies showing diminished cerebral blood flow.

A large (n = 86) recent study (Fallon et al., 2004) further characterized this decreased global activity with nicotine administration. ¹⁸F-fluorodeoxyglucose (FDG) PET was performed while smokers and ex-smokers performed the Bushman aggression task (designed to elicit an aggressive state) and wearing either a 0, 3.5, or 21 mg nicotine patch. Smokers who were rated high on the personality trait hostility had widespread cerebral metabolic decreases while wearing the 21 mg patch and performing the aggression task. Low hostility smokers did not have these changes during PET, suggesting that personality profile may determine which smokers have global metabolic decreases in response to nicotine.

In studies examining regional activity responses to nicotine or smoking, the most common findings are relative increases in activity in the prefrontal cortex (including the dorsolateral prefrontal cortex, and inferior frontal, medial frontal, and orbitofrontal gyri) (Domino et al., 2000b; Rose et al., 2003; Stein et al., 1998), thalamus (Domino et al., 2000a; Domino et al., 2000b; London et al., 1988a,b; Stein et al., 1998; Zubieta et al., 2001), and visual system (Domino et al., 2000a; Domino et al., 2000b; London et al., 1988a,b) (Jed Rose, Personal Communication). Additionally, a Xe 133 inhalation study reported increases in frontal lobe and thalamic blood flow in smokers who smoked a cigarette (Nakamura et al., 2000). The human studies here examined cigarette smokers, while the animal studies here used non-dependent rats, with strong concordance of findings between these sets of studies. Functional brain imaging studies of nicotine or cigarette administration to human non-smokers have not yet been reported, and would be important for a more complete understanding of the effects of tobacco on brain activity. While this group of studies demonstrate specific regional activation with nicotine or smoking, they also imply activation of cortico-basal ganglia-thalamic brain circuits (Alexander et al., 1990) that mediate the subjective effects of smoking (see Section 4).

Since regional activity was normalized to whole brain activity in at least some of these studies, and whole brain activity has been found to decrease with nicotine or cigarette administration (cited above), the regional findings presented here may represent either increased regional activity, or possibly, less of a decrease in regional activity than in other brain areas. Regional decreases in activity are generally not seen with nicotine or cigarette administration, though at least two studies found relatively decreased activity in the left (Rose et al., 2003) and right (Zubieta et al., 2001) amygdala.

2.2. Effect of nicotine on brain activation during cognitive tasks

The most commonly replicated cognitive effect of nicotine administration is improved performance on tasks that require vigilant attention in nicotine-dependent smokers (Newhouse et al., 2004). Nicotine administration also has been reported to improve reaction time (regardless of smoking status) as well (Ernst et al., 2001a). Consistent with these findings are studies which demonstrate that acute abstinence from smoking (within 12 h) results in slowed response times (Bell et al., 1999; Gross et al., 1993; Thompson et al., 2002).

In examining brain mediation of the cognitive effects of smoking, several groups have performed functional imaging studies in subjects performing cognitive tasks during administration of nicotine (compared to a control condition) (Table 2). For most of these studies, subjects performed a cognitive task that involved visual recognition and working memory, such as the n-back task. Results of these studies have been somewhat mixed, showing both decreased (Ernst et al., 2001b; Ghatan et al., 1998) and increased (Jacobsen et al., 2004; Kumari et al., 2003) ACC activation in response to nicotine administration while performing the task. Brain activation responses to nicotine during cognitive tasks have been more consistent in other brain areas such as the thalamus (Jacobsen et al., 2004; Lawrence et al., 2002) and visual cortex (Ghatan et al., 1998; Lawrence et al., 2002), while nicotine had no effect on the visual cortex during photic stimulation (Jacobsen et al., 2002). This last finding indicates that nicotine activates the visual cortex only during demanding visual tasks, rather than simple stimulation.

Table 2.

