Cortical control of VTA function and influence on nicotine reward

Biochemical Pharmacology xxx (2013) xxx–xxx

G Model

BCP-11700; No. of Pages 8

Review

Cortical control of VTA function and influence on nicotine reward

Jie Wu a,b,*, Ming Gao a, Jian-Xin Shen b, Wei-Xing Shi c, Andrew M. Oster d, Boris S. Gutkin e

a Divisions of Neurology, Barrow Neurological Institute, St. Joseph’s Hospital and Medical Center, Phoenix, AZ 85013-4496, USAb Departments of Physiology, Shantou University Medical College, Shantou, Guangdong, Chinac Department of Pharmaceutical and Basic Sciences, Loma Linda University Schools of Pharmacy and Medicine, Loma Linda, CA 92350, USAd Department of Mathematics, Eastern Washington University, Cheney, WA 99004, USAe Group for Neural Theory, LNC INSERM U960, Department d‘Etudes Cognitive, Ecole Normale Superieure, 29, rue d‘Ulm, 75005 Paris, France

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

2. Cortical control of VTA neurons: anatomical and functional evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

3. Neuronal slow oscillation (SO) as an important indicator of PFC–VTA functional coupling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

4. Role of computational modeling in order to understand cortical control of nicotine reward. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

5. How does PFC–VTA coupling mediate nicotine reward? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

6. Hypothesis of two systems for drug addiction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

7. Prospective and future directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 000

A R T I C L E I N F O

Article history:

Received 1 May 2013

Accepted 16 July 2013

Available online xxx

Keywords:

Prefrontal cortex

Nicotinic acetylcholine receptor

Nicotine reward

Ventral tegmental area

Dopamine neurons

A B S T R A C T

Tobacco use is a major public health problem. Nicotine acts on widely distributed nicotinic acetylcholine

receptors (nAChRs) in the brain and excites dopamine (DA) neurons in the ventral tegmental area (VTA).

The elicited increase of DA neuronal activity is thought to be an important mechanism for nicotine

reward and subsequently the transition to addiction. However, the current understanding of nicotine

reward is based predominantly on the data accumulated from in vitro studies, often from VTA slices.

Isolated VTA slices artificially terminate communications between neurons in the VTA and other brain

regions that may significantly alter nicotinic effects. Consequently, the mechanisms of nicotinic

excitation of VTA DA neurons under in vivo conditions have received only limited attention. Building

upon the existing knowledge acquired in vitro, it is now time to elucidate the integrated mechanisms of

nicotinic reward on intact systems that are more relevant to understanding the action of nicotine or

other addictive drugs. In this review, we summarize recent studies that demonstrate the impact of

prefrontal cortex (PFC) on the modulation of VTA DA neuronal function and nicotine reward. Based on

existing evidence, we propose a new hypothesis that PFC–VTA functional coupling serves as an

integration mechanism for nicotine reward. Moreover, addiction may develop due to nicotine perturbing

the PFC–VTA coupling and thereby eliminating the PFC-dependent cognitive control over behavior.

� 2013 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Biochemical Pharmacology

jo u rn al h om epag e: ww w.els evier .c o m/lo cat e/b io c hem p har m

Abbreviations: nAChR(s), nicotinic acetylcholine receptor(s); PFC, prefrontal cortex;

VTA, ventral tegmental area; DA, dopamine; ACh, acetylcholine.

* Corresponding author at: Division of Neurology, Barrow Neurological Institute,

St. Joseph’s Hospital and Medical Center, 350 West Thomas Road, Phoenix, AZ

85013-4496, USA. Tel.: +1 602 406 6376; fax: +1 602 406 7172.

E-mail address: [email protected] (J. Wu).

Please cite this article in press as: Wu J, et al. Cortical control of VTA(2013), http://dx.doi.org/10.1016/j.bcp.2013.07.013

0006-2952/$ – see front matter � 2013 Elsevier Inc. All rights reserved.

http://dx.doi.org/10.1016/j.bcp.2013.07.013

1. Introduction

Tobacco use is a major public health problem. Nicotine acts onwidely distributed nicotinic acetylcholine receptors (nAChRs) inthe brain and excites dopamine (DA) neurons in the ventraltegmental area (VTA), which elevates DA release from VTA to thenucleus accumbens (NA) and the prefrontal cortex (PFC) with bothnicotine reward and reinforcement generated as a result [1–3]. Themammalian VTA (A10) is a midbrain region that acts as anintegrative center mediating incentive and motivational effects for

function and influence on nicotine reward. Biochem Pharmacol


mailto:[email protected]


http://www.sciencedirect.com/science/journal/00062952


J. Wu et al. / Biochemical Pharmacology xxx (2013) xxx–xxx2

G Model


almost all addictive drugs, including nicotine. The VTA DA neuronsand their associated ascending projections to their targets (NA andPFC) comprise the well-characterized mesolimbic and mesocor-tical pathways. The VTA receives both direct and indirectexcitatory glutamatergic as well as indirect inhibitory GABAergicinputs from the mPFC [4,5]. During cigarette smoking, nicotinerapidly acts on widely distributed nicotinic acetylcholine receptors(nAChRs) in the brain, which in turn, alters a number of neuronalcircuits and leads to an increase in VTA DA neural firing andelevates the extracellular dopamine levels in DA targets, such asthe NA and the medial prefrontal cortex (mPFC). This process islikely a key pathway responsible for nicotine reward andreinforcement [6]. Although it is well known that the PFCefficiently controls VTA function [7–11], the impact of this corticalcontrol in nicotine reward is not fully understood [12]. In thisreview, we summarize recent advances that validate the role of thePFC in modulation of VTA neuronal function and nicotine reward.

2. Cortical control of VTA neurons: anatomical and functionalevidence

Anatomically, mPFC pyramidal neurons send descendingglutamatergic projections to the VTA both directly and indirectly.It is known that VTA DA neurons receive glutamatergic innerva-tions from the mPFC and that the enhancement of thisglutamatergic pathway underlies drug addiction [13,14]. However,emerging data indicate that most mPFC projections do not directlyterminate on VTA DA neurons [15,16], but rather form directcontacts onto GABAergic cells [7], and then, indirectly innervateVTA DA neurons. Therefore, the mPFC modulates VTA DA neuronalfunction through complex mechanisms including direct (and/orindirect) excitatory and indirect inhibitory innervations on VTA DAneurons.

