Addiction - Making the Connection Between Behavioral Changes and Neuronal Plasticity in Specific...

8/6/2019 Addiction - Making the Connection Between Behavioral Changes and Neuronal Plasticity in Specific Patways

1/12

Addiction: Making theConnection Between BehavioraChanges and Neuronal Plasticity

in Specific Pathways

146

There is an emerging consensus that drug addiction is a form of maladaptive learning. Drugs of abuse usurp the neuronal circuitryinvolved in motivation and reward, leading to aberrant engagement

of learning processes. As a result, drug-associated cues can trigger

craving and compulsive drug-seeking behavior, and voluntary control

over drug use is lost. Abused drugs can also modulate long-term

potentiation (LTP) and long-term depression (LTD) in neuronal

circuits associated with the addiction process, suggesting a way for

the behavioral consequences of drug-taking to become reinforced by

learning mechanisms. This review will assess progress in correlating

these effects on LTP and LTD with behavioral changes in animal

models of addiction, particularly behavioral sensitization.

Marina E. Wolf Department of Neuroscience, The Chicago Medical School,

3333 Green Bay RoadNorth Chicago, IL 60064-3095

The image evokes the alluring yet sinister nature of cocaine. By usurping

neuronal pathways normally involved in motivated behavior, cocaine promotes "learning" of compulsive behavioral responses that underlie drug craving and

addiction.


2/12

INTRODUCTION

Cocaine and amphetamines enter the brain and interfere with thefunction of cell membrane transporters that normally removedopamine and other monoamines (serotonin and norepinephrine)from the synapse after these transmitters are released. Thisinterference produces a rapid elevation of dopamine levels thatresults in psychomotor stimulation: a feeling of being high.

Although these interactions are becoming better understood, itremains a mystery how this initial elevation in dopamine levelsleads, in some people, to a compulsive pattern of drug-seekingand drug-taking behavior. Even if abstinence is achieved, cocaineaddicts remain vulnerable for years to episodes of craving andrelapse triggered by stimuli previously associated with drugs (1, 2).This persistent vulnerability to drug-conditioned stimuli is also

observed in animal models of addiction (3).These features of addiction suggest that it may be anexceptionally powerful form of neuronal plasticity, which can bebroadly defined as the ability of the nervous system to modify itsresponse to a stimulus based on prior experience, and is believedto underlie learning and memory. Plasticity might also underlieaddiction, because signaling through glutamate, the keyneurotransmitter for producing and maintaining synaptic plasticity,is important for the formation of behavioral sensitizationaprominent animal model of addiction (4, 5). This role forglutamate receives further support from imaging studies of thebrains of human cocaine addicts: stimuli previously associatedwith drug use (e.g., drug paraphernalia) trigger intense drug

craving, and at the same time, activate glutamate-rich neuronalcircuits implicated in learning and memory (6). Animal studiesimplicate the same glutamate-rich circuits, and suggest that theprogression to addiction is a form of habit-based learning (7).Parallels between addiction and learning are reflected in thestriking similarities between the signal transduction cascades andmolecular adaptations associated with both processes (8). Forexample, chronic administration of amphetamine or cocaineproduces dramatic changes in the morphology of dendritic spinesin addiction-related brain regions, which closely resemble thoseassociated with learning (9, 10).

ADDICTION AND LONG-TERMPOTENTIATION

Learning and memory are thought to be encoded by changes inthe usage of interneuronal connections (11). In other words,synapses that experience frequent use will be strengthened,whereas those that receive less use are weakened. In this way, anexperience (which results in activation of pathways and increasedsynaptic use) may produce a long-lasting enhancement of communication between neurons in those pathways (resulting in amemory). For thirty years, long-term potentiation (LTP) has beenthe most compelling model for studying the synaptic basis of use-

dependent changes in the strength of interneuronal connections.In LTP, brief repetitive activation of excitatory glutamate-containing pathways (produced by high frequency stimulation, ortetanus) leads to an increase in the strength of glutamate synapsesthat lasts many hours in brain slices and even for weeks in intactanimals (12, 13). Thus, a test stimulus applied to a pathwaybefore the tetanus will produce a postsynaptic response of acertain magnitude, whereas the same test stimulus applied to apathway after tetanus will produce a larger response. Subsequentstudies showed that weaker patterns of stimulation produce anopposite change, termed long-term depression (LTD).

The cellular basis of LTP and LTD continues to be the focusof intense investigation. In general, an N-methyl- D-aspartate(NMDA) receptormediated increase in postsynaptic calciumconcentrations is required for the development of both LTD and

LTP, with small increases in calcium preferentially activatingprotein phosphatases and producing LTD, whereas larger increasesin calcium preferentially activate protein kinases and produce LTP(14). The expression of LTP involves enhanced transmissionthrough the -amino-3-hydroxy-5-methylisoxazole-4-propionate(AMPA) receptor in the postsynaptic cell. Important mechanismscontributing to the formation of LTP include phosphorylation of the AMPA receptor and increased insertion of AMPA receptors atsynaptic membranes, whereas AMPA receptor dephosphorylationand internalization may contribute to LTD (15, 16, 17).

LTP was first demonstrated in the hippocampus, a brain regionimportant for learning, which was part of the reason that LTP wasspeculated to underlie memory formation. However, many

investigators now question whether LTP plays a simple role inlearning and memoryparticularly the kinds of complex associativelearning that contribute to drug-seeking behavior and relapse (18).

Alternatively, LTP (and LTD) occur in many brain regions other thanthe hippocampus, suggesting that these phenomena representgeneral mechanisms for altering synaptic strength in response tochanges in synaptic use (regardless of whether this directly underliesmemory formation). We have hypothesized that drugs of abuse alterLTP or LTD in neuronal pathways, by altering activity in networksrelated to motivation and reward, or influence the ability of normal patterns of activity to produce appropriate forms of LTPand LTD (19). Abnormal engagement of LTP or LTD may be the first

step in the cascade leading to structural changes that underliepersistent modifications in brain function (20). Recent studies haveprovided general support for this hypothesis, showing that drugs of abuse can, in fact, modify or induce LTP and LTD in neuronalpathways related to addiction.