Functional brain imaging studies of nicotine or cigarette administration during cognitive tasks/stimulation

Authors	Subjects	Method/task	Intervention	Effect of nicotine during task
Ghatan PH et al (Ghatan et al., 1998)	12 smokers; 6 non- smokers	15O-butanol PET/ computerized maze	IV nic infusion versus abstinence	↓ACC and cerebellum; ↑ occ ctx
Ernst et al (Ernst et al.. 2001b)	11 smokers; 11 former smokers	15O-PET/2-back	2 pieces of 2 mg nic gum vs. placebo gum	↓ACC and PFC activation in smokers
Jacobsen et al (Jacobsen et al., 2002)	9 smokers	fMRI/photic stimulation	IV nic 10 mcg/kg vs. saline	No effect on visual cortex
Lawrence et al (Lawrence et al., 2002)	15 smokers	fMRI/rapid visual information-processing	21 mg nic vs. placebo patch	↑ parietal and occipital ctx,, thal, caudate
Kumari et al (Kumari et al., 2003)	11 non-smoking men	fMRI/n-back	SC nic (1 mg) vs. saline	↑ ACC, superior frontal ctx, superior parietal ctx
Jacobsen et al (Jacobsen et al., 2004)	13 schizophrenic smokers; 13 smokers	fMRI/n-back	28 or 35 mg nic vs. placebo patch	↑ ACC and bilateral thal activation (schizophrenic > non- schizophrenic)

PET = positron emission tomography; IV = intravenous; nic = nicotine; ACC = anterior cingulate cortex; occ ctx = occipital cortex; PFC = prefrontal cortex; fMRI = functional magnetic resonance imaging; thal = thalamus.

2.3. Brain dopamine responses to nicotine and smoking

A common pathway for the positive reinforcement associated with most, if not all, addictive drugs is the brain dopamine (DA) system (Koob, 1992; Leshner and Koob, 1999). Laboratory animal studies demonstrate that DA release in the ventral striatum (VST)/nucleus accumbens (NAc) underlies the reinforcing properties of nicotine (Koob, 1992; Leshner and Koob, 1999). Microdialysis (Damsma et al., 1989; Di Chiara and Imperato, 1988; Pontieri et al., 1996; Sziraki et al., 2001) and lesion (Corrigall et al., 1992) studies in rats indicate that nicotine-induced DA release is strongest in this region, and is more robust than the DA release found in associated structures receiving dopaminergic input, such as the dorsal striatum (Di Chiara and Imperato, 1988). These studies generally used nicotine dosages that simulated human cigarette smoking. Acute exposure to cigarette smoke and nicotine has been found to up-regulate dopamine transporter mRNA in the ventral tegmental area (VTA) and substantia nigra (Li et al., 2004), and chronic exposure to cigarette smoke, more so than chronic nicotine alone, has also been found to up-regulate D₁ and D₂ receptor mRNA in the VST (Bahk et al., 2002). Additionally, many in vitro studies of the VST have reported DA release in response to nicotine (Connelly and Littleton, 1983; Marien et al., 1983; Rowell et al., 1987; Sakurai et al., 1982; Westfall et al., 1983).

Functional brain imaging studies of the DA system (Table 3) corroborate and expand upon these laboratory studies. Striatal DA release in response to a nicotine or cigarette challenge has been demonstrated repeatedly in both non-human primates and humans (Brody et al., 2004a; Dewey et al., 1999; Marenco et al., 2004; Tsukada et al., 2002), with the majority of these studies using PET and the radiotracer ¹¹C-raclopride (a relatively specific D₂ receptor binder) to demonstrate DA release through radiotracer displacement. These studies have reported a wide range of DA concentration change. In two studies that examined the question directly (Marenco et al., 2004; Tsukada et al., 2002), nicotine was found to result in less radiotracer displacement than amphetamine, while it has also been reported that nicotine-induced DA release is comparable in magnitude to that induced by other addictive drugs (Pontieri et al., 1996). In addition, an association between ¹¹C-raclopride displacement and the hedonic effects of smoking (defined as elation and euphoria) has been demonstrated (Barrett et al., 2004), though this study did not find an overall difference between the smoking and non-smoking conditions. Thus, while the majority of studies do provide evidence for nicotine/smoking-induced DA release, there are disparities between studies in the extent of human smoking-induced DA release, leaving this issue currently unresolved. Disparities between these studies may be due to differences in methodology (e.g., nicotine administration versus cigarette smoking) and/or technical complexities in performing such studies. (As an aside, effects of smoking on dopamine projections to the prefrontal cortex (Goldman-Rakic et al., 1989) have not yet been reported with functional brain imaging.)

Table 3.