Functionally, the mPFC and VTA are closely coupled. Stimula-tion of the mPFC increases burst firing in VTA DA neurons, whileinhibition of the PFC induces the opposite effect [7–11]. Undernon-stimulation conditions, the activity of VTA DA neurons co-varies with mPFC neuronal activity [17]. A slow oscillation (ofapproximately 4 Hz) has been identified that is thought toadaptively synchronize the PFC, VTA and hippocampal activities[18]. Further experimental studies, using simultaneous dual fieldpotential recordings from the mPFC and the VTA, showed that slowoscillations (0.5–1.5 Hz) in the mPFC lead those in the VTA byapproximately 30 ms, suggesting that these slow oscillations fromthe mPFC propagate to the VTA [17]. This idea is further supportedby experiments, in which, a dysfunction of the mPFC abolished theneuronal slow oscillations within the VTA. These lines of evidencesuggest that the mPFC control of VTA neuronal firing may bethrough a slow oscillatory pattern [17]. In addition, accumulatingstudies demonstrate that the mPFC–VTA circuit plays a key role in

Fig. 1. Two functional states of VTA DA neurons in vivo. (A) Representative typical traces

from the VTA, shown in (Aa) and (Ab), respectively. Identification of DA neurons was pe

spontaneous spiking, and the bottom trace shows the corresponding smoothed 50 ms

oscillation power of high slow oscillation (red) and low slow oscillation (black) neurons.

oscillations. (D) In VTA slices, all recorded DA neurons are low slow oscillation neuron


the attribution of incentive value to addictive drug-associatedcues. The mPFC is an associative brain region that contributes tocognitive processes such as attention, spatial learning, behavioralplanning and working memory [19]. Nicotine, acting on mPFCnAChRs, enhances working memory and attention [20–22]. Thereis strong evidence suggesting that mPFC is instrumental to manydimensions of drug abuse, including the expression of behavioralsensitization to psychostimulants [23–25] and drug-primedreinstatement of drug-seeking behavior [26]. Therefore, the mPFCprovides functionally important input to the VTA neurons, whichmay directly control their function. Moreover, systemic exposureto nicotine alters mPFC–VTA coupling, which may also underlie anexecutive mechanism for both nicotine reward and reinforcement.

3. Neuronal slow oscillation (SO) as an important indicator ofPFC–VTA functional coupling

1) What is the slow-oscillation? VTA DA neurons exhibit robustoscillatory activity in anesthetized animals over a rangefrequencies (0–10 Hz). After spectral analysis, oscillations be-tween 0.5 and 1.5 Hz were found to be most dominant in VTA DAneurons, with this range defined as the slow oscillation (SO) [27].Neurons demonstrating considerable spontaneous SO constituted50–70% of all VTA DA neurons under resting conditions; we referto these neurons as high SO neurons. The remaining neurons arecollectively referred to as non-SO or low-SO neurons.

2) How is the SO produced? Experimental results from rat brainslices and rats with forebrain hemisections show a lack of SO,suggesting that forebrain input plays an important role ingenerating the SO (Fig. 1). Consistent with this possibility, SOhas been observed in mPFC pyramidal neurons [28,29] and inNA neurons [30–33]. Both structures are located in the forebrainand both regions project to VTA DA neurons. Peters et al. showthat local field potential recordings in the VTA display a SO thatwas highly synchronized with oscillations observed in the mPFC[34]. The findings of Peters et al. and Gao et al. suggest that theSO in VTA neurons is related to the activity in the mPFC [17,34].

3) What is the significance of neuronal SO? The frequency of the SO isin the low delta band. At the electroencephalographic (EEG)level, delta rhythms are most apparent in slow wave sleep andin general anesthesia but are also detected in wakeful states. Forexample, human cortical delta oscillations are significantlyincreased while performing the Wisconsin Card Sorting Test[35]. In the NA of freely moving rats, delta rhythms are highestin amplitude and regularity during wakeful immobility and facewashing [35]. Oscillatory firing (1 Hz) of VTA DA neurons maylead to increased DA release [36]. Oscillations may also play arole in somatic and dendritic information processing in DAneurons. Neuronal oscillations functionally regulate inputselection, facilitation of synaptic plasticity, and the promotion

of high slow oscillation (h-SO) and low slow oscillation (l-SO) DA neurons recorded

rformed using a well-documented protocol. The upper trace illustrates segments of

bin-width firing-rate histograms. (B) Spectral analysis showed an averaged slow-

(C) Local infusion of TTX (10 ng/ml in PBS, 2 ml) into the mPFC abolished high slow

s (n = 16).



J. Wu et al. / Biochemical Pharmacology xxx (2013) xxx–xxx 3

G Model


of synchrony between coupled neurons [37]. The potentialsignificance of the SO is also highlighted by the fact that the SOcan be differentially modulated by addictive drugs [38]. Thus,neuronal SO is potentially an important index for the functionalcoupling of PFC and VTA neurons. The evaluation of the effects ofaddictive drugs on the SO may provide significant informationin drug-induced alterations of neuronal network function.

4. Role of computational modeling in order to understandcortical control of nicotine reward.

Nicotine reward is inexorably tied to the endogenous actions ofacetylcholine. Several general approaches to understanding its rolehave been developed, with some efforts focusing specifically on thefunctional mechanisms of nicotine reward. The benefit ofcomputational modeling, in this context, is to build upon ourknowledge of the specific receptor mechanisms and to inferfunctional consequences of such. Here we review a selection ofsuch modeling studies:

(1) A top-down computational view of cholinergic neuromodula-tion: uncertainty and ACh. Top-down functional approaches tomodeling the consequence of neuromodulators come frommachine learning literature. Primarily, theories have built onreinforcement learning algorithms that are capable of model-ing behavioral conditioning [39]. A functional top-downmethodology has been applied to the role of ACh, where Yuand Dayan utilized a Bayesian statistical framework thatincorporated ACh as a measure of the uncertainty of a stimulusin a given task [40]. That is, ACh signifies contextualuncertainty between top-down, bottom-up, and context-bound processing. In layman’s terms, ACh tracks and signalsthe ‘expected’ uncertainty associated with an environment.They found that high levels of ACh resulted in an improvedestimation of the stimulus via top-down processing. In furtherwork, they went on to test their Bayesian statistical modelframework with ACh level representing the degree of expecteduncertainty as it relates to the context-dependent Posner task[41]. Nicotine acts through the same receptor pathways as theACh, hence, within this framework, nicotine would functionallyalter the estimation of ‘expected’’ uncertainty – biasing it tohigher levels and hence altering behavior [39].