This review focuses on the challenging task of makingconnections between behavioral changes related to cocaine andamphetamine addiction and LTP or LTD in specific neuronalcircuits. After reviewing the neuronal circuits involved in drugaction, we then consider how addiction may be related to changesin neuronal pathways converging in the nucleus accumbens. Weknow that these pathways are important for addiction, but we are

Addiction and Neuronal Plasticity

1 June 2002

Volume 2, Issue 3


3/12

just beginning to understand how drugs of abuse modify theiractivity. Additionally, we will consider plasticity in the ventraltegmental area (VTA). Here, we are closer to making connectionsbetween plasticity and specific changes in neuronal activity thatcontribute to an animal model for addiction, behavioralsensitization.

NEURONAL PATHWAYS IMPORTANT IN

ADDICTION

The nucleus accumbens (also termed ventral striatum) plays acentral role in the neuronal circuits that are responsible formotivated, goal-directed behaviorsin other words, the type of behaviors that underlie compulsive drug-seeking in cocaineaddicts (Figure 1). Goal-directed behaviors are produced by

glutamate-containing projections that originate in limbic andcortical regions [most importantly, the basolateral amygdala ( BLA),the hippocampus and the prefrontal cortex] and converge on acommon postsynaptic target: the medium spiny neuron of thenucleus accumbens. The output of these nucleus accumbensneurons, conveyed through their projections to the ventralpallidum and ventral mesencephalon, is responsible for motorexecution of these goal-directed behaviors. Thus, the nucleusaccumbens serves as an interface between limbic and motorsystems (21, 22), and ultimately, drug-seeking behavior dependson glutamate transmission in the nucleus accumbens (23, 24).

Dopamine has long been the focus of addiction researchbecause cocaine and other stimulants act initially by increasing the

release of dopamine in the brain, and because dopamine neuronsparticipate in the elicitation of normal behaviors related tomotivation and reward. Indeed, dopamine is readily integratedinto glutamate-based models for addiction because glutamate- anddopamine-containing pathways converge in many brain regions,including the nucleus accumbens (Figure 1). Thus, the samenucleus accumbens neurons that receive limbic and corticalglutamate inputs also receive dopamine inputs from the VTA (25).Dopamine regulates interactions between glutamate inputs tonucleus accumbens neurons (26, 27), both by directly influencingsynaptic transmission and by modulating voltage-dependentconductance (28), in a manner complicated by the ability of thesame glutamate inputs to influence dopamine transmission (29).

VTA dopamine neurons also project to limbic and cortical regionsimportant for addiction, such as the prefrontal cortex and the

amygdala. Dopamine transmission in these regions modulates theactivity of addiction-related circuits (30) and is necessary for sometypes of drug-seeking behavior (31, 32, 33).

NUCLEUS ACCUMBENS GLUTAMATE AND

DRUG-SEEKING BEHAVIOR

Animal models have been used extensively to study the circuitsand neurotransmitters involved in drug-seeking behavior. Ratslearn to associate drug availability with a stimulus (such as aflashing light or a particular environment), and will performconditioned behaviors that are related to drug seeking whenpresented subsequently with that stimulus. Rats can also learn to

self-administer drugs(e.g., by pressing a barthat delivers anintravenous druginfusion) in responseto a priming injectionof drug or aconditioned stimulusthat was previouslyassociated with drugavailability.

Different

glutamate inputs tonucleus accumbensneurons have differentfunctions in drug-seeking behavior.Projections from theBLA to the nucleusaccumbens areimportant for learnedassociations betweenstimuli and drugreward, and for the

Review

148

PrefrontalCortex

GlutamateGABADopamine

NucleusAccumbens

Hippocampus

BLA

Thalamus

VentralPallidum

BehavioralOutput

VTA

Figure 1. The neural circuitry involved in drug addiction. The nucleus accumbens, an interface between limbic and motorsystems, plays a central role in this circuitry. Goal-directed behaviors are elicited by glutamate projections from limbic andcortical regions to the nucleus accumbens, while projections from the nucleus accumbens to motor regions are responsible formotor execution of these behaviors. Dopamine neurons located in the VTA innervate many limbic and cortical regionsimportant for addiction, including nucleus accumbens and prefrontal cortex. Dopamine and glutamate inputs typicallyconverge on common postsynaptic targets. BLA, basolateral amygdala; VTA, ventral tegmental area. [Based on (7).]


4/12

expression of motor behavior that is driven by motivationallyrelevant stimuli. These projections play a key role in drug-seekingbehavior elicited by discrete cues that were previously paired withdrug exposure (34, 35). On the other hand, drug-seeking behaviortriggered by injection of a low priming dose of cocaine does notrequire the amygdalaperhaps because this involves a response todrug itself, rather than to a conditioned stimulusbut doesrequire a circuit that includes the VTA, prefrontal cortex, nucleusaccumbens, and ventral pallidum (31, 34). The hippocampus maybe involved in conditioning to general contexts, such as aparticular cage or room environment previously associated withdrug availability. The direct and indirect connections of thehippocampus to the nucleus accumbens may enable suchcontextual cues to influence drug-seeking behavior (36), althoughthe role of the hippocampus in cue-elicited craving has been

studied far less extensively than that of the amygdala. Theprefrontal cortex is implicated in executive control of behavioraloutput based on stimulus value and expected outcome, and playsan important role in animal models of addiction (37). Excitatoryinputs from the prefrontal cortex, BLA, and hippocampus areintegrated by dopamine to determine the net level of excitatoryoutput to neurons in the nucleus accumbens (26, 38).

WHAT UNDERLIES THE TRANSITION TO

ADDICTION?

The dopamine neurons that participate in motivation and rewardhave their cell bodies in the VTA and project to a number of

limbic and cortical brain regions, including the nucleusaccumbens. These dopamine neuronsare activated by natural, rewardingstimuli such as food or sex, resulting inincreased dopamine release in targetbrain regions. Activation of dopamine-releasing neurons is important inpredicting the likelihood of reward whenan animal is presented with reward-related stimuli (39). Thus, one normalfunction of dopamine may be to helpconsolidate stimulus-response learning,

so individuals acquire the habit of pursuing those stimuli that are bothrewarding and necessary for survival.

If the same neuronal circuits areinvolved in responding for goodrewards such as food or sex, and badrewards such as cocaine, why is it thatcocaine, but usually not food or sex,leads to addiction? The answer isprobably related to the fact that cocaineand amphetamines increase dopaminerelease in a more prolonged and

unregulated manner than natural stimuli, because these drugsinteract with the dopamine transporter and interfere with theremoval of dopamine from the synaptic cleft. Over the span of several repeated drug exposures, compensatory changes occur indopamine-receptive neurons that presumably start a chain reactionof events that ultimately influences basic mechanisms of synapticplasticity. Possible mechanisms involved in this process have beenreviewed recently (19). The important point is that drugs of abuseusurp normal learning mechanisms and thereby cementbehavioral responses related to drug-seeking behavior. In somepeople, this progresses to a form of habit-based learning so strongthat it persists even in the face of tremendous adverse personalconsequences (7).