Functional imaging studies of the effects of nicotine or cigarette smoking on the dopamine (DA) system

Authors	Subjects	Method	Intervention	Results/conclusions
Dewey et al (Dewey et al., 1999)	16 baboons	11C-raclopride PET (double bolus)	IV nic (0.3 mg)	↓DV tracer (indicating ↑ DA concentration) in Nac
Dagher et al (Dagher et al., 2001)	11 smokers; 18 non- smokers	11C-SCH 23390 PET		↓BP in smokers (indicating ↓D1 receptor density) in ventral striatum
Tsukada et al (Tsukada et al., 2002)	4 Macada mulatto monkeys	11C-raclopride PET (B/I)	IV nic (B/I)	Slight ↓BP (indicating ↑ DA concentration) in anesthetized, but not conscious monkeys, in dorsal striatum
Salokangas et al (Salokangas et al., 2000b)	9 smokers; 10 non- smokers	18F-DOPA PET		↑ uptake (indicating ↑ DA activity) in cd and Put of smokers
Krause et al (Krause et al., 2002)	11 smoks w/ADHD; 11 non-smok w/ADHD	[99mTc]TRODAT SPECT		↓ DAT (striatal) in smokers
Staley et al (Staley et al., 2001)	21 smokers; 21 non- smokers	[123I]beta-CIT SPECT		No overall binding difference between smokers and non- smokers; ↑ brainstem 5- HT transporters in male smokers
Marenco et al (Marenco et al., 2004)	5 rhesus monkeys	11C-raclopride PET (double bolus & B/I)	IV nic (0.01 to 0.06 mg/ kg)	↓ BP (indicating ↑DA concentration) in basal ganglia with nic administration
Brody et al (Brody et al., 2004a)	20 smokers	11C-raclopride PET (B/I)	Single cigarette versus no smoking	↓ BP (indicating ↑ DA concentration) in smoking, but not no smoking, condition in L ventral cd and put
Barrett et al (Barrett et al., 2004)	10 smokers	11C-raclopride PET (double bolus)	Smoking every 12 minutes versus no smoking	↓BP correlated with hedonic response to smoking in cd and posterior put

PET = positron emission tomography; IV = intravenous; nic = nicotine; DV = volume of distribution; DA = dopamine; BP = binding potential; B/I = bolus-plus-infusion; cd = caudate; put = putamen; SPECT = single photon emission computed tomography; DAT = dopamine transporter; ADHD = attention deficit hyperactivity disorder; beta-CIT = (2 beta-carbomethoxy-3 beta-(4-iodophenyl)-tropane); 5-HT = serotonin.

Nicotine-induced DA release in the NAc has been reported to be mediated by stimulation of nicotinic acetylcholine receptors (nAChRs) on cells of the ventral tegmental area (VTA) that project to the NAc rather than by nicotinic receptors within the NAc itself (Nisell et al., 1994). Lesioning of mesolimbic VTA neurons projecting to the NAc leads to decreased nicotine self-administration (Corrigall et al., 1992; Lanca et al., 2000). Additionally, the effects of nicotine on the dopaminergic system appear to be modulated by glutamatergic and GABAergic neurons (Picciotto and Corrigall, 2002), with nicotine stimulation of gluatamatergic tracts from the prefrontal cortex to the VTA leading to increased DA neuron firing (Kenny and Markou, 2001) and GABA agonism leading to a dampening of DA neuron responses (Cousins et al., 2002). Recent work indicates that nicotine administration causes prolonged depression of GABAergic firing leading to relatively greater excitatory (glutamatergic) input into the mesolimbic DA system and increased DA neuron firing (Mansvelder et al., 2002).

Other functional imaging studies of the DA system have reported decreased D1 receptor density (Dagher et al., 2001), increased ¹⁸F-DOPA uptake (a marker for increased DA turnover) (Salokangas et al., 2000a), and both decreased (Krause et al., 2002) and no alterations (Staley et al., 2001) in dopamine transporter binding in smokers.

To summarize these studies of the DA system, there is extensive evidence that nicotine administration and smoking result in activation of the brain DA mesolimbic pathway, resulting in increased DA release and turnover in the VST/NAc. Because dopaminergic input to the NAc modulates neurotransmission through cortico-basal ganglia-thalamic circuitry (Haber and Fudge, 1997), smoking-induced increases in DA concentration may explain some of the clinical effects of smoking as discussed below (Section 4).