(2) A dynamical view of cholinergic neuromodulation of neuronalnetwork function: memory read-in and recall. Computationalapproaches to study how ACh influences neural populationdynamics have recently received significant attention in theliterature. ACh has been equated with attention and, at times,network models of attentional modulation make a heuristicargument that parameter changes reflecting attentionalmodulation (e.g., [42]) are due to increases in ACh inputs tothe cortex. Notably, attentional modulation has been modeledas an increase in the excitability of local neuronal populations,with the justification that such an increase is due to muscarinicAChR-dependent down-regulation of slow potassium channels(e.g., the M current) [43]. Alternatively, attentional modulationhas also been modeled as an increase in a top-down excitatoryinput [44]. In addition to the above mechanisms, cholinergicinfluence has been suggested to give rise to selectiveglutamatergic neurotransmission. Synaptic modulatory effectsof ACh are thought to be mediated through nicotinic receptors.In this framework, ACh would bias the dynamics of the networkbetween memory encoding and memory consolidation. Obvi-ously, nicotine, acting through nAChRs, would disturb thisprocess, potentially preferentially consolidating smokingrelated memories and behavioral patterns over others.


(3) Large-scale neurodynamical framework for nicotine in DAsignaling and action-selection: Gutkin and colleagues intro-duced a neuro-computational framework for nicotine addic-tion that integrated nicotine effects on the DA neuronpopulation at the receptor population level (signaling re-ward-related information), together with a simple model ofaction-selection (Fig. 2) [45]. This model incorporated a noveldopamine-dependent learning rule that gives distinct roles tothe phasic and tonic dopamine neurotransmission. The authorsstrove to tease out the relative roles of the positive (rewarding)and opponent processes in the acquisition and maintenance ofdrug taking behavior, as well as, the development of suchbehavior into a rigid habit. The details of the mathematicalmethods, equations, and simulation details can be found in[45]. The model showed that the transition to addiction (herespecifically modeled as self-administration of nicotine) iscritically dependent on complementary mechanisms: theability of nicotine to evoke a phasic dopamine response theleads to over-learning and behaviorally apparent over-valua-tion of drug-seeking choices, and tonic dopamine control thatin turn gates cortical plasticity, and the subsequent drug-induced opponent process that withdraws tonic dopamine andstops plasticity – thereby instilling the habitual drug-seekingbehavior. In this sense, this model marries together the positiveand the negative reinforcement theories of addiction, and itpredicts a generalized learning deficit in addicted individuals.Recently this global approach was focused to examine the roleof local VTA circuitry on the nicotine-evoked changes in DA cellactivity, proposing that the endogenous ACh levels in consortwith the descending glutamatergic inputs play a central role indetermining the various mechanisms for nicotine action [39].

5. How does PFC–VTA coupling mediate nicotine reward?

Nicotine stimulates brain reward circuits and promotes therelease of the neurotransmitter dopamine in the reward pathwaysof the brain, which leads to a reinforcement of associatedbehaviors, termed nicotine reward. Nicotine reward is criticalfor the initiation of the resulting nicotine dependence andaddiction when nicotine is repetitively used. Our currentunderstanding of the mechanisms underlying nicotine-inducedexcitation of VTA DA neurons is mainly based on studies usingbrain slices [46–48]. For instance, in VTA slices, nicotine activatesa7-nAChRs on glutamatergic terminals to increase glutamaterelease onto DA neurons with excitatory consequences [47,49,50].At the same time, nicotine activates and desensitizes a4b2nAChRs. Although a4b2 nAChRs are expressed on both VTA DA andGABA neurons, the effects of smoking-relevant concentrations ofnicotine (e.g., 500 nM) on the VTA and DA and GABA neuralresponses to nicotine differ since the endogenous cholinergicinnervations on these two types of VTA neurons differ. Based on aprevious report, most cholinergic terminals innervate GABAneurons, while only few cholinergic terminals innervate the DAneurons [51]. That is, while nearly all DA neurons in the VTAexpress nAChRs, only 5% of the DA neurons actually receivecholinergic projections [52]. One might posit that the receptors onthe DA neurons remain in a non-activated state, while thereceptors on the GABA neurons are tonically activated byendogenous ACh inputs (see discussion in [39]). Thus, nicotineacts on VTA neuronal somatic/dendritic b2-containing nAChRs(e.g., a4b2-nAChRs) to directly depolarize and activate thoseneurons [50,53]. Thereafter, with relevantly long-lasting exposure,nicotine desensitizes a4b2-nAChRs on VTA both DA and GABAneurons, yet the nicotine-induced nAChR desensitization mainlyeliminates the tonic cholinergic excitation of GABA neurons



Fig. 2. A computational model to simulate nicotinic effects on VTA DA neurons. (Aa) A scheme of the VTA circuits and nAChR distribution, in which r is a parameter introduced

in the model to change continuously the dominant site of a4b2 nAChR action. (Ab) Different states of the nAChRs, in which, a indicates an activated state and s indicates a

desensitized state. a (purple lines) and s (orange lines) variables are shown during and after the exposure to a constant nicotine concentration of 10 mM (c) and 100 mM (d) for

200 ms starting at t = 50 ms. The normalized receptor activation (a–s) is shown in blue. Peak current and net charges mediated during the exposure are illustrated in the

panels. The insets show the dynamics of the same variables on a longer time scale. (Ba) Disinhibition of the DA circuit can be achieved in the computational model. Here, ACh is

taken to be at a high level and the expression level of the a4b2 nAChR subtype is more heavily expressed by the VTA GABA cells. However, in Bb, direct excitation of the DA

circuit occurs when ACh is low and the DA cells have a high degree of a4b2 nAChR subtype. These results are interesting when compared to experiment findings (Bc,d), where

similar dynamics were found for the type I and type II DA cell dynamics (associated with high SO and low SO, respectively).