The behavioral changes that underlie the transition from drugexperimentation to compulsive drug-seeking behavior may be due

to synaptic plasticity in the neuronal circuits depicted in Figure 1.Based on what is known about the role of particular pathways indrug-seeking behavior, we can speculate about how plasticity inparticular pathways might contribute to various aspects of addiction (Figure 2). For example, strengthening the synapticconnections between the amygdala and the nucleus accumbensmay underlie the abnormally strong stimulus-reward learning thatoccurs in addiction, enabling drug-associated cues to elicit cravingand relapse even after long periods of abstinence. This increase inconditioned control of behavior may be enhanced by sensitizationof brain systems to the incentive-motivational effects of drugs, aneffect that may underlie the transition from wanting to craving,and at least initially, may involve LTP in VTA dopamine neurons.

Another important change may involve an increase in impulsive


1 June 2002

Volume 2, Issue 3

PrefrontalCortex

NucleusAccumbens

BLADrug-related cuesdrive behavior

Sensitization(wanting craving)

Loss of inhibitory control

Compulsivedrug-seeking anddrug-taking behavior

VTA

Figure 2. Possible relationships between neuronal plasticity in particular pathways and behavioralchanges that lead to addiction. Several types of behavioral changes may contribute to compulsive drug-seeking and drug-taking behavior in addicts. Alterations in transmission between the BLA and thenucleus accumbens may strengthen stimulus-reward learning, increasing the ability of drug-related cuesto control behavior. Sensitization of the incentive-motivational effects of drugs, due to alterations intransmission between the VTA and the nucleus accumbens, may transform drug wanting to craving.

Alterations in transmission between the prefrontal cortex and limbic targets, including nucleusaccumbens, may lead to impairment of inhibitory control mechanisms that normally govern reward-seeking behavior. Ultimately, drug-seeking behavior depends on glutamate transmission in the nucleusaccumbens. BLA, basolateral amygdala; VTA, ventral tegmental area. [Based on (37).]


5/12

behavior owing to the loss of inhibitory control mechanisms. Theprefrontal cortex is responsible for cognitive functions related toworking memory and planning. Normally, descending projectionsfrom the prefrontal cortex (to the nucleus accumbens, amygdalaand other brain regions) exert inhibitory control over reward-seeking behaviors. Recent evidence, from both rat and humanstudies, indicates that chronic cocaine exposure impairs theseinhibitory control mechanisms. Decreased inhibitory control,combined with enhanced responsiveness to drug-related cues andsensitization of drug wanting, may explain the compulsivenature of drug-seeking behavior in addicts (7, 35, 37).

Does intensification of stimulus-reward associations reflectLTP in pathways between BLA and nucleus accumbens? Does lossof inhibitory control reflect LTD in pathways between prefrontalcortex and its limbic targets? While it is not yet possible to answer

these questions, recent studies have made significant progresstowards characterizing LTP and LTD in reward-related brainregions, including the prefrontal cortex (40, 41) and amygdala(42). The remainder of this article will consider the role inaddiction of synaptic plasticity in the nucleus accumbens and the

VTA.

SYNAPTIC PLASTICITY IN THE NUCLEUS

ACCUMBENS

Both LTP and LTD are produced in the nucleus accumbens inresponse to repetitive stimulation of projections from theprefrontal cortex ( 43). LTP is also produced by repetitive

stimulation of hippocampal projections to the nucleus accumbens(44). NMDA receptor activation is required for both LTP and LTDin nucleus accumbens neurons (43, 45, 46). While LTP may bemodulated by dopamine, LTD is not affected by application of dopamine ( 46). It is notable that the properties of LTP and LTD inthe nucleus acumbens (ventral striatum) differ from those indorsal striatum (8, 19), where these processes have been bettercharacterized ( 47).

Neurons in the nucleus accumbens are normally quiescent,and their activation requires synchronous stimulation of multipleexcitatory inputs (38). Therefore, any drug-induced change thateither enhances or depresses their response to a particular input

will have a very significant influence on their activity. Recentstudies suggest that two kinds of changes may occur. First,modulatory effects of stimulants (or dopamine itself) on LTP maybe altered after repeated drug administration. Li and Kauerreported that the application of amphetamine blocked theinduction of LTP in nucleus accumbens slices prepared fromcontrol rats, whereas amphetamine did not block LTP in slicesprepared from rats pretreated with amphetamine for six days (48).

A relative shift towards LTP that may be attributable to loss of D2receptor-mediated inhibitory effects has also been reported instriatal slices prepared from rats chronically treated withmethamphetamine or ethanol (49, 50). Thus, the response of

nucleus accumbens or striatal neurons to drug challenge might bedifferent after repeated drug exposure because inhibitorymechanisms are dampened. The second kind of change mayinvolve an alteration in the strength of particular synapticconnections that long outlasts the period of drug administration.For example, nucleus accumbens slices prepared from mice twoweeks after discontinuing repeated cocaine treatment exhibitedLTD at excitatory synapses between cortical afferents and nucleusaccumbens neurons (51). This observation is consistent withprevious work indicating that the nucleus accumbens is lessexcitable after chronic stimulant exposure (19), and suggests thatdecreases in the cortical regulation of nucleus accumbens neuronalactivity could be related to the loss of inhibitory control that isimplicated in addiction and thought to involve prefrontal cortexdysfunction (37). Loss of inhibitory control at the level of the

prefrontal cortex may also contribute to the expression of behavioral sensitization (7).These are exciting findings; nonetheless, we are a long way

from understanding the relationship between LTP or LTD in aparticular pathway and a specific behavioral change in animalmodels or human studies of addictive behavior. Part of thedifficulty is that we do not yet understand how dopamine andglutamate work together to determine the output of nucleusaccumbens neurons under normal conditions (28, 52), let aloneafter chronic drug administration. In contrast, we are closer tounderstanding how plasticity in the VTA may contribute tobehavioral sensitization, an animal model of addiction.