2.4. Functional imaging of nicotinic acetylcholine receptors

Because stimulation of nicotinic acetylcholine receptors (nAChRs) is intimately linked with effects of smoking, a longstanding and still developing area of research is the labeling of nAChRs using functional brain imaging. Nicotinic acetylcholine receptors are ligand-gated ion channels consisting of α and β subunits (Court et al., 2000; Hogg et al., 2003). At least twelve nAChRs have been identified with the heteromeric α₄β₂ being the most common subtype in the brain and the homomeric α₇ being the next most common. Post-mortem (Benwell et al., 1988; Breese et al., 1997) and laboratory (Yates et al., 1995) studies demonstrate that smokers have widespread up-regulation of nAChRs, likely related to desensitization of these receptors from nicotine exposure. (Many animal studies also demonstrate up-regulation of nAChRs in response to chronic nicotine administration) e.g. (Pauly et al., 1996; Shoaib et al., 1997; Zhang et al., 2002). Thus, nAChRs are a natural target for tracer development in the pursuit of a greater understanding of tobacco dependence and other illnesses with abnormal nAChR levels.

Animal research demonstrates that nicotine binds to nAChRs in the brain to mediate a variety of behavioral states (Lukas, 1998; Paterson and Nordberg, 2000)such as heightened arousal and improved reaction time and psychomotor function (Paterson and Nordberg, 2000). Nicotine administration also produces reward through DA release in the NAc, at least in part through stimulation of nAChRs in the ventral tegmental area (Blaha et al., 1996; Corrigall et al., 1994; Nisell et al., 1994; Yeomans and Baptista, 1997; Yoshida et al., 1993). Nicotinic acetylcholine receptors are widespread throughout the brain, with a rank order distribution of nAChR density being: thalamus > basal ganglia > cerebral cortex > hippocampus > cerebellum (Broussolle et al., 1989; Cimino et al., 1992; Clarke et al., 1984; Davila-Garcia et al., 1999; Dávila-García et al., 1997; London et al., 1995, 1985; Pabreza et al., 1991; Pauly et al., 1989; Perry and Kellar, 1995; Valette et al., 1998; Villemagne et al., 1997).

Innovative researchers have developed tracers for the nAChR in recent years, with labeled A-85380 (3-(2(S)-azetidinylmethoxy) pyridine) (Koren et al., 1998) compounds having the most widespread use. Radiolabeling of A-85380 was a major advance in imaging nAChRs, because administration of radiolabeled nicotine (used for previous imaging studies) results in high non-specific binding and short drug–receptor interaction times (Sihver et al., 2000). In recent years, 2-[¹⁸F]F-A-85380 or simply 2-FA and related compounds (Chefer et al., 1999; Horti et al., 1998; Koren et al., 1998) have been developed for PET imaging, and 5-[^123/125I]iodo-A85380 has been used for SPECT imaging (Chefer et al., 1998; Horti et al., 1999; Mukhin et al., 2000) of α₄β₂ nAChRs.

Studies of non-human primates and humans have examined distributions of nAChRs with these new tracers, and found regional densities of these receptors similar to those in the animal work cited above (Chefer et al., 1999, 2003; Fujita et al., 2002; Fujita et al., 2003; Kimes et al., 2003; Valette et al., 1999). In initial human studies, no subjective or cardiovascular effects of 2-FA have been reported; however, studies of tobacco dependent subjects have not yet been published. Finally, two recent studies of baboons examined effects of nicotine or tobacco smoke on nAChR availability. In a 2-FA PET study (Valette et al., 2003), IV nicotine (0.6 mg), inhalation of tobacco smoke from one cigarette (0.9 mg nicotine), and IV nornicotine were all found to reduce the volume of distribution of the tracer by roughly 30–60% in the thalamus and putamen at 80 min, and this reduction of 2-FA binding was relatively long-lived (up to 6 h). Similarly, a 50% reduction in nAChR availability was found with IV nicotine administration to baboons using an epibatidine analog and PET scanning (Ding et al., 2000). Taken together, these studies demonstrate that radiotracers for nAChRs can be administered safely to measure nAChR densities, and that nicotine and smoking substantially decrease α₄β₂ nAChR availability.