G Model


Please cite this article in press as: Wu J, et al. Cortical control of VTA function and influence on nicotine reward. Biochem Pharmacol(2013), http://dx.doi.org/10.1016/j.bcp.2013.07.013



G Model


because of the high ACh innervation of these GABA neurons [51].Consequently, nicotine reduces the GABA neuron firing, and leadsto an increase in VTA DA neuron firing through a disinhibitionmechanism [48,50,53]. These collective results lead to thehypothesis that nicotine reward is promoted by nicotinicstimulation of the nAChRs through multiple pathways (bothdirect stimulation on the DA neurons and disinhibition via the VTAGABA cells). Activation of these nAChR-associated pathwaysincreases DA release from the VTA to its DA targets: NA and PFC.

On the other hand, systemic exposure to nicotine affects theactivity of DA neurons through mechanisms that are morecomplex, involving the VTA and other brain regions [54,55], likethe mPFC [56,57]. Emerging evidence demonstrates that the mPFCplays an important role in controlling VTA DA neuronal function[58]. The mPFC contributes to cognitive processes such asattention, spatial learning, behavioral planning and workingmemory [19]. Nicotine has been reported to act through mPFCnAChRs to enhance working memory and attention [20–22,59,60].In addition, the mPFC has been suggested to be a key brain regionunderlying the neural mechanisms of drug addiction and craving[61]. Chen et al. examined the effects of PFC dysfunction on the VTAneuronal function and on the nicotinic effects. They found that asystemic injection of nicotine (single i.p. injection) inducedenhanced AMPA/NMDA ratios in VTA DA neurons (after 24 h)with no significant difference between intact and dysfunctionalPFC [62]. They used wavelet analysis to show that in PFC intact rats,systemic nicotine increased energy contents of the 1–1.5 Hzfrequency band in VTA DA neurons, while in PFC dysfunctional rats,nicotine failed to induce a similar effect [63]. In addition, using theLempel-Ziv estimator, this group further found that systemicnicotine triggered a significant increase in the complexity of VTADA neuronal firing patterns when communication between PFCand VTA was present, while transection of PFC obliterated theeffect of nicotine [64]. Together, these studies suggest theimportance of PFC in the control VTA DA neuronal function andnicotinic effects on VTA DA neural firing. Therefore, an importantquestion is whether mPFC participates in (or controls) nicotinereward. We recently addressed this question using electrophysio-logical recording from single VTA DA neurons (extracellular singleunit recording) in anesthetized rats [65].

We found that systemic injection of nicotine (0.5 mg/kg, i.v.)induced two types of changes to the firing rate of VTA DA neurons.One group of neurons exhibited a bi-phasic change of firing rate, aninitial decrease with a subsequent increase in firing rate for aprolonged duration, while another group of neurons exhibited amono-phasic increase of firing rate. We classified the VTA DAneurons as either type-I or type-II according to their response tonicotine with either a biphasic or monophasic response, respec-tively (Fig. 2Bc and Bd). Further analysis showed that the VTA DA

Fig. 3. Effects of systemic nicotine on mPFC pyramidal neuron firing. (A) A typical case of n

rate (FR), bursting fraction (BF) and slow-oscillation power (Pso) after nicotine. *p < 0.05

(PFC LFP) and VTA (VTA LFP). (Ba) A typical response of hSO VTA DA neuron to nicotine. T

(baseline) and 12 and 14 min after nicotine injection.


neurons classified by these two response types exhibited distinctfiring characteristics. Prior to nicotine exposure, the type-I neuronsdisplayed irregular firing patterns with strong variations of theaverage firing frequency and more pronounced slow oscillations,whereas the type-II neurons maintained a regular firing patternwith a low slow oscillation power. These results suggest thepossible existence of two subtypes of VTA DA neurons: differen-tiable by their firing characteristics and response to systemicnicotine. However, there may be only one group of VTA DA neuronswith different functional states that respond differently tosystemic nicotine. Further study is needed in order to distinguishbetween these possibilities.

Beyond identifying two types of DA neurons, we examined theeffects of systemic nicotine with the control of the VTA by PFCdisrupted. We altered mPFC function with local infusion of TTX toboth sides of mPFC or by physical transection of mPFC, then testedthe effects of systemic injection of nicotine. The results demon-strated that with a disruption of the mPFC circuit, nicotine failed toelicit an increase of the firing rate for the type-I neurons, but thetype-II DA neural response remained unperturbed. In addition,after the inactivation of the mPFC, the power of the slow oscillationfor type-I neurons was dramatically reduced (to levels similar tothat of the type-II neurons). These results suggest that the VTAtype-I, but not type-II, DA neurons are functionally coupled withmPFC. Fig. 3 shows typical traces (Fig. 3A) that present the effectsof systemic nicotine on VTA firing rate (top trace) and on both PFCand VTA SO (bottom traces). The results demonstrate that systemicnicotine administration increases VTA DA neuron firing (e.g., at12 min) but reduces SO power during simultaneous dual recordingof field potentials from the PFC and the VTA. These results suggestthat systemic exposure to nicotine perturbs this coupling(indicated by the reduction of the slow oscillations in VTA) andincreases nicotinic excitation of type-I neurons.

We, also, considered whether nicotine activates PFC pyramidalneurons, which would alter the descending cortical input from thePFC to the VTA. To address this question, we performed threeexperiments. (1) We directly recorded pyramidal neuronal firing inPFC layer 5 in anesthetized rats and found that systemic nicotineinitially increased PFC neuron firing, which then declined to a firingrate below the baseline level (Fig. 3B). (2) We locally infused nAChRantagonist mecamylamine into both sides of mPFC and found thatpharmacological inhibition of PFC nAChRs eliminated systemicnicotine-induced excitation in VTA type-I neurons. (3) We locallyinfused 0.5 mM nicotine into the mPFC and examined the responseof VTA type-I DA neurons. We observed that the local nicotineinfusion was insufficient to increase the DA neuronal firing rate butdid enhance the power of the SO. Together, these results suggestthat systemic nicotine acts on nAChRs in PFC neurons, therebyaltering PFC neuronal function, and, in turn, alters the descending

icotinic effects on mPFC pyramidal cells. Inset showed Normalized changes of firing