GLUTAMATE AND BEHAVIORAL

SENSITIZATION

Behavioral sensitization refers to the progressive enhancement of species-specific behavioral responses that occurs during repeateddrug administration and persists even after long periods of withdrawal. Sensitization occurs in humans (53), but most studieshave focused on sensitization of locomotor activity in rodents.There is compelling evidence that locomotor sensitization providesa useful model for sensitization of drug craving. Both depend on

VTA dopamine neurons projecting to limbic and cortical regions(Figure 1). Prior exposure to cocaine or amphetamine, resulting in

locomotor sensitization, promotes drug self-administration andenhances stimulus-reward learning and responding forconditioned reward (7). Finally, both sensitization in rats and drugcraving in humans are strongly modulated by environmentalstimuli, conditioning and stress (54, 55). Once established,sensitization is very long-lastingamphetamine sensitization canlast up to a year in rats, a species that lives only one or two years(56)and is accompanied by dramatic changes in addiction-related neuronal circuits at molecular, cellular, and systems levels(7, 5759).

Robinson and Berridge have developed an incentive-sensitization theory of addiction (60, 61) that remains one of the

Review

150


6/12

most influential in the field. A key element of this theory is thatliking and wanting are produced by distinct neuronalmechanisms, and it is wanting that intensifies with repeated druguse. In other words, repeated drug exposure does not increase thepleasure produced by a drug, but does increase craving. Accordingto this theory, addictive drugs produce long-lasting adaptations inbrain systems mediating incentive-motivational effects, leading tosensitization of these systems to drugs and, importantly, to cuesthat trigger pursuit of rewarding stimuli (62). Sensitization of incentive-motivational effects would promote compulsive behavior,perhaps by contributing to the enhanced stimulus-reward learningdiscussed above in relation to projections from the amygdala tothe nucleus accumbens (Figure 2). In fact, although we will focuson VTA below, the amygdala might also participate in theinduction of sensitization (57).

SYNAPTIC PLASTICITY IN THE VTA

The circuitry involved in sensitization is complex, becausesensitization represents a cascade of events involving differentbrain regions and different transmitter systems at differentwithdrawal times. However, microinjection studies haveestablished that the initiation of sensitization to cocaine oramphetamine occurs in the VTA. Thus, direct injection of amphetamine into the VTA has no acute effect on locomotoractivity, but repeated intra-VTA administration results in asensitized locomotor response to systemic injections of amphetamine, cocaine, or morphine. In contrast, sensitization is

expressed in the nucleus accumbens, because direct injection of amphetamine into nucleus accumbens is sufficient to elicit asensitized response in rats that received repeated systemic or intra-

VTA injections (57, 58).The anatomical separation of sites for initiation and for

expression implies a transfer of sensitization from the VTA to thenucleus accumbens, presumably as a result of changes in the firingrate or pattern of dopamine neurons projecting from VTA tonucleus accumbens. Indeed, the firing rate of VTA dopamineneurons is increased for several days after discontinuing repeatedadministration of cocaine or amphetamine. Firing rates thennormalize within a week or so. This transient activation of

dopamine-releasing neurons is believed to be a key step in thedevelopment of long-lasting behavioral sensitization. Themechanism underlying the activation may involve increasedglutamate transmission in the VTA during drug administration andthe early withdrawal period. This idea is consistent with studiesshowing that the development of sensitization is prevented byblocking glutamate transmission in the VTA, or by lesioning theprefrontal cortex, which sends glutamate projections to the VTA(57, 58).

One way to account for these findings is to propose thatrepeated drug administration results in LTP at synapses betweenglutamate nerve terminals and VTA dopamine neurons, leading to

a transient increase in the firing rate of dopamine neurons (57).This, in turn, would increase dopamine release in the nucleusaccumbens and other addiction-related brain regions that are thetargets of dopamine projections (including prefrontal cortex,amygdala and hippocampus). Drug-induced changes in dopaminerelease may be one of the triggers for the persistentelectrophysiological and neurochemical changes observed in theseregions after discontinuing drug use ( 57, 58).

Once it has developed, LTP is expressed as an enhancementof postsynaptic AMPA receptor transmission. Thus, if the initiationof behavioral sensitization is associated with LTP at glutamatesynapses on VTA dopamine neurons, and if LTP is responsible fortransient activation of dopamine cell firing during earlywithdrawal, then dopamine neurons in sensitized rats shouldexhibit increased responsiveness to AMPA at early withdrawal

times, but not at later times. This prediction has been confirmedin two ways. First, VTA dopamine neurons recorded from chroniccocaine- or amphetamine-pretreated rats are more responsive tothe excitatory effects of AMPA (63). Second, intra-VTAadministration of AMPA produces a greater increase in nucleusaccumbens dopamine levels in rats previously treated withamphetamine (64). The latter study suggests that the enhancementof AMPA receptor transmission occurs in those VTA dopamineneurons that project to the nucleus accumbens. In both studies,no change in responsiveness to NMDA was observed and theincreased responsiveness to AMPA was evident in rats tested threedays after discontinuing drug administration but not at later timeswhen dopamine cell firing rates have normalized.

But can a drug regimen that produces sensitization in a wholeanimal influence synaptic plasticity in the VTA? In preparedmidbrain slices from control mice and from mice injected the daybefore with a single injection of saline or cocaine, the cocainetreatmentwhich produces one-shot behavioral sensitizationproduced an enhancement of AMPA receptor transmission in VTAdopamine neurons that was attributable to LTP (65). Cocaine-induced LTP was observed in slices prepared five days, but not tendays, after cocaine injection. Thus, cocaine-induced LTP is atransient phenomenon, similar to the enhancement of AMPAreceptor transmission detected in vivo (63, 64). Furthermore, LTPwas blocked by an NMDA receptor antagonist, as was

sensitization. These findings argue strongly for linkage betweencocaine-induced LTP and sensitization.How do dopamine-releasing psychomotor stimulants promote

LTP in the VTA? Amphetamine or dopamine, acting through D2dopamine receptors, can block the induction of LTD in midbraindopamine neurons, through a mechanism that may involveinhibition of high threshold calcium channels (46, 66). InhibitingLTD through cocaine-dependent stimulation of dopaminereceptors may increase neuronal excitability and thus promote LTP.LTP may also be promoted by stimulant-induced increases inglutamate levels in the VTA, although exactly how cocaine andamphetamine modulate VTA glutamate levels is a matter of some


1 June 2002

Volume 2, Issue 3


7/12

debate (19). Dopamine receptor stimulation may also increase theexcitability of midbrain dopamine neurons by reducing inhibitoryeffects of metabotropic glutamate receptors (67).