2.5. Glutamatergic (and other) effects of nicotine/cigarette smoking

Recent autoradiography studies of rodents are determining effects of nicotine/smoking in brain systems that may be activated by nAChR stimulation. For example, in response to nicotine, glutamate release has been demonstrated in the prelimbic prefrontal cortex (Gioanni et al., 1999), and glutamate and aspartate release have been demonstrated in the VTA (Schilstrom et al., 2000). The finding of nAChR-induced glutamate release in the prefrontal cortex has also been demonstrated by measuring spontaneous excitatory postsynaptic currents (Lambe et al., 2003). Importantly, one of these studies (Gioanni et al., 1999) also demonstrated that nicotine administration facilitates thalamo-cortical neurotransmission through stimulation of nAChRs on glutamatergic neurons.

Other autoradiography studies of rats have demonstrated that chronic administration of nicotine increases glucose transporter (Glut1 and Glut3) densities in an array of brain areas (Duelli et al., 1998) and that chronically administered low dose nicotine is protective against neurodegenerative agents in the striatum (a model for Parkinson’s Disease) (Ryan et al., 2001).

3. Brain function responses to chronic nicotine administration and cigarette smoking

3.1. Functional brain imaging of cigarette craving

Turning to brain imaging of tobacco/nicotine dependence, chronic cigarette smokers experience craving for cigarettes (urge to smoke) within minutes after the last cigarette, and the intensity of craving rises over the next 3–6 h (Jarvik et al., 2000; Schuh and Stitzer, 1995). Cigarette-related cues have been shown to reliably enhance craving during this period, when compared to neutral cues (Carter and Tiffany, 1999).

Two recent studies used a cigarette versus neutral cue paradigm paired with functional imaging to evaluate brain mediation of cigarette craving. In one study (Due et al., 2002), 6 smokers and 6 non-smokers underwent event-related fMRI when presented with smoking images (color photographs) compared with neutral images, for 4 s each. For the smoker group, craving increased during the testing session and exposure to smoking images resulted in activation of mesolimbic (right posterior amygdala, posterior hippocampus, ventral tegmental area, and medial thalamus) and visuospatial cortical attention (bilateral prefrontal and parietal cortex and right fusiform gyrus) circuitry, while the non-smoker group did not have these changes. In the second study (Brody et al., 2002), 20 smokers and 20 non-smokers underwent two FDG-PET sessions. For one PET session, subjects held a cigarette and watched a cigarette-related video, while for the other, subjects held a pen and watched a nature video (randomized order) during the 30-min uptake period of FDG. When presented with smoking-related (compared to neutral) cues, smokers had higher regional metabolism in bilateral anterior cingulate cortex (ACC), left orbitofrontal cortex (OFC), and left anterior temporal lobe. Change in craving scores was also positively correlated with change in metabolism in the OFC, dorsolateral prefrontal cortex, and anterior insula bilaterally.

Taken together, these studies of cigarette craving indicate that immediate responses to visual smoking-related cues (fMRI study) activate the brain reward system, limbic regions, and the visual processing system, while longer exposure to cues (FDG-PET study) leads to activation of the ACC, which mediates anxiety, alertness, and arousal (Chua et al., 1999; Critchley et al., 2001; Kimbrell et al., 1999; Naito et al., 2000; Rauch et al., 1999) and the OFC, which functions in part as a secondary processing center for sensory information (Rolls et al., 1998; Rolls and Baylis, 1994).

In a related preliminary study, seventeen smokers underwent the same FDG-PET craving versus neutral cue protocol as in the second study of craving listed above (Brody et al., 2002) after treatment with a standard course of bupropion HCl (tapered up to 150 mg per oral twice a day for a mean 5.6 weeks). This group of treated subjects had a significant reduction in smoking levels from pre- to post-treatment (mean 27.1 cigs/d pre-treatment to a mean of 3.7 cigs/d post-treatment). Bupropion-treated smokers also had reduced cigarette cue-induced craving and diminished ACC activation when presented with cigarette-related cues, compared to untreated smokers (Brody et al., 2004b). This diminished ACC activation was due to elevated baseline normalized ACC activity in treated smokers, giving an indication that bupropion treatment of smokers increases resting ACC metabolism.