. **p < 0.01. (B) Simultaneous recordings of field potential (FP) from left side of PFC

he slow-oscillation power (Bb) was simultaneously measured between 0 and 2 min




G Model


information from the PFC to the VTA. This signal from the PFC tothe VTA modulates VTA DA neuronal (type-I) function. However,smoking-relevant levels of nicotine in PFC alone (locally infusedinto PFC) are insufficient to alter VTA DA neuron firing rate in in

vivo conditions. Finally, we evaluated the role of GABA neurons innicotinic modulation of PFC–VTA coupling and nicotine reward.After systemic injection of GABAA receptor antagonist bicuculline(BMI), systemic injection nicotine failed to alter VTA type-I DAneuron firing (both initial negative and followed positive phases).A systemic injection of bicuculline did not alter VTA DA neuronfiring rate and bursting fraction, yet significantly reduced slowoscillation power before nicotine injection. These data suggest thatbicuculline eliminates the PFC–VTA functional coupling. Ourprevious work [17] showed that the PFC negatively couples tothe VTA, suggesting that the PFC may couple to VTA DA neuronsthrough the GABAergic neurons. Based on these lines of evidence, itis likely that GABA neurons between PFC and VTA DA neurons areimportant for nicotinic excitation in DA neurons.

Taken together, systemic nicotine administration significantlyreduces the power of the slow oscillation in type-I VTA DA neuronswhile increasing their firing rate. The findings in our study areconsistent with nicotine acting to functionally uncouple themPFC–VTA circuit. In addition, exposure to BMI lowers type-I VTADA neuronal Pso and blocks both the initial inhibitory and the lateexcitatory effects of nicotine on type-I VTA DA neuronal firing. Thisis consistent with nicotine’s mediatory effects and the mPFC–VTADA neuronal negative coupling via an inhibitory, GABAergicconnection. It has been reported that GABAergic neurons in theVTA fire at rate of higher than 10 Hz and exhibit underlyingoscillations at a frequency range between 1 and 2 Hz [66] and thatVTA GABAergic neurons receive glutamatergic innervations fromthe PFC [15,67]. Thus, VTA GABAergic neurons may play a role inthe transfer of the SO from the mPFC to the type-I VTA DA neurons.The mPFC nAChRs influences the excitatory response of type-I VTADA neurons as is evidenced by the response sensitivity tomecamylamine applied locally in the mPFC [65]. Moreover, theability of BMI administration to prevent the late inhibitory phase ofnicotine’s actions on mPFC neuronal firing suggests that GABA andGABAA receptors, at least partly, mediate that effect. However, theinsensitivity to BMI of the initial excitatory effect of nicotineexposure on mPFC neuronal firing is consistent with an effect bynAChR-mediated activation of release from glutamatergic term-inals in the mPFC [68–70]. These results also suggest that nicotine-induced initial inhibition on type-I DA neuron firing may bemediated by the GABA neurons that may be subject to a degree ofPFC control.

In the mPFC, nicotine exposure culminates in a delayed butlonger-lasting reduction in mPFC neuron firing, likely throughactivation of GABAergic mechanisms [71]. Such a process wouldlead to lasting activation of VTA DA neurons and has beenimplicated in morphine-, muscimol- and benzodiazepine-inducedVTA DA neuron excitation [72,73]. However, our studies suggestthat consideration should be given to the nicotine exposure-induced impairment of inverse coupling between the mPFC andthe VTA via the GABA neurons as defined by diminished power ofthe SO as a novel alternate mechanism for nicotine-induced DAneuron excitation in vivo.

6. Hypothesis of two systems for drug addiction

The major feature of drug addiction is a reduced ability toregulate control over the desire to procure drugs regardless of therisks involved. Traditional models implicated the neural ‘reward’(i.e., the VTA) system as the key brain region in providing aneurobiological model of addiction. Computational models of thelocal VTA circuitry, suggest that a biphasic pattern of response to


nicotine could be mediated by nicotine action on the nAChRs atlocal GABA neurons and the subsequent disinhibition of the DAneurons [39,74]. However, this single rewarding system theory hasbeen expanded as two separate, but interconnected systems, thelimbic system in the incentive sensitization of drugs (VTA) and thePFC in regulating inhibitory cognitive control over drug use. Ourexperimental lines of evidence support this novel two-systemmodel; this is particularly significant when put into the largercontext of drug addiction theory. The development of drugaddiction has been likened to a progressive habitualization ofbehavior and a loss of cognitive control [75,76], possibly through adisconnection of higher cortical influence over reward circuitry[77]. Consistent with the notion that the mPFC plays a role in drugaddiction through the modulation of descending inhibitory system[76], and in line with the concept that the mPFC mediates cognitivecontrol (and executive function), the inverse coupling that weshow here suggests that the mPFC exerts this control throughmulti-synaptic gating of inhibition and excitation of DA transmis-sion and that nicotine selectively disrupts this inhibitory control,leading to a dysfunctional modulation of type-I VTA DA neurons.Furthermore, recent computational theories propose an oppositionof cognitive control versus reward-driven habitual action [78],suggesting that a model-based cortical system can ensure flexibleachievement of long-term goals, whereas a model-free, reward-based, DA system promotes habitual seeking of immediaterewards. The theory proposes that an optimal integration betweenthe two systems for control of behavior is key and may be mediatedby cholinergic signals encoding the uncertainty of the behavioralcontext [41]. Based on our results, we speculate that nicotinewould pathologically remove the influence of the cortical model-based cognitive control system over the DA circuitry and promotehabitual immediate-reward-seeking behavior. Thus, our results,which point to a nicotine-induced uncoupling of the mPFC-mediated inhibitory control over reward circuitry, provide apotential neural basis for the bias toward impulsivity and loss ofcognitive control seen in nicotine-dependent individuals andpossibly other types of drug dependencies.