Based on results summarized above, a model for the initiationof sensitization in the VTA should include: 1) a role for LTPinduction in VTA dopamine neurons, 2) a key role for VTAdopamine neurons projecting to the nucleus accumbens, and 3) anecessary role for prefrontal cortex (recall that lesions of prefrontalcortex prevent the development of sensitization). The simplestsolution would be to propose that the LTP leading to sensitizationoccurs at synapses between prefrontal cortex terminals and VTAdopamine neurons projecting to nucleus accumbens.Unfortunately, this hypothesis is not consistent with recentanatomical findings. Sesack and colleagues have shown thatprefrontal cortex neurons do not make excitatory synapses on

dopamine-releasing VTA neurons that project to the nucleusaccumbens; however, they do synapse on -aminobutyric acid(GABA)-releasing VTA neurons that project to the nucleusaccumbens. Conversely, prefrontal cortex neurons synapse on VTAdopamine neurons, but not on VTA GABA neurons, that projectback to the prefrontal cortex (68) (Figure 3). Neurochemicalfindings are consistent with this anatomical arrangement ( 69).

An alternative hypothesis is that the LTP important forsensitization occurs at synapses between prefrontal cortex neuronsand VTA dopamine neurons projecting to the prefrontal cortex(site 1 in Figure 3), consistent with evidence for a direct excitatorylink between these populations. In this model, changes inprefrontal cortex occur first, and then trigger longer-lived changes

in the nucleus accumbens through the many interconnections thatexist between the two regions (Figure 1). Indeed, the prefrontalcortex is necessary for cocaine-induced adaptations occurring inthe nucleus accumbens (70), and there is evidence for changes inboth glutamate and dopamine transmission in the prefrontalcortex shortly after discontinuing repeated cocaine oramphetamine administration (7). However, this proposed site forLTP (site 1 in Figure 3) is not consistent with evidence that AMPAreceptor transmission is enhanced in those VTA dopamine neuronsthat project to nucleus accumbens (64).

Another possibility is that VTA dopamine neurons projecting tonucleus accumbens are excited by glutamate (and cholinergic)

projections from the laterodorsal tegmentum (LDT) and neighboringmesopontine nuclei (71). Thus, the LTP important for initiatingsensitization might occur at synapses between the LDT and VTAdopamine neurons projecting to nucleus accumbens (site 2 in Figure3). Although promising, this hypothesis raises questions about howthe prefrontal cortex is brought into the circuit. The prefrontal cortexprojects to the LDT, and it has been proposed that this indirect routeto the VTA enables the prefrontal cortex to exert excitatory effects onmidbrain dopamine neurons (72). This hypothesis is consistent withexperiments showing that the prefrontal cortex activates VTAdopamine neurons through polysynaptic pathways rather thanthrough a direct monosynaptic projection (73) (Figure 4). However,

even this indirect route is not consistent with all findings, as a recentstudy showed that stimulation of the prefrontal cortex atphysiologically relevant frequencies actually inhibits dopaminerelease in the nucleus accumbens (74). Of course, the prefrontal

Review

152

Figure 3. Possible sites for LTP in the ventral tegmental area duringthe induction of behavioral sensitization. The VTA contains dopamineneurons that project to nucleus accumbens (mesoaccumbens dopamineneurons) and dopamine neurons that project to prefrontal cortex(mesoprefrontal dopamine neurons). Likewise, the VTA containsmesoaccumbens GABA neurons and mesoprefrontal GABA neurons. VTAdopamine and GABA neurons also innervate other brain regions notshown in the diagram. Glutamate projections from the prefrontal cortexsynapse on mesoaccumbens GABA neurons but not mesoaccumbens DAneurons, and on mesoprefrontal DA neurons but not mesoprefrontal GABAneurons (68). Some evidence suggests that prefrontal cortex alsocommunicates with mesoaccumbens dopamine neurons indirectly, viaprojections to the LDT. The LDT sends both glutamate and cholinergic

projections to mesoaccumbens dopamine neurons (71). The cholinergicfibers are not shown in the diagram. Based on this anatomy and otherevidence, two possible sites are proposed for cocaine-induced LTP leadingto the initiation of behavioral sensitization. Site 1: Excitatory synapsesbetween prefrontal cortical projections and mesoprefrontal dopamineneurons. Site 2: Excitatory synapses between LDT projections andmesoaccumbens dopamine neurons. VTA, ventral tegmental area; LDT,laterodorsal tegmentum. [Based on (68).]

Prefrontal

Cortex

GlutamateGABADopamine

Nucleusccumbens

LDT

V T A

21


8/12

cortex could be brought into the circuit later, after changes haveoccurred in the pathway from LDT to VTA to nucleus accumbens,through interconnections between these regions (Figure 1).

A cornerstone of the above hypotheses is that the inductionof sensitization is prevented by blocking NMDA receptors in the

VTA, which also prevents the induction of LTP by systemiccocaine (65), but it is possible that the two are not linked. Analternative possibility is that the important effect of NMDAreceptor blockade in the VTA is to alter the activity of VTA GABAneurons that project locally as well as to the nucleus accumbensand other forebrain regions ( 57, 75). Unlike dopamine neurons,GABA neurons of the VTA do not exhibit LTP under normalconditions (76) or after cocaine treatment (65). While VTA GABA

neurons can exhibit LTD (66), effects of cocaine on this processhave yet to be examined. Nevertheless, there is evidence that theactivity of these GABA neurons is altered by repeated drugadministration. For example, GABA B receptor transmission in the

VTA is enhanced after withdrawal from amphetamine and otherdrugs (77). Another exciting finding is that brief (two or threeminutes) co-activation of group I metabotropic and NMDAglutamate receptors in the ventral midbrain produced a long-lasting (greater than one hour) enhancement in the rhythmicity of GABAergic inhibitory synaptic activity (78). This effect involveddirect electrical coupling of neurons through gap junctions ( 78), aphenomenon that is altered by chronic amphetamine treatment in

other brain regions (79). A role for this mechanism in sensitizationwould account for the fact that the development of sensitization isinhibited by administration of either NMDA or metabotropicglutamate receptor antagonists into the VTA (80). Alterations ininhibitory tone, which controls basal firing characteristics of VTAdopamine neurons, would be expected to have major effects onthe response of these neurons to excitatory inputs (29).