3.2. Functional brain imaging of cigarette withdrawal

Brain activity changes (measured with fMRI) during cigarette withdrawal were recently reported for nicotine-dependent rats (Shoaib et al., 2004). In this study, subcutaneous mecamylamine (1 mg/kg), a nicotine receptor antagonist, was administered to precipitate withdrawal during scanning, and this state was compared to a control state after subcutaneous saline administration. After subcutaneous mecamylamine, nicotine dependent rats had bilateral increases in nucleus accumbens activity compared to the control state.

3.3. Monoamine oxidase function in smokers

Fowler and colleagues have performed a series of elegant studies demonstrating decreases in monoamine oxidase (MAO) A and B activity in cigarette smokers using the PET tracers [¹¹C]clorgyline (Fowler et al., 1996b) and ([¹¹C]L-deprenyl-D2) (Fowler et al., 1996a, 1998b), respectively. When compared to former smokers and non-smokers, average reductions for current smokers are 30% and 40% for MAO A and B (Fowler et al., 2003a). These reductions are the result of chronic smoking behavior rather than a single administration of intravenous nicotine (Fowler et al., 1998a) or smoking a single cigarette (Fowler et al., 1999, 2000), and are less than those seen with antidepressant MAO inhibitors (Fowler et al., 1994; Fowler et al., 1996b). Additionally, a human post-mortem study of chronic smokers demonstrated a modest reduction in MAO A binding that did not reach statistical significance (Klimek et al., 2001). Peripheral MAO B is also reduced in cigarette smokers (Fowler et al., 2003b).

MAO participates in the catabolism of dopamine, norepinephrine, and serotonin (Berlin and Anthenelli, 2001; Fowler et al., 2003a), and it has been postulated that some of the clinical effects of smoking are due to MAO inhibition, leading to decreases in monoamine breakdown with a subsequent increase in monoamine availability (Berlin and Anthenelli, 2001). Thus, smoking may enhance DA availability and the rewarding properties of smoking both through DA release (as described above) and MAO inhibition. Smoking may also alter mood and anxiety through MAO inhibition effects on norepinephrine and serotonin availability and turnover. Comprehensive reviews of the role of MAO in tobacco dependence have recently been published (Berlin and Anthenelli, 2001; Fowler et al., 2003a).

4. Discussion: functional neuroanatomy of tobacco use and dependence

Both acute and chronic effects of nicotine/cigarette exposure have been elucidated with functional brain imaging. Replicated responses to acute administration of nicotine/smoking include: a reduction in global brain activity (perhaps most prominently in smokers with high levels of hostility as a personality trait); activation of the prefrontal cortex, thalamus, and visual system; activation of the thalamus and visual cortex (and possibly ACC) during visual cognitive tasks; and increased DA concentration in the ventral striatum/NAc. Replicated responses to chronic nicotine/cigarette exposure include decreased MAO A and B activity and a substantial reduction in α₄β₂ nAChR availability in the thalamus and putamen (accompanied by an overall up-regulation of these receptors).

This group of findings demonstrates a number of ways in which smoking might enhance neurotransmission through cortico-basal ganglia-thalamic circuits (Alexander et al., 1990) (in addition to demonstrating direct effects of chronic nicotine exposure on nAChR availability) (Fig. 1). Given that the thalamus (Groenewegen et al., 1999; Herrero et al., 2002; Sommer, 2003) and ventral striatum/Nac (Groenewegen et al., 1999; Herrero et al., 2002) function as relay centers for information and for paralimbic and motor processing in the brain, the net effect of smoking may be to enhance neurotransmission along cortico-basal ganglia-thalamic loops originating in prefrontal and paralimbic cortices. Neurotransmission through these circuits may be stimulated directly by the interconnected (Sherman, 2001; Sillito and Jones, 2002) nAChR-rich thalamus and visual systems, and/or indirectly through effects on MAO inhibition and DA release in the ventral striatum/NAc (as well as through nicotine stimulation of excitatory glutamatergic input to the dopamine system (Mansvelder et al., 2002)). In the thalamus, for example, nicotine has direct agonist action on excitatory thalamocortical projection neurons and local circuit neurons, although nicotine also stimulates GABAergic interneurons, so that the relationship between nicotine stimulation and thalamocortical stimulation may be complex (Clarke, 2004). There is mixed evidence as to whether or not nicotine stimulates corticothalamic neurons (Clarke, 2004).