7. Prospective and future directions

Mechanisms of nicotine reward are very complex includingnicotine-induced alterations from the levels of genes, molecules,synapses, receptors, circuits, networks and behaviors. Althoughextensive efforts have been made, our understanding is stillimmature, especially at the circuit and network levels. In thisarticle, we summarized recent advances aimed at understandingthe roles of the PFC in nicotine-induced excitation of VTA DAneurons. Based on the existing data and computational models, weproposed a new hypothesis of how nicotine modulates PFC–VTAcoupling and how PFC neurons control nicotinic excitation of VTADA neurons. However, there are still many important questionsthat need to be addressed in future studies. (1) In VTA DA neurons,one group is coupled to PFC (type-I) and another group is not (type-II). However, the properties of these two groups of neurons are notwell understood and it remains to be seen whether they belong totwo subtypes or are, in fact, of the same type, but in two differentstates. For example, do they have different locations (ventral ordorsal region in the VTA), different projections (to NA or PFC) ordifferent innervations from PFC? In addition, it is of interest todetermine whether these two groups of VTA DA neurons havedifferent functional roles (e.g., reward vs. aversive). (2) VTA DAneuron SO is a useful index to show the functional couplingbetween PFC and VTA in anesthetized animals, but it is muchweaker in waking animals, thus the significance of this SO in VTADA neurons needs to be carefully validated. (3) Although we havedemonstrated the importance of the PFC in controlling nicotinic




G Model


effects on some VTA DA neurons, the details of neuronal circuits(e.g., direct and indirect innervations) underlying this modulationare unclear. (4) GABA neurons are likely very important for PFC–VTA coupling, but it is uncertain where these GABA neurons arelocated, within the VTA or outside of the VTA (or both). (5)Computational simulation provides a powerful approach forstudying complex systems such as this. Mathematical modelsmay provide important predictions of how PFC exerts control overVTA function and nicotine reward. The present models mainlyfocus on the VTA local circuit, and there is need for thedevelopment of a more complex PFC–VTA model. Nevertheless,recent advancements have furthered our understanding of theintegrated signals between the PFC and the VTA and providedinsight into the impact of the PFC in nicotine reward anddependence. With new conceptual insights, methodologies, andadvances in computational modeling, there is great promise thatthe role of cortical modulations underlying nicotine reward andaddiction will be elucidated.

Acknowledgements

Work toward this project was supported by the BarrowNeurological Foundation and by grants from the Arizona Biomedi-cal Research Commission (0028 and 0057), the Institute for MentalHealth Research, Philip Morris International through their ExternalResearch Program, and the National Institutes of Health (R01NS040417 and R01 DA015389). BSG and AMO were supported inpart by the ANR ‘‘Dopanic’’ grant, and AMO was supported in partby a NERF grant.

References

[1] Wise RA. Dopamine, learning and motivation. Nature Reviews Neuroscience2004;5:483–94.

[2] Mansvelder HD, McGehee DS. Cellular and synaptic mechanisms of nicotineaddiction. Journal of Neurobiology 2002;53:606–17.

[3] Fields HL, Hjelmstad GO, Margolis EB, Nicola SM. Ventral tegmental areaneurons in learned appetitive behavior and positive reinforcement. AnnualReview of Neuroscience 2007;30:289–316.

[4] Kalivas PW. Neurotransmitter regulation of dopamine neurons in the ventraltegmental area. Brain Research 1993;18:75–113.

[5] Charara A, Smith Y, Parent A. Glutamatergic inputs from the pedunculopontinenucleus to midbrain dopaminergic neurons in primates: phaseolus vulgaris-leucoagglutinin anterograde labeling combined with postembedding gluta-mate and GABA immunohistochemistry. Journal of Comparative Neurology1996;364:254–66.

[6] Mathon DS, Kamal A, Smidt MP, Ramakers GM. Modulation of cellular activityand synaptic transmission in the ventral tegmental area. European Journal ofPharmacology 2003;480:97–115.

[7] Tong ZY, Overton PG, Clark D. Stimulation of the prefrontal cortex in the ratinduces patterns of activity in midbrain dopaminergic neurons which resem-ble natural burst events. Synapse 1996;22:195–208.

[8] Gariano RF, Groves PM. Burst firing induced in midbrain dopamine neurons bystimulation of the medial prefrontal and anterior cingulate cortices. BrainResearch 1988;462:194–8.

[9] Svensson TH, Tung CS. Local cooling of pre-frontal cortex induces pacemaker-like firing of dopamine neurons in rat ventral tegmental area in vivo. ActaPhysiologica Scandinavica 1989;136:135–6.

[10] Murase S, Grenhoff J, Chouvet G, Gonon FG, Svensson TH. Prefrontal cortexregulates burst firing and transmitter release in rat mesolimbic dopamineneurons studied in vivo. Neuroscience Letters 1993;157:53–6.

[11] Overton PG, Tong ZY, Brain PF, Clark D. Preferential occupation of mineralo-corticoid receptors by corticosterone enhances glutamate-induced burst firingin rat midbrain dopaminergic neurons. Brain Research 1996;737:146–54.

[12] Tolu S, Eddine R, Marti F, David V, Graupner M, Pons S, et al. Co-activation ofVTA DA and GABA neurons mediates nicotine reinforcement. MolecularPsychiatry 2013;18:382–93.

[13] Takahata R, Moghaddam B. Activation of glutamate neurotransmission in theprefrontal cortex sustains the motoric and dopaminergic effects of phencycli-dine. Neuropsychopharmacology 2003;28:1117–24.

[14] Kalivas PW, Lalumiere RT, Knackstedt L, Shen H. Glutamate transmission inaddiction. Neuropharmacology 2009;56(Suppl 1):169–73.

[15] Carr DB, Sesack SR. Projections from the rat prefrontal cortex to the ventraltegmental area: target specificity in the synaptic associations with mesoac-cumbens and mesocortical neurons. Journal of Neuroscience 2000;20:3864–73.


[16] Sesack SR, Carr DB, Omelchenko N, Pinto A. Anatomical substrates for gluta-mate-dopamine interactions: evidence for specificity of connections andextrasynaptic actions. Annals of the New York Academy of Sciences2003;1003:36–52.

[17] Gao M, Liu CL, Yang S, Jin GZ, Bunney BS, Shi WX. Functional coupling betweenthe prefrontal cortex and dopamine neurons in the ventral tegmental area.Journal of Neuroscience 2007;27:5414–21.

[18] Fujisawa S, Buzsaki G. A 4 Hz oscillation adaptively synchronizes prefrontal,VTA, and hippocampal activities. Neuron 2011;72:153–65.

[19] Fuster JM. Executive frontal functions. Experimental Brain Research 2000;133:66–70.

[20] Granon S, Poucet B, Thinus-Blanc C, Changeux JP, Vidal C. Nicotinic andmuscarinic receptors in the rat prefrontal cortex: differential roles in workingmemory, response selection and effortful processing. Psychopharmacology1995;119:139–44.