FUTURE DIRECTIONS

The demonstration that drugs of abuse alter LTP and LTD inreward-related neuronal pathways is a very exciting development.


1 June 2002

Volume 2, Issue 3

PrefrontalCortex

GlutamateGABA

Dopamine

NucleusAccumbens

LaterodorsalTegmentum

(LDT)

VentralTegmental Area

(VTA)

Figure 4. Addiction-relatedpathways in the rat brain. The VTAcontains both dopamine and GABAneurons that innervate the nucleus

accumbens, prefrontal cortex, andother forebrain targets not shown inthe diagram. The prefrontal cortexsends glutamate projections to VTA,LDT, nucleus accumbens, and otherlimbic regions not shown in thediagram (see Figure 1). In the VTA,glutamate inputs from prefrontalcortex synapse on mesoaccumbensGABA neurons and mesoprefrontalDA neurons. Mesoaccumbensdopamine neurons also receiveglutamate and cholinergic inputs fromthe LDT (cholinergic inputs notshown). See legend to Figure 3 for

more details.

Figure 5. Dopamine regulates cell surface AMPA receptor expression.Insertion and removal of AMPA receptors from synaptic sites is importantfor LTP and LTD (16,17). In cultured nucleus accumbens neurons, D1dopamine receptor stimulation increased the density of AMPA receptorpuncta on the cell surface, suggesting a mechanism by which dopaminemight influence LTP and LTD ( 83). Primary cultures were prepared fromnucleus accumbens of postnatal day one rats. AMPA receptors on the cellsurface were visualized by incubating live cells (twenty-one days in vitro)

with antibody that recognizes the extracellular portion of the AMPAreceptor subunit GluR1, fixing cultures, and then incubating withfluorescent secondary antibody. To quantify GluR1 surface expression, astandard length of neurite was defined using a circle 25 M in diameterand the number of cell surface GluR1 puncta on a single neurite withinthe circle was counted using MetaMorph Imaging software. Left panel:Control conditions. Right panel: Incubation with the D1 agonist SKF81297 (1 M, for fifteen minutes) increased the density of surface GluR1puncta. Nucleus accumbens interneurons are shown in these pictures, butthe same results were obtained for medium spiny neurons, the projectionneurons of the nucleus accumbens (83). In the same culture system, D1receptor stimulation increased phosphorylation of GluR1 at the proteinkinase A site, suggesting another way that dopamine may influence basicmechanisms of synaptic plasticity ( 81).


9/12


10/12

17. Sheng, M. and Lee, S.H. AMPAreceptor trafficking and the control of synaptic transmission. Cell105 ,

825828 (2001).18. Martin, S.J., Grimwood, P.D., and

Morris, R.G.M. Synaptic plasticity andmemory: An evaluation of thehypothesis. Annu. Rev. Neurosci.23 ,649711 (2000).

19. Wolf, M.E. Addiction and glutamate-dependent plasticity, in Glutamate and Addiction(Herman, B.H., Frankenheim, J., Litten, R., Sheridan, P.H., Weight,F.F. and Zukin, S.R. eds), The HumanaPress, Inc. (in press).

20. Lscher, C., Nicoll, R.A., Malenka,R.C., and Muller, D. Synaptic plasticityand dynamic modulation of thepostsynaptic membrane. Nat. Neurosci.3, 545550 (2000).

21. Kelley, A.E. Neural integrative activitiesof nucleus accumbens subregions inrelation to learning and motivation.Psychobiology27 , 198213 (1999).

22. Groenewegen, H.J., Wright, C.I., Beijer, A.V., and Voorn, P. Convergence andsegregation of ventral striatal inputsand outputs. Ann. N.Y. Acad. Sci.877 ,4964 (1999).

22. Cornish, J.L. and Kalivas, P.W.Glutamate transmission in the nucleusaccumbens mediates relapse in cocaineaddiction. J. Neurosci.20 , RC89 (2000).

23. Di Ciano, P., Cardinal, R.N., Cowell,R.A., Little, S.J., and Everitt B.J.Differential involvement of NMDA,

AMPA/kainate, and dopamine receptorsin the nucleus accumbens core in theacquisition and performance of Pavlovian approach behavior. J.Neurosci.21 , 94719477 (2001).

24. Sesack, S.R. and Pickel, V.M. Prefrontalcortical efferents in the rat synapse onunlabeled neuronal targets of catecholamine terminals in the nucleusaccumbens septi and on dopamineneurons in the ventral tegmental area. J. Comp. Neurol.320 , 145160 (1992).

25. Floresco, S.B., Blaha, C.D., Yang, C.R.and Phillips, A.G. Modulation of hippocampal and amygdalar-evoked

activity of nucleus accumbens neuronsby dopamine: Cellular mechanisms of input selection. J. Neurosci.21 ,28512860 (2001).

26. Floresco, S.B., Blaha, C.D., Yang, C.R.,and Phillips, A.G. Dopamine D1 andNMDA receptors mediate potentiationof basolateral amygdala-evoked firingof nucleus accumbens neurons. J.Neurosci.21 , 63707376 (2001).

27. Nicola, S.M., Surmeier, D.J., andMalenka, R.C. Dopaminergicmodulation of neuronal excitability inthe striatum and nucleus accumbens. Annu. Rev. Neurosci.232 , 185215(2000).

28. Floresco, S.B., Todd, C.L. and Grace, A.A. Glutamatergic afferents from thehippocampus to the nucleusaccumbens regulate activity of ventraltegmental area dopamine neurons. J.Neurosci.21 , 49154922 (2001).

29. Rosenkranz, J.A. and Grace, A.A.

Dopamine attenuates prefrontal corticalsuppression of sensory inputs to thebasolateral amygdala of rats. J. Neurosci.21 , 40904103 (2001).

30. McFarland, K. and Kalivas, P.W. Thecircuitry mediating cocaine-inducedreinstatement of drug-seeking behavior. J. Neurosci.21 , 86558663 (2001).

31. Ciccocioppo, R., Sanna, P.P., and Weiss,F. Cocaine-predictive stimulus inducesdrug-seeking behavior and neuralactivation in limbic brain regions aftermultiple months of abstinence:Reversal by D1 antagonists. Proc. Natl. Acad. Sci. U.S.A.98 , 19761981(2001).

32. See, R.E., Kruzich, P.J., and Grimm, J.W. Dopamine, but not glutamate,receptor blockade in the basolateralamygdala attenuates conditionedreward in a rat model of relapse tococaine-seeking behavior.Psychopharmacology154 , 301310

(2001).