Fig. 1.

Simplified representation of cortico-basal ganglia-thalamic brain circuitry that mediates effects of nicotine/smoking on attentional control, craving, mood and anxiety. Potential targets for nicotine/smoking to enhance attention (and improve craving, mood, and anxiety) include: (1) direct stimulation of nicotinic acetylcholine receptors (nAChRs) in cortex, (2) stimulation of the nAChR-rich thalamus and basal ganglia (which function as relay stations for this circuitry), (3) activation of dopaminergic mesolimbic reward pathways originating in the ventral tegmental area and projecting to the striatum, and (4) monoamine oxidase (MAO) inhibition in the basal ganglia. NAc = nucleus accumbens; VTA = ventral tegmental area.

Enhancement of neurotransmission through prefrontal and paralimbic cortico-basal ganglia-thalamic circuits may account for the most commonly reported cognitive effect of cigarette smoking, namely improved attentional performance (Newhouse et al., 2004), and also related effects, such as improvements in reaction times (Hatsukami et al., 1989; Pritchard et al., 1992; Shiffman et al., 1995), arousal (Parrott and Kaye, 1999), motivation (Powell et al., 2002), and sustained attention (Rusted et al., 2000). Prefrontal (including both dorsolateral and ventrolateral) (Duncan and Owen, 2000; Rees and Lavie, 2001; Smith and Jonides, 1999) and ACC (Carter et al., 1999; Duncan and Owen, 2000; Peterson et al., 1999; Smith and Jonides, 1999) cortices are reported to activate during attentional control tasks (especially visuospatial tasks) (Pessoa et al., 2003). Cigarette smoking may enhance attentional control through direct stimulation of nAChRs within these structures or perhaps through subcortical stimulation of nAChRs in the thalamus and via DA release and/or MAO inhibition in the basal ganglia.

In addition to improvement in attention, smoking improves withdrawal symptoms, such as depressed mood, anxiety, and irritability in tobacco dependent smokers (Cohen et al., 1991; Parrott, 2003), and all of these effects depend (at least in part) on the expectations of the smoker (Perkins et al., 2003). Though nicotine administration generally results in increased activity along prefrontal and paralimbic brain circuits, it is interesting that both increased and decreased ACC activation during cognitive task performance has been reported (see Section 2.2). ACC activity has been associated with anxiety and mood, with increased activity being associated with greater anxiety (Chua et al., 1999; Kimbrell et al., 1999) and decreased activity being associated with depressed mood (Drevets et al., 1997). This combination of findings suggests a potential interaction between expectation of the effects of smoking (e.g. mood improvement, anxiety reduction, or decreased irritability) and direction of ACC activity change during cognitively demanding tasks. Perhaps smokers who expect to and do have anxiety alleviation from smoking have deactivation or decreased activation of the ACC while performing cognitive tasks, while those who expect to and do experience mood improvement from smoking have increased activation of the ACC.

In addition to these primary effects of nicotine and smoking, other functional imaging studies reviewed here focus on smoking-related states, such as cue-induced cigarette craving. Such studies are part of a large body of literature examining cue-induced craving for addictive drugs. Studies specific for cigarette cues/craving reveal that exposure to visual cigarette cues immediately activates mesolimbic (ventral tegmental area, amygdala, and hippocampus) and visuospatial cortical attention areas of the brain, and acutely (over a 30 min time period) activate paralimbic regions (ACC and OFC), and that this cue-induced activation may be diminished by a course of bupropion treatment. These results are similar to those of functional imaging studies for drugs other than tobacco (Goldstein and Volkow, 2002; Miller and Goldsmith, 2001). and it has been posited that at least some of the activations seen with cigarette-related cues (cortical attention areas and OFC) are associated with an expectation of smoking in the non-treatment seeking subjects who participated in these studies (Wilson et al., 2004).

In summary, functional brain imaging studies of nicotine/cigarette smoking have demonstrated a link between nicotine/cigarette administration and brain circuitry that mediates visuospatial attentional processing and withdrawal symptoms. Future studies utilizing newer PET tracers and enhanced MRI techniques will undoubtedly further elucidate the brain mediation of tobacco dependence, and may accelerate the development of targeted smoking cessation therapies.