[21] Levin ED. Nicotinic systems and cognitive function. Psychopharmacology1992;108:417–31.

[22] Levin ED, McClernon FJ, Rezvani AH. Nicotinic effects on cognitive function:behavioral characterization, pharmacological specification, and anatomic lo-calization. Psychopharmacology 2006;184:523–39.

[23] Wolf ME, Dahlin SL, Hu XT, Xue CJ, White K. Effects of lesions of prefrontalcortex, amygdala, or fornix on behavioral sensitization to amphetamine:comparison with N-methyl-D-aspartate antagonists. Neuroscience 1995;69:417–39.

[24] Li Y, Wolf ME, White FJ. The expression of cocaine sensitization is notprevented by MK-801 or ibotenic acid lesions of the medial prefrontal cortex.Behavioural Brain Research 1999;104:119–25.

[25] Li Y, Hu XT, Berney TG, Vartanian AJ, Stine CD, Wolf ME, et al. Both glutamatereceptor antagonists and prefrontal cortex lesions prevent induction of co-caine sensitization and associated neuroadaptations. Synapse (New York NY)1999;34:169–80.

[26] McFarland K, Lapish CC, Kalivas PW. Prefrontal glutamate release into the coreof the nucleus accumbens mediates cocaine-induced reinstatement of drug-seeking behavior. Journal of Neuroscience 2003;23:3531–7.

[27] Shi WX. Slow oscillatory firing: a major firing pattern of dopamine neurons inthe ventral tegmental area. Journal of Neurophysiology 2005;94:3516–22.

[28] Cowan RL, Wilson CJ. Spontaneous firing patterns and axonal projections ofsingle corticostriatal neurons in the rat medial agranular cortex. Journal ofNeurophysiology 1994;71:17–32.

[29] Lewis BL, O‘Donnell P. Ventral tegmental area afferents to the prefrontal cortexmaintain membrane potential ‘up’ states in pyramidal neurons via D(1)dopamine receptors. Cerebral Cortex 2000;10:1168–75.

[30] Goto Y, O‘Donnell P. Network synchrony in the nucleus accumbens in vivo.Journal of Neuroscience 2001;21:4498–504.

[31] Goto Y, O‘Donnell P. Synchronous activity in the hippocampus and nucleusaccumbens in vivo. Journal of Neuroscience 2001;21:RC131.

[32] Leung LS, Yim CY. Rhythmic delta-frequency activities in the nucleus accum-bens of anesthetized and freely moving rats. Canadian Journal of Physiologyand Pharmacology 1993;71:311–20.

[33] O‘Donnell P, Grace AA. Synaptic interactions among excitatory afferents tonucleus accumbens neurons: hippocampal gating of prefrontal cortical input.Journal of Neuroscience 1995;15:3622–39.

[34] Peters Y, Barnhardt NE, O‘Donnell P. Prefrontal cortical up states are synchro-nized with ventral tegmental area activity. Synapse 2004;52:143–52.

[35] Gonzalez-Hernandez JA, Cedeno I, Pita-Alcorta C, Galan L, Aubert E, Figueredo-Rodriguez P. Induced oscillations and the distributed cortical sources duringthe Wisconsin card sorting test performance in schizophrenic patients: newclues to neural connectivity. International Journal of Psychophysiology2003;48:11–24.

[36] Gonon FG. Nonlinear relationship between impulse flow and dopamine re-leased by rat midbrain dopaminergic neurons as studied by in vivo electro-chemistry. Neuroscience 1988;24:19–28.

[37] Buzsaki G, Draguhn A. Neuronal oscillations in cortical networks. Science2004;304:1926–9.

[38] Shi WX, Pun CL, Zhou Y. Psychostimulants induce low-frequency oscillationsin the firing activity of dopamine neurons. Neuropsychopharmacology2004;29:2160–7.

[39] Graupner M, Gutkin B. Modeling nicotinic neuromodulation from globalfunctional and network levels to nAChR based mechanisms. Acta Pharmaco-logica Sinica 2009;30:681–93.

[40] Yu AJ, Dayan P. Acetylcholine in cortical inference. Neural Networks TheOfficial Journal of the International Neural Network Society 2002;15:719–30.

[41] Yu AJ, Dayan P. Uncertainty, neuromodulation, and attention. Neuron 2005;46:681–92.

[42] Compte A, Wang XJ. Tuning curve shift by attention modulation in corticalneurons: a computational study of its mechanisms. Cerebral Cortex 2006;16:761–78.

[43] Hasselmo ME, Bower JM. Cholinergic suppression specific to intrinsic notafferent fiber synapses in rat piriform (olfactory) cortex. Journal of Neuro-physiology 1992;67:1222–9.

[44] Deco G, Rolls ET. Attention, short-term memory, and action selection: aunifying theory. Progress in Neurobiology 2005;76:236–56.

[45] Gutkin BS, Dehaene S, Changeux JP. A neurocomputational hypothesis fornicotine addiction. Proceedings of the National Academy of Sciences of theUnited States of America 2006;103:1106–11.


http://refhub.elsevier.com/S0006-2952(13)00449-8/sbref0005





































































































































G Model


[46] Dani JA, Heinemann S. Molecular and cellular aspects of nicotine abuse.Neuron 1996;16:905–8.

[47] Mansvelder HD, McGehee DS. Long-term potentiation of excitatory inputs tobrain reward areas by nicotine. Neuron 2000;27:349–57.

[48] Mansvelder HD, Keath JR, McGehee DS. Synaptic mechanisms underlie nico-tine-induced excitability of brain reward areas. Neuron 2002;33:905–19.

[49] Wooltorton JR, Pidoplichko VI, Broide RS, Dani JA. Differential desensitizationand distribution of nicotinic acetylcholine receptor subtypes in midbraindopamine areas. Journal of Neuroscience 2003;23:3176–85.

[50] Pidoplichko VI, Noguchi J, Areola OO, Liang Y, Peterson J, Zhang T, et al.Nicotinic cholinergic synaptic mechanisms in the ventral tegmentalarea contribute to nicotine addiction. Learning and Memory 2004;11:60–9.

[51] Garzon M, Vaughan RA, Uhl GR, Kuhar MJ, Pickel VM. Cholinergic axonterminals in the ventral tegmental area target a subpopulation of neuronsexpressing low levels of the dopamine transporter. Journal of ComparativeNeurology 1999;410:197–210.