33. Grimm, J.W. and See, R.E. Dissociationof primary and secondary reward-relevant limbic nuclei in an animalmodel of relapse.Neuropsychopharmacology22 , 473479(2000).

34. Everitt, B.J., Dickinson, A.D., andRobbins, T.W. The neuropsychologicalbasis of addictive behaviour. Brain Res.Rev. 36 , 129138 (2001).

35. Vorel, S.R., Liu, X., Hayes, R.J., Spector, J.A., and Gardner E.L. Relapse tococaine-seeking after hippocampal

theta burst stimulation. Science292 ,11751178 (2001).

36. Jentsch, J.D. and Taylor, J.R.Impulsivity resulting from frontostriataldysfunction in drug abuse: Implicationsfor the control of behavior by reward-related stimuli. Psychopharmacology146 , 373390 (1999).

37. ODonnell, P. and Grace, A.A. Synapticinteractions among excitatory afferentsto nucleus accumbens neurons:Hippocampal gating of prefrontalcortical input. J. Neurosci.15 ,36223639 (1995).

38. Schultz, W. and Dickinson, A.Neuronal coding of prediction errors. Annu. Rev. Neurosci.21 , 473500(2000).

39. Otani, S., Auclair, N., Desce, J.-M.,Roisin, M.-P., and Crpel, F. Dopaminereceptors and groups I and II mGluRscooperate for long-term depressioninduction in rat prefrontal cortex

through converging postsynapticactivation of MAP kinases. J. Neurosci.19 , 97889802 (1999).

40. Gurden, H., Takita, M. and Jay, T.M.Essential role of D1 but not D2receptors in the NMDA receptor-dependent long-term potentiation athippocampal-prefrontal cortex synapsesin vivo. J. Neurosci.20 , RC106 (2000).

41. Maren, S. Long-term potentiation inthe amygdala: A mechanism for


1 June 2002

Volume 2, Issue 3


11/12

emotional learning and memory. TrendsNeurosci.22 , 561567 (1999).

42. Pennartz, C.M.A., Ameerun, R.F.,Groenewegen, H.J., and Lopes da Silva,F.H. Synaptic plasticity in an in vitroslice preparation of the rat nucleusaccumbens. Eur. J. Neurosci.5, 107117(1993).

43. Boeijinga, P.H., Mulder, A.B., Pennartz,C.M.A., Manshanden, I., and Lopes daSilva, F.H. Responses of the nucleusaccumbens following fornix/fimbriastimulation in the rat. Identificationand long-term potentiation of mono-and polysynaptic pathways.Neuroscience53 , 10491058 (1993).

44. Kombian, S.B. and Malenka, R.C.Simultaneous LTP of non-NMDA- andLTD of NMDA-receptor-mediatedresponses in the nucleus accumbens.Nature 368 , 242246 (1994).

45. Thomas, M.J., Malenka, R.C., andBonci, A. Modulation of long-termdepression by dopamine in themesolimbic system. J. Neurosci.20 ,55815586 (2000).

46. Calabresi, P., Centonze, D., Gubellini,P., Marfia, G.A., Pisani, A., Sancesario,G., and Bernardi, G. Synaptictransmission in the striatum: Fromplasticity to neurodegeneration. Prog.Neurobiol.61 , 231265 (2000).

47. Li, Y. and Kauer, J. Effects of amphetamine on long-termpotentiation in the nucleus accumbens.Soc. Neurosci. Abstr.26 , 1398 (2000).

48. Yamamoto, Y., Nakanishi, H., Takai, N.,Shimazoe, T., Watanabe, S., and Kita,H. Expression of N-methyl- D-aspartatereceptor-dependent long-termpotentiation in the neostriatal neuronsin an in vitro slice after ethanolwithdrawal of the rat. Neuroscience91 ,5968 (1999).

49. Nishioku, T., Shimazoe, T., Yamamoto, Y., Nakanishi, H., and Watanabe, S.Expression of long-term potentiation of the striatum in methamphetamine-

sensitized rats. Neurosci. Lett.268 ,8184 (1999).

50. Thomas, M.F., Beurrier, C., Bonci, A.,and Malenka, R. Long-term depressionin the nucleus accumbens: A neuralcorrelate of behavioral sensitization tococaine. Nat. Neurosci.4, 12171223(2001).

51. West, A.R. and Grace, A.A. Oppositeinfluences of endogenous D1 and D2receptor activation on activity statesand electrophysiological properties of striatal neurons: Studies combining invivo intracellular recordings andreverse microdialysis. J. Neurosci.22 ,294304 (2002).

52. Strakowski, S.M. and Sax, K.W.,Progressive behavioral response torepeated d-amphetamine challenge:Further evidence for sensitization inhumans. Biol. Psych.44 , 11711177(1998).

53. Robinson, T.E., Browman, K.E.,Crombag, H.S., and Badiani, A.Modulation of the induction orexpression of psychostimulant

sensitization by the circumstancessurrounding drug administration.Neurosci. Biobehav. Rev.22 , 347354(1998).

54. Shaham, Y., Erb, S., and Stewart, J.Stress-induced relapse to heroin andcocaine seeking in rats: A review. BrainRes. Rev.33 , 1333 (2000).

55. Paulson, P.E., Camp, D.M., andRobinson, T.E. The time course of transient behavioral depression and

persistent behavioral sensitization inrelation to regional monoamineconcentrations during amphetaminewithdrawal in rats. Psychopharmacol.103 , 480492 (1991).

56. Wolf, M.E. The role of excitatory aminoacids in behavioral sensitization topsychomotor stimulants. Prog.Neurobiol.54 , 679720 (1998).

57. Vanderschuren, L.J. and Kalivas, P.W. Alterations in dopaminergic and

glutamatergic transmission in theinduction and expression of behavioralsensitization: A critical review of

preclinical studies. Psychopharmacology151 , 99120 (2000).

58. Nestler, E.J. Molecular basis of long-term plasticity underlying addiction.Nat. Rev. Neurosci.2, 119128 (2001).

59. Robinson, T.E. and Berridge, K.C. Theneural basis of drug craving: Anincentive-sensitization theory of addiction. Brain Res. Rev.18 , 247291(1993).

60. Robinson, T.E. and Berridge, K.C. The

psychology and neurobiology of addiction: An incentive-sensitizationview. Addiction95 , S91S117 (2000).

61. Wyvell, C.L. and Berridge, K.C.Incentive sensitization by previousamphetamine exposure: Increased cue-triggered wanting for sucrose reward. J. Neurosci.21 , 78317840 (2001).

62. Zhang, X.-F., Hu, X.-T., White, F.J., and Wolf, M.E. Increased responsiveness of ventral tegmental area dopamineneurons to glutamate after repeatedadministration of cocaine oramphetamine is transient andselectively involves AMPA receptors. J.Pharmacol. Exp. Ther. 281 , 699706(1997).

63. Giorgetti, M., Hotsenpiller, G., Ward,P., Teppen, T., and Wolf, M.E.

Amphetamine-induced plasticity of AMPA receptors in the ventraltegmental area: Effects on extracellularlevels of dopamine and glutamate in

freely moving rats. J. Neurosci.21 ,63626369 (2001).

64. Ungless, M.A., Whistler, J.L., Malenka,R.C., and Bonci, A. Single cocaineexposure in vivo induces long-termpotentiation in dopamine neurons.Nature 411 , 583587 (2001).

65. Jones, S., Kornblum, J.L., and Kauer, J.A. Amphetamine blocks long-termsynaptic depression in the ventraltegmental area. J. Neurosci.20 ,

Review

156


12/12

55755580 (2000).

66. Paladini, C.A., Fiorillo, C.D.,Morikawa, H., and Williams, J.T.

Amphetamine selectively blocksinhibitory glutamate transmission indopamine neurons. Nat. Neurosci.4,275281 (2001).

67. Carr, D.B. and Sesack, S.R. Projectionsfrom the rat prefrontal cortex to theventral tegmental area: Targetspecificity in the synaptic associationswith mesoaccumbens and mesocorticalneurons. J. Neurosci.20 , 38643873(2000).

68. Takahata, R. and Moghaddam, B.Target-specific glutamatergic regulationof dopamine neurons in the ventraltegmental area. J. Neurochem.75 ,17751778 (2000).

69. Li, Y., Hu, X.-T., Berney, T.G., Vartanian, A.J., Stine, C., Wolf, M.E.,and White, F.J. Both glutamate receptorantagonists and prefrontal cortexlesions prevent the induction of cocaine sensitization and associatedneuroadaptations. Synapse34 ,169180

(1999).70. Forster, G.L. and Blaha, C.D.

Laterodorsal tegmental stimulationelicits dopamine efflux in the ratnucleus accumbens by activation of acetylcholine and glutamate receptorsin the ventral tegmental area. Eur. J.Neurosci.12 , 35963604 (2000).

71. Lokwan, S.J.A., Overton, P.G., Berry,M.S., and Clark, D. Stimulation of thepedunculopontine tegmental nucleus inthe rat produces burst firing in A9

dopaminergic neurons. Neuroscience92 ,245254 (1999).

72. Tong, Z.-Y., Overton, P.G., and Clark,D. Stimulation of the prefrontal cortexin the rat induces patterns of activity inmidbrain dopaminergic neurons whichresemble natural burst events. Synapse22 , 195208 (1996).

73. Jackson, M.E., Frost, A.S. andMoghaddam, B. Stimulation of

prefrontal cortex at physiologicallyrelevant frequencies inhibits dopaminerelease in the nucleus accumbens. J.

Neurochem.78 , 920923 (2001).74. Cornish, J.L., Nakamura, M., and

Kalivas, P.W. Dopamine-independentlocomotion following blockade of N-methyl- D-aspartate receptors in theventral tegmental area. J. Pharmacol.Exp. Ther. 298 , 226233 (2001).

75. Bonci, A. and Malenka, R.C. Propertiesand plasticity of excitatory synapses ondopaminergic and GABAergic cells inthe ventral tegmental area. J. Neurosci.19 , 37233730 (1999).

76. Giorgetti, M., Hotsenpiller, G., Froestl, W., and Wolf, M.E. In vivomodulationof ventral tegmental area dopamine andglutamate efflux by local GABABreceptors is altered after repeatedamphetamine treatment. Neuroscience109 , 585595 (2002).

77. Beretta, N., Paolucci, E., Bernardi, G.,and Mercuri, N.B. Glutamate receptorstimulation induces a persistentrhythmicity of the GABAergic inputs to

rat midbrain dopaminergiic neurons.Eur. J. Neurosci.14 , 777784 (2001).

78. Onn, S.P. and Grace, A.A. Amphetamine withdrawal altersbistable states and cellular coupling inrat prefrontal cortex and nucleusaccumbens neurons recorded in vivo. J.Neurosci.20 , 23322345 (2000).

79. Kim, J.-H. and Vezina, P. Metabotropicglutamate receptors are necessary forsensitization by amphetamine.Neuroreport 9, 403306 (1998).

80. Chao, S.Z., Lu, W., Lee, H.-K., Huganir,R.L., and Wolf, M.E. D1 dopaminereceptor stimulation increases GluR1phosphorylation in postnatal nucleusaccumbens cultures. J. Neurochem., inpress (2002).

81. Bibb, J.A., Chen, J., Taylor, J.R.,Svenningsson, P., Nishi, A., Snyder,G.L., Yan, Z., Sagawa, Z.K., Ouimet,C.C., Nairn, A.C., Nestler, E.J., and

Greengard, P. Effects of chronicexposure to cocaine are regulated bythe neuronal protein Cdk5. Nature

410 , 376380 (2001).82. Chao, S.Z., Peterson, D.A., and Wolf,

M.E. Dopamine and glutamatereceptors regulate surface expression of GluR1 in postnatal nucleus accumbenscultures. Soc. Neurosci. Abstr. 26 , 789(2000).

83. Dewey, S.L., Morgan, A.E., Ashby Jr.,C.R., Horan, B., Kushner, S.A., Logan,

J., Volkow, N.D., Fowler, J.S., Gardner,E.L. and Brodie, J.D. A novel strategyfor the treatment of cocaine addiction.Synapse30 , 119129 (1998).


June 2002

Marina E. Wolf, PhD, is a Professor and isthe acting-chair in the Department of Neuroscience at the Finch University of Health Sciences, Chicago Medical School.Please address correspondence to MEW. E-mail: [email protected]; Fax

(847)-578-3428

Addiction - Making the Connection Between Behavioral Changes and Neuronal Plasticity in Specific...

Documents

Transcript of Addiction - Making the Connection Between Behavioral Changes and Neuronal Plasticity in Specific...