[52] Fiorillo CD, Williams JT. Cholinergic inhibition of ventral midbrain dopamineneurons. Journal of Neuroscience 2000;20:7855–60.

[53] Pidoplichko VI, DeBiasi M, Williams JT, Dani JA. Nicotine activates and desen-sitizes midbrain dopamine neurons. Nature 1997;390:401–4.

[54] Mameli-Engvall M, Evrard A, Pons S, Maskos U, Svensson TH, Changeux JP,et al. Hierarchical control of dopamine neuron-firing patterns by nicotinicreceptors. Neuron 2006;50:911–21.

[55] Avale ME, Faure P, Pons S, Robledo P, Deltheil T, David DJ, et al. Interplay ofbeta2* nicotinic receptors and dopamine pathways in the control of sponta-neous locomotion. Proceedings of the National Academy of Sciences of theUnited States of America 2008;105:15991–96.

[56] Erhardt S, Schwieler L, Engberg G. Excitatory and inhibitory responses ofdopamine neurons in the ventral tegmental area to nicotine. Synapse 2002;43:227–37.

[57] Ikemoto S, Qin M, Liu ZH. Primary reinforcing effects of nicotine are triggeredfrom multiple regions both inside and outside the ventral tegmental area.Journal of Neuroscience 2006;26:723–30.

[58] Mansvelder HD, Mertz M, Role LW. Nicotinic modulation of synaptic trans-mission and plasticity in cortico-limbic circuits. Seminars in Cell and Devel-opmental Biology 2009;20:432–40.

[59] Counotte DS, Goriounova NA, Li KW, Loos M, van der Schors RC, Schetters D,et al. Lasting synaptic changes underlie attention deficits caused by nicotineexposure during adolescence. Nature Neuroscience 2011;14:417–9.

[60] Guillem K, Bloem B, Poorthuis RB, Loos M, Smit AB, Maskos U, et al. Nicotinicacetylcholine receptor beta2 subunits in the medial prefrontal cortex controlattention. Science (New York NY) 2011;333:888–91.

[61] Tzschentke TM. Pharmacology and behavioral pharmacology of the mesocor-tical dopamine system. Progress in Neurobiology 2001;63:241–320.

[62] Chen T, Zhang D, Dragomir A, Kobayashi K, Akay Y, Akay M. Investigating theinfluence of PFC transection and nicotine on dynamics of AMPA and NMDAreceptors of VTA dopaminergic neurons. Journal of NeuroEngineering andRehabilitation 2011;8:58.


[63] Chen TY, Zhang D, Dragomir A, Akay Y, Akay M. The effects of nicotineexposure and PFC transection on the time–frequency distribution of VTADA neurons’ firing activities. Medical & Biological Engineering & Computing2011;49:605–12.

[64] Chen TY, Zhang D, Dragomir A, Akay YM, Akay M. Complexity of VTA DA neuralactivities in response to PFC transection in nicotine treated rats. Journal ofNeuroEngineering and Rehabilitation 2011;8:1–8.

[65] Zhang D, Gao M, Xu D, Shi WX, Gutkin BS, Steffensen SC, et al. Impact ofprefrontal cortex in nicotine-induced excitation of ventral tegmental areadopamine neurons in anesthetized rats. Journal of Neuroscience 2012;32:12366–75.

[66] Steffensen SC, Svingos AL, Pickel VM, Henriksen SJ. Electrophysiological char-acterization of GABAergic neurons in the ventral tegmental area. Journal ofNeuroscience 1998;18:8003–15.

[67] Geisler S, Wise RA. Functional implications of glutamatergic projections to theventral tegmental area. Reviews in the Neurosciences 2008;19:227–44.

[68] Dickinson JA, Kew JN, Wonnacott S. Presynaptic alpha 7- and beta 2-containingnicotinic acetylcholine receptors modulate excitatory amino acid release fromrat prefrontal cortex nerve terminals via distinct cellular mechanisms. Molec-ular Pharmacology 2008;74:348–59.

[69] Livingstone PD, Srinivasan J, Kew JN, Dawson LA, Gotti C, Moretti M, et al.Alpha7 and non-alpha7 nicotinic acetylcholine receptors modulate dopaminerelease in vitro and in vivo in the rat prefrontal cortex. European Journal ofNeuroscience 2009;29:539–50.

[70] Konradsson-Geuken A, Gash CR, Alexander K, Pomerleau F, Huettl P, GerhardtGA, et al. Second-by-second analysis of alpha 7 nicotine receptor regulation ofglutamate release in the prefrontal cortex of awake rats. Synapse 2009;63:1069–82.

[71] Aracri P, Consonni S, Morini R, Perrella M, Rodighiero S, Amadeo A, et al. Tonicmodulation of GABA release by nicotinic acetylcholine receptors in layer v ofthe murine prefrontal cortex. Cerebral Cortex 2009.

[72] Gysling K, Wang RY. Morphine-induced activation of A10 dopamine neuronsin the rat. Brain Research 1983;277:119–27.

[73] Lokwan SJ, Overton PG, Berry MS, Clark D. The medial prefrontal cortex playsan important role in the excitation of A10 dopaminergic neurons followingintravenous muscimol administration. Neuroscience 2000;95:647–56.

[74] Ahmed SH, Graupner M, Gutkin B. Computational approaches to the neurobi-ology of drug addiction. Pharmacopsychiatry 2009;42(Suppl 1):S144–52.

[75] Everitt BJ, Robbins TW. Neural systems of reinforcement for drug addiction:from actions to habits to compulsion. Nature Neuroscience 2005;8:1481–9.

[76] Feil J, Sheppard D, Fitzgerald PB, Yucel M, Lubman DI, Bradshaw JL. Addiction,compulsive drug seeking, and the role of frontostriatal mechanisms in regu-lating inhibitory control. Neuroscience and Biobehavioral Reviews 2010;35:248–75.

[77] Kalivas PW. The glutamate homeostasis hypothesis of addiction. NatureReviews Neuroscience 2009;10:561–72.

[78] Daw ND, Niv Y, Dayan P. Uncertainty-based competition between prefrontaland dorsolateral striatal systems for behavioral control. Nature Neuroscience2005;8:1704–11.





































































































Documents

Cortical control of VTA function and influence on nicotine reward